The DOI matches its Crossref record; 65 of the 65 references it cites were found in a registry or an index; 32 of the 39 sources its claims cite were checked for support.
High-performance brain-to-text communication via handwriting
Open accessPublished version of a preprint
Abstract
Brain-computer interfaces (BCIs) can restore communication to people who have lost the ability to move or speak. So far, a major focus of BCI research has been on restoring gross motor skills, such as reaching and grasping 1-5 or point-and-click typing with a computer cursor 6,7. However, rapid sequences of highly dexterous behaviours, such as handwriting or touch typing, might enable faster rates of communication. Here we developed an intracortical BCI that decodes attempted handwriting movements from neural activity in the motor cortex and translates it to text in real time, using a recurrent neural network decoding approach. With this BCI, our study participant, whose hand was paralysed from spinal cord injury, achieved typing speeds of 90 characters per minute with 94.1% raw accuracy online, and greater than 99% accuracy offline with a general-purpose autocorrect. To our knowledge, these typing speeds exceed those reported for any other BCI, and are comparable to typical smartphone typing speeds of individuals in the age group of our participant (115 characters per minute)8. Finally, theoretical considerations explain why temporally complex movements, such as handwriting, may be fundamentally easier to decode than point-to-point movements. Our results open a new approach for BCIs and demonstrate the feasibility of accurately decoding rapid, dexterous movements years after paralysis.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
20 Own result · 11 Own interpretation · 7 Supported · 7 Partly supported · 3 No basis given · 2 No full text of the sources · 1 Supported so far · 1 Not supported · 1 Own method · accuracy not yet measured
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“During parameter optimization, the cursor gain is increased so as to increase typing rate, until the cursor moves so quickly that it becomes uncontrollable due to decoding errors22.”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
What the sources say
- Partly supportedcites [22] Willett, F. R. et al. Signal-independent noise in intracortical brain–computer interfaces causes movement time properties inconsistent with Fitts’ law. J. Neural Eng. 14, 026010 (2017). sourceExperimental — accuracy not yet measured
The Results section, “Effect of gain on movement time” (page unavailable), reports that increasing gain initially shortened target-acquisition time, but higher gains made cursor movements fast, inaccurate, and uncontrollable around targets. The paper relates this difficulty to decoding noise, but it studies cursor target acquisition, not typing rate, and does not describe the claimed typing-speed optimization procedure. (LLM, conf=0.85)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The speed of point-and-click BCIs is limited primarily by decoding accuracy.”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
What the sources say
- Partly supportedcites [22] Willett, F. R. et al. Signal-independent noise in intracortical brain–computer interfaces causes movement time properties inconsistent with Fitts’ law. J. Neural Eng. 14, 026010 (2017). sourceExperimental — accuracy not yet measured
The Results (Effect of gain on movement time; page unavailable) report that increasing cursor speed made movements fast but inaccurate, and the Introduction identifies signal-independent decoding noise as the root cause of the movement-time properties observed. This supports a speed–accuracy constraint in the intracortical cursor systems studied, but not the broader claim that decoding accuracy is the primary speed limit for point-and-click BCIs generally. (LLM, conf=0.84)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“EEG spellers based on oddball potentials or motor imagery typically achieve 1–5 characters per minute28–32.”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Could not verifycites [28] Vansteensel, M. J. et al. Fully implanted brain–computer interface in a locked-in patient with ALS. N. Engl. J. Med. 375, 2060–2066 (2016). sourceExperimental — accuracy not yet measured
The Results section, “Communication and Independent Home Use” (page unavailable), reports spelling times for one patient using a fully implanted interface. Our copy does not give a typical rate for EEG spellers based on oddball potentials or motor imagery; searches for those methods and their characters-per-minute rates found no such passage. (LLM, conf=0.76)
- Partly supportedcites [29] Nijboer, F. et al. A P300-based brain–computer interface for people with amyotrophic lateral sclerosis. Clin. Neurophysiol. 119, 1909–1916 (2008). sourceExperimental — accuracy not yet measured
The Results report 1.2 and 2.1 selections per minute for the study’s P300 speller, which uses an oddball paradigm (Introduction). This supports the stated range for this speller in people with ALS, but the paper does not establish that EEG spellers generally achieve that rate or provide a rate for motor-imagery spellers. Page numbers were not available in our copy. (LLM, conf=0.86)
- Partly supportedcites [30] Townsend, G. et al. A novel P300-based brain–computer interface stimulus presentation paradigm: moving beyond rows and columns. Clin. Neurophysiol. 121, 1109–1120 (2010). sourceExperimental — accuracy not yet measured
The Discussion (§4.2; page unavailable) reports 4.10 selections/min for a previously studied P300 speller and 4.36 selections/min for this paper’s checkerboard P300 paradigm, both within the claimed range. It does not establish that 1–5 characters/min is typical across EEG spellers or report a rate for motor-imagery spellers. (LLM, conf=0.84)
- Partly supportedcites [31] McCane, L. M. et al. P300-based brain–computer interface (BCI) event-related potentials (ERPs): people with amyotrophic lateral sclerosis (ALS) vs. age-matched controls. Clin. Neurophysiol. 126, 2124–2131 (2015). sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) identifies the P300 speller as an oddball-response system that provides slow communication. The Methods describe a characters-per-minute measure, and the Results report communication rates for this study, but the passages do not state that EEG spellers typically achieve 1–5 characters per minute or address motor-imagery spellers. (LLM, conf=0.77)
- Partly supportedcites [32] Wolpaw, J. R. et al. Independent home use of a brain–computer interface by people with amyotrophic lateral sclerosis. Neurology 91, e258–e267 (2018). sourceExperimental — accuracy not yet measured
The Results report an average of 2.9 (±1.3 SD) selections per minute for BCI applications used by people with ALS, within the claimed numerical range. The paper does not establish a typical rate of 1–5 characters per minute across EEG spellers based on oddball potentials or motor imagery; selections are also not necessarily characters. (LLM, conf=0.86)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“For intracortical BCIs, the highest performing method has been point-and-click typing with a 2D computer cursor, peaking at 40 characters per minute 7”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
What the sources say
- Partly supportedcites [7] Pandarinath, C. et al. High performance communication by people with paralysis using an intracortical brain–computer interface. eLife 6, e18554 (2017). sourceExperimental — accuracy not yet measured
The Introduction describes two-dimensional cursor control with click selection, and the Results say this study outperformed previous BCIs tested with people with motor impairment. The Results report T5’s average typing rate as 39.2 correct characters per minute, approximately 40, but do not identify 40 as a peak or an upper limit. (LLM, conf=0.85)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“The prior state-of-the-art for free typing in intracortical BCIs was 24.4 correct characters per minute 7.”
§ High-performance brain-to-text communication via handwriting > Decoding handwritten sentences
What the sources say
- Supportedcites [7] Pandarinath, C. et al. High performance communication by people with paralysis using an intracortical brain–computer interface. eLife 6, e18554 (2017). source
The Results report T6’s mean free typing rate as 24.4 correct characters per minute and describe those sessions as achieving, to the authors’ knowledge, the highest typing rate to date for a person with a physical disability using a BCI. This supports the citing paper’s description of that rate as the prior benchmark for free typing. (LLM, conf=0.93)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“We used a time-alignment technique to remove temporal variability11, revealing remarkably consistent underlying patterns of neural activity that are unique to each character (Fig. 1c).”
§ High-performance brain-to-text communication via handwriting > Neural representation of handwriting
What the sources say
- Supportedcites [11] Williams, A. H. et al. Discovering precise temporal patterns in large-scale neural recordings through robust and interpretable time warping. Neuron 105, 246–259 (2020). source
The citation is attached to the time-alignment technique, not the citing authors’ character-specific finding. The source’s Results section (“Time Warping Framework”) describes trial-by-trial time warping to correct variability in the timing of neural activity. (LLM, conf=0.91)
boosteddirectSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Current microelectrode array technology has been shown to retain functionality for 1000+ days post implant37,38 (including here; see Extended Data Fig. 7)”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Supportedcites [37] Simeral, J. D., Kim, S.-P., Black, M. J., Donoghue, J. P. & Hochberg, L. R. Neural control of cursor trajectory and click by a human with tetraplegia 1000 days after implant of an intracortical microelectrode array. J. Neural Eng. 8, 025027 (2011). source
The Methods section (2.2) places the study sessions on implant days 999–1003. The Discussion (4.1) reports successful point-and-click control on day 1003, supporting retained array functionality beyond 1000 days. The parenthetical “including here” refers to the citing authors’ own study, not a finding attributed to this source. (LLM, conf=0.94)
- Not checked — no full text of the sourcecites [38] Bullard, A. J., Hutchison, B. C., Lee, J., Chestek, C. A. & Patil, P. G. Estimating risk for future intracranial, fully implanted, modular neuroprosthetic systems: a systematic review of hardware complications in clinical deep brain stimulation and experimental human intracortical arrays. Neuromodulation 23, 411–426 (2020). source
boostedSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“and has enabled the highest BCI performance to date compared to other recording technologies (EEG, ECoG) for restoring communication7, arm control2,5, and general-purpose computer use39.”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Supportedcites [7] Pandarinath, C. et al. High performance communication by people with paralysis using an intracortical brain–computer interface. eLife 6, e18554 (2017). source
Citation 7 is attached to restoring communication, not to arm control or general-purpose computer use, which have separate citations. In the Results (Table 1; page unavailable), the paper compares its intracortical communication BCI with prior systems, including EEG systems, and states that it outperforms all previous BCIs tested with people with motor impairment. (LLM, conf=0.90)
- Supportedcites [2] Collinger, J. L. et al. High-performance neuroprosthetic control by an individual with tetraplegia. Lancet 381, 557–564 (2013). source
This citation is attached to arm control, not to communication or general-purpose computer use. The Methods (Array implantation; page unavailable) describe intracortical microelectrode arrays, and the Introduction (page unavailable) contrasts this study’s robust seven-dimensional prosthetic-arm control with earlier human studies limited to three-dimensional translation and discrete grasp. That supports the citing sentence’s arm-control comparison. (LLM, conf=0.86)
- Partly supportedcites [5] Ajiboye, A. B. et al. Restoration of reaching and grasping movements through brain-controlled muscle stimulation in a person with tetraplegia: a proof-of-concept demonstration. Lancet 389, 1821–1830 (2017). sourceExperimental — accuracy not yet measured
The Methods describe implanted microelectrode arrays controlling reaching and grasping, and the Discussion says intracortical BCIs offer greater functional performance than non-invasive systems, specifically contrasting robotic-arm control with EEG. The paper does not establish a performance comparison with ECoG. Communication and general-purpose computer use have separate citations in the citing sentence and are not judged against this source. (LLM, conf=0.84)
- Partly supportedcites [39] Nuyujukian, P. et al. Cortical control of a tablet computer by people with paralysis. PLoS One 13, e0204566 (2018). sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) says prior EEG and ECoG work had not shown general control of a computing device such as a commercial tablet. The Abstract and Conclusion (pages unavailable) report high-performance control of an unmodified general-purpose tablet using implanted arrays. The paper does not establish that this achieved the highest BCI performance relative to EEG or ECoG; the separate communication and arm-control claims carry other citations in the citing sentence. (LLM, conf=0.84)
boosteddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We found that when only 2–7 days passed between sessions, performance was reasonable with no decoder retraining (11.1% raw error rate, 1.5% with a language model)”
§ High-performance brain-to-text communication via handwriting > Daily decoder retraining
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Importantly, typing rates were high, plateauing at 90 characters per minute with a 5.4% character error rate (Fig. 2c, average of red circles).”
§ High-performance brain-to-text communication via handwriting > Decoding handwritten sentences
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- It may state two separate claims as one. 20
- Probably not a substantial claim. 8
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“which is comparable to state-of-the-art speech recognition systems (e.g. word error rates of 4–5% 14,15), putting it well within the range of usability.”
§ High-performance brain-to-text communication via handwriting > Decoding handwritten sentences
What the sources say
- Disputedcites [14] Xiong, W. et al. The Microsoft 2017 Conversational Speech Recognition System. Preprint at https://arxiv.org/abs/1708.06073 (2017). sourceExperimental — accuracy not yet measured
The Conclusions (p. 4) identify 6.2% as the state-of-the-art word error rate on the Switchboard test data, not 4–5%. The paper does not assess whether an error rate is usable; the citing sentence presents usability as its own inference. (LLM, conf=0.88)
Checked by a reasoning model — may be wrong.
- Not checked — no full text of the sourcecites [15] He, Y. et al. Streaming end-to-end speech recognition for mobile devices. In 2019 IEEE International Conference on Acoustics, Speech and Signal Processing 6381–6385 (2019). source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Disputed
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Prior BCI studies have shown that the motor intention for gross motor skills, such as reaching, grasping or moving a computer cursor, remains neurally encoded in motor cortex after paralysis1–7.”
§ High-performance brain-to-text communication via handwriting > Neural representation of handwriting
What the sources say
- Supportedcites [1] Hochberg, L. R. et al. Reach and grasp by people with tetraplegia using a neurally controlled robotic arm. Nature 485, 372–375 (2012). source
The Abstract reports that people with longstanding tetraplegia previously used neural signals to move and click a computer cursor. It also reports that two participants used motor-cortex activity generated by intended arm and hand movements to reach and grasp with a robotic arm, supporting the claim that these intentions remain neurally encoded after paralysis. (LLM, conf=0.91)
- Supportedcites [2] Collinger, J. L. et al. High-performance neuroprosthetic control by an individual with tetraplegia. Lancet 381, 557–564 (2013). source
The Methods describe intracortical electrodes implanted in the motor cortex of a person with tetraplegia, and the Introduction reports neural control of reaching and grasping. The Discussion also reports earlier human intracortical studies controlling computer cursors. The copy provides section labels but no page numbers. (LLM, conf=0.91)
- Partly supportedcites [3] Aflalo, T. et al. Neurophysiology. Decoding motor imagery from the posterior parietal cortex of a tetraplegic human. Science 348, 906–910 (2015). sourceExperimental — accuracy not yet measured
The passage comparing this study with previous human M1 recordings (indexed section labeled Abstract; page unavailable) reports that those studies decoded movement trajectories, supporting the motor-cortex portion in part. This paper’s own recordings of imagined reaching, grasping, and cursor control were from the posterior parietal cortex of a tetraplegic participant, not motor cortex. It does not establish that the earlier motor-cortex studies demonstrated all three movements named in the claim. (LLM, conf=0.88)
- Not checked — no full text of the sourcecites [4] Bouton, C. E. et al. Restoring cortical control of functional movement in a human with quadriplegia. Nature 533, 247–250 (2016). source
- Supportedcites [5] Ajiboye, A. B. et al. Restoration of reaching and grasping movements through brain-controlled muscle stimulation in a person with tetraplegia: a proof-of-concept demonstration. Lancet 389, 1821–1830 (2017). source
The Introduction reports that people with paralysis had used intracortical BCIs to control cursors and robotic reaching and grasping. The Methods locate the arrays in motor cortex, and the Results report neural activity strongly related to intended movements in the participant with tetraplegia. Together, these passages support the claim that movement intentions remain neurally encoded after paralysis. (LLM, conf=0.90)
- Supportedcites [6] Jarosiewicz, B. et al. Virtual typing by people with tetraplegia using a self-calibrating intracortical brain–computer interface. Sci. Transl. Med. 7, 313ra179 (2015). source
The Introduction reports that people with tetraplegia have controlled cursors and robotic arms through intracortical BCIs by imagining arm movement. The Results locate the recorded neural signals in the hand/arm area of motor cortex, and the References identify an earlier study of neurally controlled reaching and grasping by people with tetraplegia. No page numbers are available in our copy. (LLM, conf=0.88)
- Partly supportedcites [7] Pandarinath, C. et al. High performance communication by people with paralysis using an intracortical brain–computer interface. eLife 6, e18554 (2017). sourceExperimental — accuracy not yet measured
The Discussion (page unavailable) reports that people with no limb movement controlled a cursor using an intracortical BCI, and that monkeys given a paralytic agent retained grasp-related cortical activity. Our copy does not establish the claimed post-paralysis encoding of intended reaching; its mention of reaching is in a reference title about primates. (LLM, conf=0.82)
boostedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“as might be expected from prior work showing short-term stability of neural recordings19–21.”
§ High-performance brain-to-text communication via handwriting > Daily decoder retraining
What the sources say
- Supportedcites [19] Dickey, A. S., Suminski, A., Amit, Y. & Hatsopoulos, N. G. Single-unit stability using chronically implanted multielectrode arrays. J. Neurophysiol. 102, 1331–1339 (2009). source
The Abstract reports that 57% of the original units remained stable through 7 days, and the Discussion concludes that a subset of neurons remained stable for days to weeks. This supports the cited point about short-term stability of neural recordings; the finding concerns a subset of recorded units, not every unit. Page numbers were not provided in the supplied passages. (LLM, conf=0.93)
- Supportedcites [20] Eleryan, A. et al. Tracking single units in chronic, large scale, neural recordings for brain machine interface applications. Front. Neuroeng. 7, 23 (2014). source
The Discussion reports stable single units across recording days and describes this paper’s method for tracking them using 15-minute recordings. That supports the broad claim of short-term stability, though the paper also notes that neural signals can fluctuate and not every unit remains stable. No page number is available in our copy. (LLM, conf=0.84)
- Supportedcites [21] Downey, J. E., Schwed, N., Chase, S. M., Schwartz, A. B. & Collinger, J. L. Intracortical recording stability in human brain–computer interface users. J. Neural Eng. 15, 046016 (2018). source
The Results section, “Most units are stable for hours,” reports that 72%–80% of recorded units remained stable across recordings one hour apart in both subjects. This supports the cited claim of short-term stability without implying that every recording was stable. (LLM, conf=0.94)
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“More generally, using the principle of maximizing the nearest neighbor distance between movements, it should be possible to optimize a set of movements for ease of classification26.”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
What the sources say
- Supportedcites [26] Cunningham, J. P., Yu, B. M., Gilja, V., Ryu, S. I. & Shenoy, K. V. Toward optimal target placement for neural prosthetic devices. J. Neurophysiol. 100, 3445–3457 (2008). source
The Methods section states that the least distinguishable pair of reach targets has the smallest KL divergence and that maximizing this minimum divergence reduces decoding error. This supports optimizing a set of movements for easier classification; the relevant distance is between their neural-response distributions, not necessarily their physical positions. (LLM, conf=0.88)
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Variability in performance across participants is also a potential concern (in a prior study, T5 achieved the highest performance of 3 participants7).”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Supportedcites [7] Pandarinath, C. et al. High performance communication by people with paralysis using an intracortical brain–computer interface. eLife 6, e18554 (2017). source
The Results report performance for all three participants: T5 had the highest average grid-task bitrate (3.7 bps, versus 2.2 for T6 and 1.4 for T7). The Results also show T5 had the highest reported average typing rate, while the Discussion explicitly notes variation in communication rates across participants. (LLM, conf=0.94)
hedgedn = 3- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Recently, speech-decoding BCIs have shown exciting promise for restoring rapid communication (e.g. 34,16,17), but their accuracies and vocabulary sizes require further improvement to support general-purpose use.”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Supportedcites [34] Mugler, E. M. et al. Direct classification of all American English phonemes using signals from functional speech motor cortex. J. Neural Eng. 11, 035015 (2014). source
The Abstract reports word identification equivalent to as much as 33.6 words per minute, supporting the promise of speech-based BCI communication. The Discussion says word identification was only 18.8% and was not sufficient for communication, while describing broader word decoding as a future step. Together, these support the claim that accuracy and vocabulary coverage need improvement for general-purpose use. (LLM, conf=0.91)
- Supportedcites [16] Anumanchipalli, G. K., Chartier, J. & Chang, E. F. Speech synthesis from neural decoding of spoken sentences. Nature 568, 493–498 (2019). source
The Abstract and Discussion describe neural speech synthesis as a promising route to communication at a natural speaking rate. Synthesis performance reports testing with only 25- or 50-word choice pools and median word error rates of 31% and 53%, respectively. Those limits support the claim that accuracy and vocabulary would need improvement for general-purpose use. (LLM, conf=0.85)
- Supportedcites [17] Makin, J. G., Moses, D. A. & Chang, E. F. Machine translation of cortical activity to text with an encoder-decoder framework. Nat. Neurosci. 23, 575–582 (2020). source
The Abstract reports decoding at natural-speech rates with word error rates as low as 3%. The Discussion qualifies that result: testing was limited to 30–50 sentences and a 250-word vocabulary, and it remains unknown how much data would be needed to extend the approach to general English. Page numbers were not provided in our copy. (LLM, conf=0.88)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“EEG spellers that use visually evoked potentials have achieved speeds of 60 characters per minute 18, but have important usability limitations, as they tie up the eyes, are not typically self-paced, and require panels of flashing lights on a screen.”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Supportedcites [18] Chen, X. et al. High-speed spelling with a noninvasive brain–computer interface. Proc. Natl Acad. Sci. USA 112, E6058–E6067 (2015). source
The citation is attached to the 60-characters-per-minute figure. The Results report that this EEG-based SSVEP speller achieved that rate; the subsequent usability characterization is not the part directly attributed to this source. (LLM, conf=0.94)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Intracortical BCIs based on 2D cursor movements give the user more freedom to look around and set their own pace of communication, but have yet to exceed 40 correct characters per minute in people7.”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Partly supportedcites [7] Pandarinath, C. et al. High performance communication by people with paralysis using an intracortical brain–computer interface. eLife 6, e18554 (2017). sourceExperimental — accuracy not yet measured
The Introduction describes continuous two-dimensional cursor control, and the Discussion describes the point-and-click interface as flexible. The Results report average copy-typing rates up to 39.2 correct characters per minute, but do not establish that no individual typing rate exceeded 40. The cited passages also do not establish the specific claim that users can look around while communicating. (LLM, conf=0.84)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Partly supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The neural activity appeared to be strong and repeatable, although the timing of its peaks and valleys varied across trials, potentially due to fluctuations in writing speed.”
§ High-performance brain-to-text communication via handwriting > Neural representation of handwriting
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We assessed the stability of the neural patterns associated with each character and found high short-term stability (mean correlation = 0.85 when 7 days apart or less), and neural changes that seemed to accumulate at a steady and predictable rate (Extended Data Fig. 4).”
§ High-performance brain-to-text communication via handwriting > Daily decoder retraining
hedgedcorrelationalSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Accuracy was extremely high in this regime (0.17% character error rate), indicating a high potential ceiling of performance, although this decoder would not be able to provide letter-by-letter feedback to the user.”
§ High-performance brain-to-text communication via handwriting > Decoding handwritten sentences
hedgedcorrelationalSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Readily recognizable letter shapes confirm that pen tip velocity is robustly encoded.”
§ High-performance brain-to-text communication via handwriting > Neural representation of handwriting
boostedcorrelationalSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We found that spatial dimensionality was only modestly larger for characters (1.24 times larger; 95% CI = [1.19, 1.30]), but that the temporal dimensionality was much greater (2.65 times larger; 95% CI = [2.58, 2.72])”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The character error rate fell to 0.89% and the word error rate fell to 3.4% averaged across all days”
§ High-performance brain-to-text communication via handwriting > Decoding handwritten sentences
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The t-SNE visualization revealed tight clusters of neural activity for each character and a predominantly motoric encoding where characters that are written similarly are closer together.”
§ High-performance brain-to-text communication via handwriting > Neural representation of handwriting
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“T5 has a high-level spinal cord injury (C4 AIS C) and was paralyzed from the neck down; his hand movements were entirely non-functional and limited to twitching and micromotion.”
§ High-performance brain-to-text communication via handwriting > Neural representation of handwriting
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Simply by allowing the trajectories to change in time (Fig. 4g), the nearest neighbor distance between the neural trajectories can be increased, resulting in an increase in classification accuracy when noise levels are large enough (Fig. 4h).”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The results confirm that high performance can also be achieved when the user writes self-generated sentences as opposed to copying on-screen prompts (73.8 characters per minute with an 8.54% character error rate in real-time, 2.25% with a language model).”
§ High-performance brain-to-text communication via handwriting > Decoding handwritten sentences
boostedSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We found that 10 sentences (8.7 minutes) were enough to achieve a raw error rate of 8.5% (1.7% with a language model), although 30 sentences were needed to match the raw online error rate of 5.9%.”
§ High-performance brain-to-text communication via handwriting > Daily decoder retraining
plaindirectSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We think this is one, but not necessarily the only, important factor that enabled a handwriting BCI to go faster than continuous-motion point-and-click BCIs.”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Retraining helps account for changes in neural recordings that accrue over time, which might be caused by neural plasticity or electrode array micromotion.”
§ High-performance brain-to-text communication via handwriting > Daily decoder retraining
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
A claim is checked only against the cited paper’s full text, never its abstract, and none is available for its sources.
In the paper
“Other discrete (classification-based) BCIs have also typically used directional movements with little temporal variety, which may have limited their accuracy and/or the size of the movement set24,25.”
§ High-performance brain-to-text communication via handwriting > Temporal variety improves decoding
What the sources say
- Not checked — no full text of the sourcecites [24] Musallam, S., Corneil, B. D., Greger, B., Scherberger, H. & Andersen, R. A. Cognitive control signals for neural prosthetics. Science 305, 258–262 (2004). source
- Not checked — no full text of the sourcecites [25] Santhanam, G., Ryu, S. I., Yu, B. M., Afshar, A. & Shenoy, K. V. A high-performance brain–computer interface. Nature 442, 195–198 (2006). source
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The demonstrated real-time system is general (the user can express any sentence), easy to use (entirely self-paced and the eyes are free to move), and accurate enough to be useful in the real-world (94.1% raw accuracy, and >99% accuracy offline with a large-vocabulary language model).”
§ High-performance brain-to-text communication via handwriting > Discussion
boostedSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“These methods could be applied more generally to any sequential behavior that cannot be observed directly (e.g., decoding speech from someone who can no longer speak).”
§ High-performance brain-to-text communication via handwriting > Discussion
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
A claim is checked only against the cited paper’s full text, never its abstract, and none is available for its sources.
In the paper
“More broadly, intracortical microelectrode array technology is still maturing, and requires further demonstrations of longevity, safety, and efficacy before widespread clinical adoption35,36.”
§ High-performance brain-to-text communication via handwriting > Discussion
What the sources say
- Not checked — no full text of the sourcecites [35] Nurmikko, A. Challenges for large-scale cortical interfaces. Neuron 108, 259–269 (2020). source
- Not checked — no full text of the sourcecites [36] Vázquez-Guardado, A., Yang, Y., Bandodkar, A. J. & Rogers, J. A. Recent advances in neurotechnologies with broad potential for neuroscience research. Nat. Neurosci. 23, 1522–1536 (2020). source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Finally, it is important to recognize that the current system is a proof-of-concept that a high-performance handwriting BCI is possible (in a single participant); it is not yet a complete, clinically viable system.”
§ High-performance brain-to-text communication via handwriting > Discussion
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The authors state this as an assumption.
It was not yet known whether the neural representation for a rapid, highly dexterous skill such as handwriting remains intact after paralysis.
“However, it is still unknown whether the neural representation for a rapid and highly-dexterous motor skill, such as handwriting, also remains intact.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“which is 3.4 times as fast as the previous record8”
From A high-performance speech neuroprosthesis (2023) · cites this paper as [8]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“During 14 days of data collection shortly after device implantation, we achieved high-performance text decoding, exceeding communication speeds of previous BCIs by a factor of 4 or more1,3,9”
From A high-performance neuroprosthesis for speech decoding and avatar control (2023) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“A second approach, specific to text decoding, involves leveraging large language models to automatically recalibrate MEA-based text neuroprostheses.”
From The speech neuroprosthesis (2024) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“A similar tradeoff can be facilitated by having a spelling framework available on the neuroprosthesis, either using attempted speech120 or handwriting9, which would allow users to spell out arbitrary sentences with high accuracy (albeit with substantially lower speeds).”
From The speech neuroprosthesis (2024) · cites this paper as [9]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“This decoding rate exceeds our participant’s typical communication rate using her assistive device (14.2 WPM; Supplementary Note 2) and is closer to naturalistic speaking rates than has been previously reported with communication neuroprostheses1,3,7–9.”
From A high-performance neuroprosthesis for speech decoding and avatar control (2023) · cites this paper as [9]
The cited source says what the claim attributes to it.
In the citing paper
“In this approach, MEA activity is decoded into subword units and converted into sentences using language models; this language model output is used as a ‘pseudo-label’ to update the RNN decoding model9,16,105.”
From The speech neuroprosthesis (2024) · cites this paper as [9]
The cited source says what the claim attributes to it.
In the citing paper
“MEA-based decoders, including text decoders9,15, have traditionally required day-to-day recalibration owing to signal instabilities and drift, requiring additional dedicated time from the user.”
From The speech neuroprosthesis (2024) · cites this paper as [9]
The cited source says what the claim attributes to it.
In the citing paper
“and several studies indicate that UEAs retain their functionality in humans for more than 1000 days (29, 31).”
From Visual percepts evoked with an intracortical 96-channel microelectrode array inserted in human occipital cortex (2021) · cites this paper as [29]
The cited source says what the claim attributes to it.
In the citing paper
“Although the duration of this study was limited to 6 months, UEAs have been successfully implanted in human subjects in numerous studies (25-29) and these devices have continued to work for many years.”
From Visual percepts evoked with an intracortical 96-channel microelectrode array inserted in human occipital cortex (2021) · cites this paper as [29]
The cited source says what the claim attributes to it.
In the citing paper
“Recently, BCIs based on hand movement activity have enabled typing speeds of between eight and 18 words per minute in people with paralysis8,29.”
From A high-performance speech neuroprosthesis (2023) · cites this paper as [8]
The cited source says what the claim attributes to it.
In the citing paper
“Neurotechnologies can decode and project very specific and often rudimentary forms of thought, such as imagined handwriting, 40 typing, 27 and other kinds of intended and directed movements. 28”
From AI and Neurotechnology (2023) · cites this paper as [40]
Citations
Cites
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 3×
“We used the gradient descent method “Adam” (Kingma and Ba, 2017) (beta1 = 0.”
Method(2008) · Journal of Machine Learning Research · Cited by 33,051
Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation
Cho, K., van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1179 · Cited by 25,545 · cited 1×
“We used a two layer, gated recurrent unit RNN (Cho et al., 2014) to convert T5’s neural activity into a time series of character probabilities (see Extended Data Fig.”
MethodHinton, G., Deng, L., Yu, D., Dahl, G., Mohamed, A.-R., Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T., et al. (2012) · IEEE Signal Processing Magazine · doi:10.1109/msp.2012.2205597 · Cited by 10,260
TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Corrado, G. S., Davis, A., Dean, J., Devin, M., et al. (2016) · arXiv · doi:10.48550/arxiv.1603.04467 · Cited by 9,408 · cited 2×
“We used the “Adam” gradient descent method (Abadi et al., 2016; Kingma and Ba, 2017).”
MethodSpeech recognition with deep recurrent neural networks
Graves, A., Mohamed, A.-R., & Hinton, G. (2013) · 2013 IEEE International Conference on Acoustics, Speech and Signal Processing · doi:10.1109/icassp.2013.6638947 · Cited by 8,893 · cited 1×
“Our methods are similar to neural network methods used in automatic speech recognition (Hinton et al., 2012; Graves et al., 2013; Zeyer et al., 2017; Xiong et al., 2017), but with some key changes made to achieve high performance on neural activity in a highly data-limited regime (1-10 hours of data, as opposed to 1,0…”
Method“Librispeech: an ASR corpus based on public domain audio books,”
Panayotov, V., Chen, G., Povey, D., & Khudanpur, S. (2015) · 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2015.7178964 · Cited by 6,072 · cited 1×
“, 2017), but with some key changes made to achieve high performance on neural activity in a highly data-limited regime (1-10 hours of data, as opposed to 1,000+ hours (Cieri et al., 2004; Panayotov et al., 2015; He et al., 2019)).”
BackgroundGraves, A., Fernández, S., Gomez, F., & Schmidhuber, J. (2006) · Proceedings of the 23rd international conference on Machine learning - ICML '06 · doi:10.1145/1143844.1143891 · Cited by 5,370
Kaldi Speech Recognition Toolkit
Povey, D. (2024) · Infoscience (Ecole Polytechnique Fédérale de Lausanne) · doi:10.57702/jb3fvbn9 · Cited by 5,270
Neuronal ensemble control of prosthetic devices by a human with tetraplegia
Hochberg, L. R., Serruya, M. D., Friehs, G. M., Mukand, J. A., Saleh, M., Caplan, A. H., Branner, A., Chen, D., Penn, R. D., & Donoghue, J. P. (2006) · Nature · doi:10.1038/nature04970 · Cited by 3,295
Self-citation
Cited by
Sorted by citations
Items 101–200 of 967
Rahman, A. U., & Halim, Z. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-13114-5 · Cited by 44
“Till now the restoration of gross motor abilities was the primary focus, however, it is observed that handwriting or touch-typing provides faster communication [35].”
BackgroundDirect speech reconstruction from sensorimotor brain activity with optimized deep learning models
Berezutskaya, J., Freudenburg, Z. V., Vansteensel, M. J., Aarnoutse, E. J., Ramsey, N. F., & van Gerven, M. A. J. (2023) · Journal of Neural Engineering · doi:10.1088/1741-2552/ace8be · Cited by 42
Multi-scale neural decoding and analysis
Lu, H., Lorenc, E. S., Zhu, H., Kilmarx, J., Sulzer, J. S., Xie, C., Tobler, P. N., Watrous, A. J., Orsborn, A. L., Lewis‐Peacock, J. A., et al. (2021) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac160f · Cited by 47
“Most recently, using 192 channels and threshold crossing spike rates as input to a recurrent neural network decoder, Willett and colleagues [67] decoded imagined handwritten characters in humans with over 99% accuracy at a decoding speed comparable to the typing speed of age-matched controls, significantly improving t…”
MethodDeep Adversarial Domain Adaptation With Few-Shot Learning for Motor-Imagery Brain-Computer Interface
Phunruangsakao, C., Achanccaray, D., & Hayashibe, M. (2022) · IEEE Access · doi:10.1109/access.2022.3178100 · Cited by 43
Saway, B. F., Palmer, C., Hughes, C. L., Triano, M. J., Suresh, R. E., Gilmore, J., George, M. S., Kautz, S. A., & Rowland, N. C. (2024) · Neurotherapeutics · doi:10.1016/j.neurot.2024.e00337 · Cited by 53
“Another clever technique was implemented using intracortical electrodes in motor cortex to decode hand-written language, which was able to achieve speeds of 90 characters per minute, a 3-fold improvement over their previous iteration and almost 20 times faster than EEG P300-based decoders [71].”
MethodA tactile oral pad based on carbon nanotubes for multimodal haptic interaction
Hou, B., Yang, D., Ren, X., Yi, L., & Liu, X. (2024) · Nature Electronics · doi:10.1038/s41928-024-01234-9 · Cited by 48
“Unlike finger movements and complex brain–computer interfaces that demand full concentration, tongue movements are more intuitive and less demanding in terms of focus 20–22 .”
Silva, A. B., Liu, J. R., Metzger, S. L., Bhaya-Grossman, I., Dougherty, M. E., Seaton, M. P., Littlejohn, K. T., Tu-Chan, A., Ganguly, K., Moses, D. A., et al. (2024) · Nature Biomedical Engineering · doi:10.1038/s41551-024-01207-5 · Cited by 54
“Similar to previous studies, we chose to report WER over blocks, given that short phrases may become overly influential 42,43,76 .”
A Pathway into Metaverse: Gesture Recognition Enabled by Wearable Resistive Sensors
Duan, S., Zhao, F., Yang, H., Hong, J., Shi, Q., Lei, W., & Wu, J. L. (2023) · Advanced Sensor Research · doi:10.1002/adsr.202200054 · Cited by 42
Natraj, N., Seko, S., Abiri, R., Miao, R., Yan, H., Graham, Y., Tu-Chan, A., Chang, E. F., & Ganguly, K. (2025) · Cell · doi:10.1016/j.cell.2025.02.001 · Cited by 45
A generic noninvasive neuromotor interface for human-computer interactionPreprint
Labs, C.-L. A. R., Sussillo, D., Kaifosh, P., & Reardon, T. R. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.02.23.581779 · Cited by 44
Stabilizing brain-computer interfaces through alignment of latent dynamics
Karpowicz, B. M., Ali, Y. H., Wimalasena, L. N., Sedler, A. R., Keshtkaran, M. R., Bodkin, K., Ma, X., Miller, L. E., & Pandarinath, C. (2022) · doi:10.1101/2022.04.06.487388 · Cited by 71
Decoding ECoG signal into 3D hand translation using deep learning
Śliwowski, M., Martin, M., Souloumiac, A., Blanchart, P., & Aksenova, T. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac5d69 · Cited by 41
“For example, studies analyzing intracortical recordings used recurrent neural networks (RNNs) [48] (classification of 31 letters and continuous decoding), RNN-based autoencoders [49] (continuous decoding) as well as CNN [50], and a combination of RNN and CNN [51] to recognize several MI classes…”
Yao, L., Zhu, B., & Shoaran, M. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac4ed1 · Cited by 41
“In contrast, the widely-used intracortical BMIs that capture spikes and local field potentials via penetrating electrodes allow the highest quality of motor control [4, 5], and have led to breakthrough innovations in the human BMI field [6, 7].”
Exploring Cognition with Brain–Machine Interfaces
Andersen, R. A., Aflalo, T., Bashford, L., Bjånes, D. A., & Kellis, S. (2022) · Annual Review of Psychology · doi:10.1146/annurev-psych-030221-030214 · Cited by 39
“For example, decoding attempted handwriting from a paralyzed individual produced faster rates of communication than did point-and-click typing (Willett et al. 2021).”
BackgroundBrain–Machine Interfaces: Closed-Loop Control in an Adaptive System
Sorrell, E., Rule, M. E., & O’Leary, T. (2021) · Annual Review of Control Robotics and Autonomous Systems · doi:10.1146/annurev-control-061720-012348 · Cited by 41
Biohybrid neural interfaces: improving the biological integration of neural implants
Boulingre, M., Lara, R. P., & Green, R. A. (2023) · Chemical Communications · doi:10.1039/d3cc05006h · Cited by 41
Gated Recurrent Units Viewed Through the Lens of Continuous Time Dynamical Systems
Jordan, I. D., Sokół, P., & Park, I. M. (2021) · Frontiers in Computational Neuroscience · doi:10.3389/fncom.2021.678158 · Cited by 70
Design a Novel BCI for Neurorehabilitation Using Concurrent LFP and EEG Features: A Case Study
Feng, Z., Sun, Y., Qian, L., Qi, Y., Wang, Y., Guan, C., & Sun, Y. (2021) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2021.3115799 · Cited by 37
Harnessing the Power of Artificial Intelligence in Otolaryngology and the Communication Sciences
Wilson, B. S., Tucci, D. L., Moses, D. A., Chang, E. F., Young, N. M., Zeng, F., Lesica, N. A., Bur, A. M., Kavookjian, H. L., Mussatto, C. C., et al. (2022) · Journal of the Association for Research in Otolaryngology · doi:10.1007/s10162-022-00846-2 · Cited by 37
“…can communicate by controlling a spelling interface using visual evoked potentials or imagined hand or arm movements (Wolpaw et al. 2002; Sellers et al. 2014; Vansteensel et al. 2016; Pandarinath et al. 2017; Brumberg et al. 2018; Linse et al. 2018; Oxley et al. 2021; Willett et al. 2021).”
BackgroundBiologicalisation in manufacturing – Current state and future trends
Wegener, K., Damm, O., Harst, S., Ihlenfeldt, S., Monostori, L., Teti, R., Wertheim, R., & Byrne, G. (2023) · CIRP Annals · doi:10.1016/j.cirp.2023.05.005 · Cited by 38
Sani, O. G., Pesaran, B., & Shanechi, M. M. (2024) · Nature Neuroscience · doi:10.1038/s41593-024-01731-2 · Cited by 65
“Second, TNDM and LFADS 16,18 , like other prior works 16,18,20,23,24,26,61 , do not provide flexible nonlinearity or explore hypotheses regarding the origin of nonlinearities because they use fixed nonlinear network structures (use-case 4).”
BackgroundNeurorights in the Constitution: from neurotechnology to ethics and politics
Ruíz, S., Valera, L., Ramos, P., & Sitaram, R. (2024) · Philosophical Transactions of the Royal Society B Biological Sciences · doi:10.1098/rstb.2023.0098 · Cited by 53
“BCIs have used motor intention-related information encoded in the motor cortex to help patients achieve gross motor skills, such as reaching, grasping or moving a computer cursor, for using computer-assisted spellers, wheelchairs and exoskeletons that restore walking [20].”
MethodRecent Progress in Bio‐Integrated Intelligent Sensing System
Liu, M., Zhang, Y., & Tao, T. H. (2022) · Advanced Intelligent Systems · doi:10.1002/aisy.202100280 · Cited by 36
Chronic stability of a neuroprosthesis comprising multiple adjacent Utah arrays in monkeys
Chen, X., Wang, F., Kooijmans, R. N., Klink, P. C., Boehler, C., Asplund, M., & Roelfsema, P. R. (2023) · Journal of Neural Engineering · doi:10.1088/1741-2552/ace07e · Cited by 45
Signal Processing for Brain–Computer Interfaces: A review and current perspectives
Wu, L., Liu, A., Ward, R. K., Wang, Z. J., & Chen, X. (2023) · IEEE Signal Processing Magazine · doi:10.1109/msp.2023.3278074 · Cited by 36
“The efficacy of spike sorting techniques in invasive BCI studies is well-documented, with significant advances in recognizing user intent, controlling digital and mechanical devices, and improving communication efficiency for paralyzed users [18].”
BackgroundA Survey on Brain-Computer Interface-Inspired Communications: Opportunities and Challenges
Hu, H., Wang, Z., Zhao, X., Li, R., Li, A., Si, Y., Wang, J., Zhou, T., & Xu, T. (2024) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2024.3396847 · Cited by 41
Pan, H., Ding, P., Wang, F., Li, T., Zhao, L., Nan, W., Fu, Y., & Gong, A. (2024) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2024.1429130 · Cited by 45
“In the case where the accuracy is greater than or equal to 50%, the ITR value has practical significance and can be evaluated after each session (Zickler et al., 2013).”
BackgroundInfluentialUtilising the triboelectricity of the human body for human-computer interactions
Zhang, R., Hummelgård, M., Örtegren, J., Olsen, M., Andersson, H., Yang, Y. F., Olin, H., & Wang, Z. L. (2022) · Nano Energy · doi:10.1016/j.nanoen.2022.107503 · Cited by 34
BRAND: a platform for closed-loop experiments with deep network models
Ali, Y. H., Bodkin, K., Rigotti-Thompson, M., Patel, K., Card, N. S., Bhaduri, B., Nason-Tomaszewski, S. R., Mifsud, D. M., Hou, X., Nicolas, C., et al. (2024) · Journal of Neural Engineering · doi:10.1088/1741-2552/ad3b3a · Cited by 32
Self-citationAn, H., Nason, S. R., Lim, J., Kwon, K., Willsey, M. S., Patil, P. G., Kim, H.-S., Sylvester, D., Chestek, C. A., & Blaauw, D. T. (2022) · IEEE Transactions on Biomedical Circuits and Systems · doi:10.1109/tbcas.2022.3175926 · Cited by 33
Neural Plasticity in Sensorimotor Brain–Machine Interfaces
Dadarlat, M. C., Canfield, R. A., & Orsborn, A. L. (2023) · Annual Review of Biomedical Engineering · doi:10.1146/annurev-bioeng-110220-110833 · Cited by 35
“Nonlinear algorithms such as recurrent neural networks have also been used for both continuous (87, 88) and discrete (2) control.”
MethodDecoding and geometry of ten finger movements in human posterior parietal cortex and motor cortex
Guan, C., Aflalo, T., KADLEC, K. H., Gámez, J., Rosario, E. R., Bari, A., Pouratian, N., & Andersen, R. A. (2023) · Journal of Neural Engineering · doi:10.1088/1741-2552/acd3b1 · Cited by 38
“non-traditional variables, such as handwriting [32] or goals [22].”
BackgroundBio-Inspired Design of Superconducting Spiking Neuron and Synapse
Schegolev, A. E., Klenov, N. V., Gubochkin, G. I., Kupriyanov, M. Y., & Soloviev, I. I. (2023) · Nanomaterials · doi:10.3390/nano13142101 · Cited by 33
“and for real-time systems needed to create and improve bi-directional brain–computer interfaces [33,41].”
BackgroundBrain-computer interface paradigms and neural coding
Tai, P., Ding, P., Wang, F., Gong, A., Li, T., Zhao, L., Su, L., & Fu, Y. (2024) · Frontiers in Neuroscience · doi:10.3389/fnins.2023.1345961 · Cited by 35
“In recent years, many important advances have been made in the innovation of BCI paradigms (Willett et al., 2021, 2023; Metzger et al., 2023).”
BackgroundMethodInfluentialNeural electrodes for brain‐computer interface system: From rigid to soft
Yang, D., Tian, G., Chen, J., Liu, Y., Fatima, E., Qiu, J., Malek, N. A. N. N., & Qi, D. (2025) · BMEMat · doi:10.1002/bmm2.12130 · Cited by 37
An Ultraflexible Electrode Array for Large‐Scale Chronic Recording in the Nonhuman Primate Brain
Tian, Y., Yin, J., Wang, C., He, Z., Xie, J., Feng, X., Zhou, Y., Ma, T., Xie, Y., Li, X., et al. (2023) · Advanced Science · doi:10.1002/advs.202302333 · Cited by 32
Reward signals in the motor cortex: from biology to neurotechnology
Derosière, G., Shokur, S., & Vassiliadis, P. (2025) · Nature Communications · doi:10.1038/s41467-024-55016-0 · Cited by 30
“BCIs are techniques used to decode motor intentions based on neural signals to restore or enhance motor functions, such as hand and arm movements 74 , locomotion 75 , or speech 76 .”
Overcoming failure: improving acceptance and success of implanted neural interfaces
Dalrymple, A. N., Jones, S. T., Fallon, J. B., Shepherd, R. K., & Weber, D. J. (2025) · Bioelectronic Medicine · doi:10.1186/s42234-025-00168-7 · Cited by 38
“…et al. 2017; Barry et al. 2023; Collinger et al. 2013; Deo et al. 2024; Drew 2024; Flesher et al. 2021; Hosman et al. 2023; Rastogi et al. 2021; Willett et al. 2023, 2021); Endovascular electrode array: (Mitchell et al. 2023; Oxley et al. 2020); Epidural spinal cord stimulation (Angeli et al.…”
Human-centric assembly in smart factories
Wang, L., Gao, R. X., Krüger, J., & Váncza, J. (2025) · CIRP Annals · doi:10.1016/j.cirp.2025.04.058 · Cited by 36
The 2021 yearbook of Neurorestoratology
Sharma, H. S., Chopp, M., Chen, L., Sarnowska, A., Xue, M., Ao, Q., Siniscalco, D., Chen, L., Hawamdeh, Z. M., & Huang, H. (2022) · Journal of Neurorestoratology · doi:10.1016/j.jnrt.2022.100008 · Cited by 31
Wong, J. K., Mayberg, H. S., Wang, D. D., Richardson, R. M., Halpern, C. H., Krinke, L., Arlotti, M., Rossi, L., Priori, A., Marceglia, S. R. F., et al. (2023) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2022.1084782 · Cited by 31
“, 2018), and high-performance brain-to-text communication using decoded handwriting (Willett et al., 2021).”
BackgroundSelf-citationCanny, E., Vansteensel, M. J., van der Salm, S. M. A., Müller-Putz, G., & Berezutskaya, J. (2023) · Journal of NeuroEngineering and Rehabilitation · doi:10.1186/s12984-023-01272-y · Cited by 38
“…[228, 273], hand movement towards a specific location [53, 301], hand gestures [37, 47, 165, 218], basic mouth movements [255], facial expressions [256] and continuous motor information, including 2D and 3D finger trajectories, speed, acceleration and force of movement [44, 53, 70, 267, 303].”
BackgroundRestoring Speech Using Brain–Computer Interfaces
Stavisky, S. D. (2025) · Annual Review of Biomedical Engineering · doi:10.1146/annurev-bioeng-110122-012818 · Cited by 42
Brain–computer interfaces in 2023–2024
Chen, S., Chen, M., Wang, X., Liu, X., Liu, B., & Ming, D. (2025) · Brain‐X · doi:10.1002/brx2.70024 · Cited by 33
Cortical Control of Virtual Self-Motion Using Task-Specific Subspaces
Schroeder, K. E., Perkins, S. M., Wang, Q., & Churchland, M. M. (2021) · Journal of Neuroscience · doi:10.1523/jneurosci.2687-20.2021 · Cited by 31
Deep brain–machine interfaces: sensing and modulating the human deep brain
Sui, Y., Yu, H., Zhang, C., Chen, Y., Jiang, C., & Li, L. (2022) · National Science Review · doi:10.1093/nsr/nwac212 · Cited by 33
“While interfacing with the human cerebral cortex allows us to decode sensory and motor signals such as visual responses, hand movements and speech in labs [8,9], there is still a long way to go before these BMIs will be deployed in daily life or on a large scale.”
BackgroundSynthesizing Speech from Intracranial Depth Electrodes using an Encoder-Decoder Framework
Köhler, J., Ottenhoff, M. C., Goulis, S., Angrick, M., Colon, A., Wagner, L., Tousseyn, S., Kubben, P., & Herff, C. (2022) · Neurons Behavior Data analysis and Theory · doi:10.51628/001c.57524 · Cited by 37
“…enable patients who have lost the ability to communicate due to injury, stroke or neuromuscular disorders, to interact with others using a robotic arm [1], high-performance cursor control [2] or even imagined handwriting [3], achieving information transfer rates up to approximately 12 bits/second.”
BackgroundShi, N., Miao, Y., Huang, C., Li, X., Song, Y., Chen, X., Wang, Y., & Gao, X. (2024) · NeuroImage · doi:10.1016/j.neuroimage.2024.120548 · Cited by 33
“Brain-computer interfaces (BCIs) strive to establish high-information-rate communication between the human brain and the external world (Anumanchipalli et al., 2019; Willett et al., 2021).”
Putting representations to use
Cao, R. (2022) · Synthese · doi:10.1007/s11229-022-03522-3 · Cited by 35
Decoding the business of brain–computer interfaces
Drew, L. (2023) · Nature Electronics · doi:10.1038/s41928-023-00929-9 · Cited by 29
The Emergence of Functional Ultrasound for Noninvasive Brain–Computer Interface
Zheng, H., Niu, L., Qiu, W., Liang, D., Long, X., Li, G., Liu, Z., & Meng, L. (2023) · Research · doi:10.34133/research.0200 · Cited by 27
“[3] proposed an intracortical BCI to decode attempted handwriting movements from neural activity in the motor cortex and translates it to text in real time.”
Eisenkolb, V. M., Held, L. M., Utzschmid, A., Lin, X., Krieg, S. M., Meyer, B., Gempt, J., & Jacob, S. N. (2023) · Cell Reports · doi:10.1016/j.celrep.2023.112467 · Cited by 30
Scaled Elastic Hydrogel Interfaces for Brain Electrophysiology
Hu, M., Ren, J., Pan, Y., Cheng, L., Xu, X., Tan, C. L., Sun, H., Shi, Y., & Yan, S. (2024) · Advanced Functional Materials · doi:10.1002/adfm.202407926 · Cited by 30
iLQR-VAE : control-based learning of input-driven dynamics with applications to neural dataPreprint
Schimel, M., Kao, T.-C., Jensen, K. T., & Hennequin, G. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.10.07.463540 · Cited by 29
Feng, X., Feng, X., Qin, B., & Liu, T. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3314642 · Cited by 36
“Additionally, all features are Z-scored as done by Willett et al. [35].”
Human motor cortex encodes complex handwriting through a sequence of stable neural states
Qi, Y., Zhu, X., Xiong, X., Yang, X., Ding, N., Wu, H., Xu, K., Zhu, J., Zhang, J., & Wang, Y. (2025) · Nature Human Behaviour · doi:10.1038/s41562-025-02157-x · Cited by 27
“Humans are superb at controlling sophisticated fine movements, such as writing, typing and musical performance 1,2 .”
Li, G., Jiang, S., Meng, J., Chai, G., Wu, Z., Fan, Z., Hu, J. Z., Sheng, X., Zhang, D., Chen, L., et al. (2022) · NeuroImage · doi:10.1016/j.neuroimage.2022.118969 · Cited by 22
“, 2018; Zheng and Lu, 2015), letters output (Willett et al., 2021; Chen et al., 2015), speech reconstruction (Hamilton et al.”
BackgroundAbbaspourazad, H., Erturk, E., Pesaran, B., & Shanechi, M. M. (2023) · Nature Biomedical Engineering · doi:10.1038/s41551-023-01106-1 · Cited by 40
“Similar to these previous generative networks, previous predictive networks with RNN-based methods 57–59 also do not allow for flexible inference.”
Yang, J. G., Awais, M., Hossain, M. A., Por, L. Y., Haowei, M., Mehedi, I. M., & Iskanderani, A. I. M. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105120 · Cited by 28
Perna, A., Angotzi, G. N., Berdondini, L., & Ribeiro, J. F. (2023) · Frontiers in Neuroscience · doi:10.3389/fnins.2023.1275908 · Cited by 30
“…the BrainGate team has presented BCIs that allow to restore communication in patients who have lost their ability to speak, either by decoding neural signals associated with handwriting (Willett et al., 2021) or by interpreting their intention to utter phonemes and words (Willett et al., 2023).”
BackgroundEndovascular Brain-Computer Interfaces in Poststroke Paralysis
Brannigan, J., Fry, A., Opie, N. L., Campbell, B. C. V., Mitchell, P. J., & Oxley, T. J. (2023) · Stroke · doi:10.1161/strokeaha.123.037719 · Cited by 32
Haghi, B., Aflalo, T., Kellis, S., Guan, C., Gámez, J., Huang, A. Y., Pouratian, N., Andersen, R. A., & Emami, A. (2024) · Nature Biomedical Engineering · doi:10.1038/s41551-024-01297-1 · Cited by 36
“…of determination ( R 2 ) compared with TCs, as the current standard feature extraction techniques in lab works with human participants 4,6,22,26,27,31 , and WTs, which have also demonstrated performance improvements in our offline analysis and in the recent studies on BCIs 20,29 , across multiple…”
Resistive memory-based zero-shot liquid state machine for multimodal event data learning
Lin, N., Wang, S., Li, Y., Wang, B., Shi, S., He, Y., Zhang, W., Yu, Y., Zhang, Y., Zhang, X., et al. (2025) · Nature Computational Science · doi:10.1038/s43588-024-00751-z · Cited by 32
“Human intracortical neural recordings are captured by a microelectrode array and represented as events 42–44 when alphabets from Extended MNIST (E-MNIST) dataset 45 are observed.”
Feasibility of decoding visual information from EEG
Wilson, H., Chen, X., Golbabaee, M., Proulx, M. J., & O'Neill, E. J. (2023) · Brain-Computer Interfaces · doi:10.1080/2326263x.2023.2287719 · Cited by 30
Xiao, Q., Fan, L. H., Ma, Q., Ning, Y., Gu, Z., Chen, L., Li, L., You, J. W., Niu, Y., & Cui, T. J. (2025) · Nature Communications · doi:10.1038/s41467-025-63326-0 · Cited by 29
“5e, f. Subsequently, thesedata streams are decrypted accurately by the method outlined in Supplementary Note 12.”
InfluentialAlgorithm and hardware considerations for real-time neural signal on-implant processing
Zhang, Z., Savolainen, O. W., & Constandinou, T. G. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac5268 · Cited by 25
Recent advances in wireless epicortical and intracortical neuronal recording systems
Ji, B., Liang, Z., Yuan, X., Xu, H., Wang, M., Yin, E., Guo, Z., Wang, L., Zhou, Y., Feng, H., et al. (2022) · Science China Information Sciences · doi:10.1007/s11432-021-3373-1 · Cited by 29
Garwood, I. C., Major, A. J., Antonini, M., Correa, J., Lee, Y., Sahasrabudhe, A., Mahnke, M. K., Miller, E. K., Brown, E. N., & Anikeeva, P. (2023) · Science Advances · doi:10.1126/sciadv.adh0974 · Cited by 25
Large-scale foundation models and generative AI for BigData neuroscience
Wang, R., & Chen, Z. S. (2024) · Neuroscience Research · doi:10.1016/j.neures.2024.06.003 · Cited by 29
Ienca, M., Valle, G., & Raspopović, S. (2025) · The Lancet Digital Health · doi:10.1016/s2589-7500(24)00222-x · Cited by 27
Implantable neural electrodes: from preparation optimization to application
Zhang, B., Xie, R., Jiang, J., Hao, S., Fang, B., Zhang, J., Bai, H., Peng, B., Li, L., Liu, Z., et al. (2023) · Journal of Materials Chemistry C · doi:10.1039/d2tc05162a · Cited by 27
Hernandez-Reynoso, A. G., Sturgill, B. S., Hoeferlin, G. F., Druschel, L. N., Krebs, O. K., Menendez, D. M., Thai, T. T. D., Smith, T. J., Duncan, J., Zhang, J., et al. (2023) · Biomaterials · doi:10.1016/j.biomaterials.2023.122351 · Cited by 25
“Because of this, MEAs have been widely used to develop brain-machine interfaces, restoring lost function in paralyzed and injured individuals [1–7].”
MethodLin, X., Zhang, X., Chen, J., & Jia Liu, B. (2025) · Advanced Materials · doi:10.1002/adma.202413938 · Cited by 25
Neurocognitive and motor-control challenges for the realization of bionic augmentation
Makin, T. R., Micera, S., & Miller, L. E. (2022) · Nature Biomedical Engineering · doi:10.1038/s41551-022-00930-1 · Cited by 23
Semi-Supervised EEG Clustering With Multiple Constraints
Dai, C., Wu, J. X., Monaghan, J. J. M., et al. (2022) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2022.3206330 · Cited by 23
“E LECTROENCEPHALOGRAM (EEG) is an established technique that has been widely applied in Brain-Computer Interfa-ces (BCIs) [1], [2], [3] as a non-invasive means of analysing brain working status and body functions [4].”
MethodBibliometric analysis on Brain-computer interfaces in a 30-year period
Yin, Z., Wan, Y., Fang, H., Li, L., Wang, T., Wang, Z., & Tan, D. (2022) · Applied Intelligence · doi:10.1007/s10489-022-04226-4 · Cited by 23
“[3] Much research on BCI applications has focused on rehabilitating the disabled, comprising brain-controlled wheelchairs [4–6], neuroprostheses [7– 9], and communications equipment [10, 11].”
BackgroundSynthesizing Speech by Decoding Intracortical Neural Activity from Dorsal Motor Cortex
Wairagkar, M., Hochberg, L. R., Brandman, D. M., & Stavisky, S. D. (2023) · International IEEE/EMBS Conference on Neural Engineering · doi:10.1109/ner52421.2023.10123880 · Cited by 25
“Current intracortical BCIs have shown great potential for high accuracy communication by decoding neural activity related to attempted handwriting [9] or hand movements in point and click typing [10] in people with tetraplegia.”
BackgroundSelf-citationNoise Efficient Integrated Amplifier Designs for Biomedical Applications
Simmich, S., Bahr, A., & Rieger, R. (2021) · Electronics · doi:10.3390/electronics10131522 · Cited by 23
“Brain–computer interfaces are used for the control of an exoskeleton [7], real-time brain-to-text translation of handwritten messages [8], or wireless on-screen selection tasks [9].”
Han, J., Gu, X., Yang, G., & Lo, B. P. L. (2023) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2023.3337072 · Cited by 22
Generative language reconstruction from brain recordings
Ye, Z., Ai, Q., Liu, Y., de Rijke, M., Zhang, M., Lioma, C., & Ruotsalo, T. (2025) · Communications Biology · doi:10.1038/s42003-025-07731-7 · Cited by 24
“Forexample, BCIsbasedonmotorrepresentations 49 – 51 or attempted language production 52 have demonstrated a usable performance, but they require extensive user training and active engagement in the input system, demanding signi fi cant user effort 50,52 .”
Intracranial brain-computer interface spelling using localized visual motion response
Liu, D., Xu, X., Li, D., Li, J., Yu, X., Ling, Z., & Hong, B. (2022) · NeuroImage · doi:10.1016/j.neuroimage.2022.119363 · Cited by 23
Stability of motor representations after paralysis
Guan, C., Aflalo, T., Zhang, C. Y., Amoruso, E., Rosario, E. R., Pouratian, N., & Andersen, R. A. (2022) · eLife · doi:10.7554/elife.74478 · Cited by 23
Petrosyan, A., Voskoboinikov, A., Sukhinin, D., Makarova, A. V., Skalnaya, A. A., Arkhipova, N., Sinkin, M. V., & Ossadtchi, A. E. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/aca1e1 · Cited by 26
“While several technologies have been proposed to restore the communication 39 function they primarily rely on brain controlled typing or imaginary handwriting [55] and appear to be practical only 40 for severely affected patients.”
BackgroundInterpol questioned documents review 2019–2022
Deviterne-Lapeyre, M., & Ibrahim, S. (2023) · Forensic Science International Synergy · doi:10.1016/j.fsisyn.2022.100300 · Cited by 22
Robust cortical encoding of 3D tongue shape during feeding in macaques
Laurence-Chasen, J. D., Ross, C. F., Arce-McShane, F. I., & Hatsopoulos, N. G. (2023) · Nature Communications · doi:10.1038/s41467-023-38586-3 · Cited by 22
“The fi nding that 3D tongue posture can be accurately decoded from the sensorimotor cortex opens up a new avenue for potential brain computer interface-based prostheses for restoring orolingual function and communication 49 .”
BackgroundDifferential expression of genes involved in the chronic response to intracortical microelectrodes
Song, S., Druschel, L. N., Chan, E. R., & Capadona, J. R. (2023) · Acta Biomaterialia · doi:10.1016/j.actbio.2023.07.038 · Cited by 22
Ethical considerations for integrating multimodal computer perception and neurotechnology
Hurley, M. E., Sonig, A., Herrington, J. D., Storch, E. A., Lázaro‐Muñoz, G., Blumenthal‐Barby, J., & Kostick, K. M. (2024) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2024.1332451 · Cited by 28
“Recent studies suggest that it is possible to infer visual content of mental processing (Wen et al., 2018), imagined handwriting (Willett et al., 2021), or covert (internal) speech (Pawar and Dhage, 2020) from neural data (Reardon, 2023; Tang et al., 2023).”
BackgroundShaeri, M., Shin, U., Yadav, A., et al. (2024) · Digest of technical papers/Digest of technical papers - IEEE International Solid-State Circuits Conference · doi:10.1109/isscc49657.2024.10454533 · Cited by 22
Neural Data Transformer 2: Multi-context Pretraining for Neural Spiking ActivityPreprint
Ye, J., Collinger, J., Wehbe, L., et al. (2023) · bioRxiv · doi:10.1101/2023.09.18.558113 · Cited by 100
“It is currently common to record from 100-200 electrodes with an interelectrode spacing of 400 µm (Blackrock Utah Arrays) [29, 42], so our budget forces 16-32 channels per token.”
BackgroundAngrick, M., Luo, S., Rabbani, Q., Candrea, D. N., Shah, S., Milsap, G. W., Anderson, W. S., Gordon, C. R., Rosenblatt, K. R., Clawson, L., et al. (2023) · doi:10.1101/2023.06.30.23291352 · Cited by 95
Berger, S., & Rossi, F. (2023) · Communications of the ACM · doi:10.1145/3529088 · Cited by 21
Merk, T., Köhler, R. M., Brotons, T. M., Vossberg, S. R., Peterson, V., Lyra, L. F., Vanhoecke, J., Chikermane, M., Binns, T. S., Li, N., et al. (2025) · Nature Biomedical Engineering · doi:10.1038/s41551-025-01467-9 · Cited by 48
Decoding motor imagery tasks using ESI and hybrid feature CNN
Fang, T., Song, Z., Zhan, G., Zhang, X., Mu, W., Wang, P., Zhang, L., & Kang, X. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac4ed0 · Cited by 20
“For example, patients with high paraplegia can use minds to type [1] or control mechanical prosthetic hands [2].”
Cost function for low-dimensional manifold topology assessment
Zdybał, K., Armstrong, E., Sutherland, J. C., & Parente, A. (2022) · Scientific Reports · doi:10.1038/s41598-022-18655-1 · Cited by 20
“…purpose of data visualization, t-distributed stochastic neighbor embedding (t-SNE) 55 and uniform manifold approximation and projection (UMAP) 60 are gaining popularity in various research disciplines 7,36–38 , Prior to applying dimensionality reduction, training datasets are often preprocessed.”
Biosignal-based co-adaptive user-machine interfaces for motor control
Madduri, M. M., Burden, S. A., & Orsborn, A. L. (2023) · Current Opinion in Biomedical Engineering · doi:10.1016/j.cobme.2023.100462 · Cited by 22
“Iteratively adding control dimensions has been used to achieve high-dimensional BCI control [3,25]. Interestingly, Silversmith and colleagues (2020) showed that, once a stable neural encoding of a two-dimensional BCI task had formed, the participant could learn to control an additional discrete behavior (a ‘mouse clic…”
MethodLiang, Z., Zheng, Z., Chen, W., Pei, Z., Wang, J., & Chen, J. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105556 · Cited by 21
Transforming modeling in neurorehabilitation: clinical insights for personalized rehabilitation
Lin, D. J., Backus, D. A., Chakraborty, S., Liew, S., Valero‐Cuevas, F. J., Patten, C., & Cotton, R. J. (2024) · Journal of NeuroEngineering and Rehabilitation · doi:10.1186/s12984-024-01309-w · Cited by 20
“…the model has immediate or more distant clinical application, there is consequential value to explicitly stating the eventual clinical application (e.g., improving brain-computer interface decoding for assistive use in people with paralysis [14–16] or optimizing rehabilitation protocols [13]).”
BackgroundCandrea, D. N., Shah, S., Luo, S., Angrick, M., Rabbani, Q., Coogan, C. G., Milsap, G., Nathan, K., Wester, B. A., Anderson, W. S., et al. (2024) · Communications Medicine · doi:10.1038/s43856-024-00635-3 · Cited by 25
Acoustic Inspired Brain-to-Sentence Decoder for Logosyllabic Language
Feng, C., Cao, L., Wu, D., Zhang, E., Wang, T., Jiang, X., Zhou, C., Chen, J., Wu, H., Lin, S., et al. (2025) · Cyborg and Bionic Systems · doi:10.34133/cbsystems.0257 · Cited by 23
“…from the brain activity, particularly within the scope of alphabetic linguistic systems (e.g., English and Dutch), by establishing a mapping between language output and different activity-related brain signals, such as hand-writing [ 1 ], spelling [ 2 – 4 ], typing [ 5 , 6 ], and speech [ 7 – 14 ].”
Firing-rate-modulated spike detection and neural decoding co-design
Zhang, Z., & Constandinou, T. G. (2023) · Journal of Neural Engineering · doi:10.1088/1741-2552/accece · Cited by 20
“With the development of deep learning, recurrent neural networks , long-short term memory (LSTM), and the variations become more and more popular [6, 50, 51, 53, 54].”
Influential
Metadata
- Authors
- Willett, F. R., Avansino, D. T., Hochberg, L. R., Henderson, J. M., & Shenoy, K. V.
- Affiliations
- Howard Hughes Medical Institute (US)
- Harvard University (US)
- Stanford Medicine (US)
- Corresponding author
- Willett, Francis R.
- Venue
- Nature
- Published
- 2021-05-12
- Volume
- 593
- Issue
- 7858
- Pages
- 249-254
- ISSN
- 0028-0836 (ISSN-L)
- 1476-4687
- PMC
- PMC8163299
- OpenAlex
- W3161343553
- Semantic Scholar
- bebaab40b42e63cd7cef7b334e5f6bacdcccfc79
- Cited by
- 1,024
- References
- 50
- Citations by year
- 2026: 187 · 2025: 225 · 2024: 203 · 2023: 203 · 2022: 134 · 2021: 67
- FWCI
- 70.16
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Licence
- public-domain
- www.springer.com/tdm (tdm)
- www.springer.com/tdm (vor)
- Locations
- Nature (publishedVersion)
- bioRxiv (Cold Spring Harbor Laboratory) (submittedVersion, public-domain)
- PubMed (publishedVersion)
- Europe PMC (PubMed Central) (acceptedVersion)
- PubMed Central (acceptedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- Funders
- Howard Hughes Medical Institute
- U.S. Department of Veterans Affairs (I01 RX002295, I50RX002864, B6453-R, N-2864-C)
- Massachusetts General Hospital
- Office of Research and Development (N2864C, B6453R)
- National Institutes of Health (5uh2ns095548-02, 5u01dc017844-06)
- Wu Tsai Neurosciences Institute, Stanford University
- National Institute on Deafness and Other Communication Disorders (R01DC009899, DC017844, R01DC014034, U01- DC017844)
- National Institute of Neurological Disorders and Stroke (UH2-NS095548, U01NS098968, NS095548)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
- Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neuroscience and Neural Engineering
- Physical Sciences › Computer Science › Human-Computer Interaction › Gaze Tracking and Assistive Technology
- Keywords
- brain-computer interface, handwriting, text entry, motor cortex, intracortical recording, spinal cord injury, paralysis, real-time communication, typing speed
- MeSH terms
- Brain, Communication, Handwriting, Humans, Spinal Cord Injuries, Time Factors, Neural Networks, Computer, Brain-Computer Interfaces
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-09-28, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-09-28 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
