Flexible Electronics and Devices as Human–Machine Interfaces for Medical Robotics

In the paper

“The advantage of this method is its small size, fast response time, and good linearity.[106,217]”

What the sources say

• Disputed

  In the section on SiNR sensors (page unavailable), the paper describes the diode temperature sensors’ nonlinear characteristics as an advantage, contrary to the claimed advantage of good linearity. Its statement that sensors have a linear and fast response concerns SiNR strain gauges, not the p–n-junction temperature sensors; the cited passages also do not establish small size as an advantage of the junction method. (LLM, conf=0.79)

  Checked by a reasoning model — may be wrong.

  cites [106] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [217] source

In the paper

“Moreover, some diseases of the locomotor system provoke abnormal elevations of metabolites and biomarkers, such as L-dopa in Parkinson’s disease[246–248] uric acid in gout,[45] which are very important evaluation factors.”

What the sources say

• ?Could not verify

  The Introduction (page unavailable) describes levodopa as a treatment for Parkinson’s disease, not as a metabolite or biomarker abnormally elevated by the disease. Searches for elevated L-dopa, biomarker concentrations, and related terms found no passage addressing that claim. The separate gout example is cited to another source, not this paper. (LLM, conf=0.76)

  cites [246] sourceExperimental — accuracy not yet measured
• Disputed

  The Levodopa Monitoring > Abstract passages describe L-dopa as a medication for Parkinson’s disease, not a biomarker elevated by the disease. The on-body study reports increased sweat L-dopa after fava-bean consumption and almost zero concentration without it. The gout example has a separate citation and is not assessed against this source. (LLM, conf=0.85)

  Checked by a reasoning model — may be wrong.

  cites [247] sourceExperimental — accuracy not yet measured
• Partly supported

  The paper’s Abstract supports monitoring L-Dopa levels to guide Parkinson’s disease treatment. It does not support the claim that Parkinson’s disease itself causes L-Dopa to rise: the reported increase followed an oral L-Dopa tablet, and the measurements were made in healthy volunteers. Uric acid in gout is attached to a different citation in the citing sentence. (LLM, conf=0.88)

  cites [248] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [45] source

In the paper

“Due to their breathability, these aforementioned 2D porous materials facilitate the outflow and evaporation of sweat and are not only comfortable to wear but also do not cause much skin irritation.[75]”

What the sources say

• Partly supported

  The Results section (page unavailable) reports that the paper’s perforated e-skin lets sweat and vapor pass through and caused no skin irritation in a one-week forearm test. The Introduction describes its low mass as suitable for imperceptible wear. These findings support the claim for this specific e-skin, but not for 2D porous materials generally. (LLM, conf=0.84)

  cites [75] sourceExperimental — accuracy not yet measured

In the paper

“Together, e-skin and soft robotics can therefore alleviate problems such as limited functionality,[37] heavy weight,[19] and poor aesthetic design.[31]”

What the sources say

• ?Not checked — no full text of the sourcecites [37] source
• ?Not checked — no full text of the sourcecites [19] source
• Partly supported

  Citation [31] is attached to the aesthetic-design part of the sentence. The Prosthetic Hands and Fingers section describes robotic or bionic prostheses with life-like, aesthetically pleasing outer shells, supporting the possibility of improved appearance. It does not establish that e-skin and soft robotics together produce that improvement. (LLM, conf=0.83)

  cites [31] sourceExperimental — accuracy not yet measured

In the paper

“Coating hydrogels on electronics[55] or mixing conductive materials such as carbon nanotubes (CNTs),[56] ionic liquids,[57] or conductive polymers[58] into the gel can help maintain biocompatibility while also preserving the electronic properties.”

What the sources say

• Supported

  The citation [55] supports the hydrogel-on-electronics approach, not the separately cited claims about mixing conductive materials into gels. In Results and Discussion, under “Hydrogel electronics,” the paper describes bonding a biocompatible PVA hydrogel to an electronic sensor-and-heater patch, placing the gel at the tissue interface while demonstrating autonomous electronic operation. (LLM, conf=0.90)

  cites [55] source
• Supported

  For the CNT portion marked [56], the Experimental Methods describe adding CNTs to GelMA hydrogel, and the Results and Discussion report lower impedance than pristine GelMA at physiologically relevant frequencies. The cardiomyocyte viability section reports higher cell viability on CNT-GelMA and no observed CNT cytotoxicity during seven days of culture. This paper does not need to support the separately cited examples of ionic liquids or conductive polymers. (LLM, conf=0.92)

  cites [56] source
• ?Not checked — no full text of the sourcecites [57] source

In the paper

“As a material whose features resemble water-rich tissue in the human body, hydrogels can greatly reduce the inflammatory response of foreign objects, regulate cell or protein attachment, and prevent device failure in vivo,[52] which is important for implants.[53,54]”

What the sources say

• Partly supported

  The ‘Assessment of the foreign body response’ section (page not specified) reports a substantially reduced response to implanted hydrogel hybrid probes after 1, 3, and 6 months. The paper also reports stable recording performance in mice, but it does not establish that hydrogels generally prevent device failure in vivo. Its Discussion presents cell-adhesion enhancers as a possible future addition, not a demonstrated ability to regulate cell or protein attachment. (LLM, conf=0.88)

  cites [52] sourceExperimental — accuracy not yet measured
• Partly supported

  The paper’s introductory passage (indexed as Abstract; page unavailable) says hydrogels resemble human tissues and may serve as coatings for long-term bio-integration. Its discussion of device reliability concerns robust hydrogel–component interfaces under deformation, not prevention of failure in vivo. The copy does not report reduced foreign-body inflammation or regulation of cell or protein attachment. (LLM, conf=0.86)

  cites [53] sourceExperimental — accuracy not yet measured
• Partly supported

  The ‘Advancing the state of the art in implanted electrodes’ section says hydrogels or other soft coatings can help clear inflammatory factors, and that functionalized surface coatings can regulate cell or protein attachment (page unavailable). It does not say hydrogels greatly reduce inflammation or prevent device failure in vivo; its examples of attachment control include PEG and ECM coatings. (LLM, conf=0.87)

  cites [54] sourceExperimental — accuracy not yet measured

In the paper

“To make the replacement from current rigid electronics, e-skin requires a low mechanical modulus (high stretchability),[28–30] aesthetic appearance,[31] low-cost,[32] lightweight design,[29] large-scale fabrication techniques,[32] self-healing properties,[30] and be thermally stable.[29]”

What the sources say

• Partly supported

  The Introduction (page not provided) contrasts low-modulus, compliant flexible electronics with bulky, rigid systems, and §2.2 identifies modulus and stretchability as important for skin contact. §2.5 describes an ultra-lightweight wearable device. The paper does not establish thermal stability as a requirement for e-skin; the citing sentence attaches this source to that point as well. Its other listed properties are assigned to other citations. (LLM, conf=0.85)

  cites [28] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [29] source
• ?Not checked — no full text of the sourcecites [30] source
• Partly supported

  The Introduction’s “Aesthetics” section says soft-tissue prostheses should look realistic and aesthetically pleasing. The cited paper discusses prosthetic materials, however, not aesthetic appearance as a requirement for e-skin replacing rigid electronics. In the citing sentence, this source appears attached only to “aesthetic appearance,” not to the other listed properties. (LLM, conf=0.82)

  cites [31] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [32] source

In the paper

“Additional shortcomings include their limited environmental feedback, power supply, and failure to autonomously respond appropriately to external (possibly dangerous) stimuli.[24]”

What the sources say

• Supported

  The review reports that relatively few active lower-limb prosthetic and orthotic devices explicitly sense and adapt to their environment (Explicit environmental sensing), and that many portable powered devices operate near their continuous-performance limits (Conclusions). Its Safety mechanisms section says users may need to handle trips and slips themselves until advanced balance-recovery controls are developed. Together, these passages support the cited summary of shortcomings, in the context of the devices reviewed. (LLM, conf=0.88)

  cites [24] source

In the paper

“Notably, 3D flexible electronics maintain conductivity while gaining an extra dimension and thus more sophisticated functions that 2D electronics do not offer (Figure 3E and 3F).[70–72]”

What the sources say

• Partly supported

  The ‘Route to morphable 3D mesostructures and electronics’ section (page unavailable) reports unchanged transistor characteristics after a 3D shape change, while its LED example shows that electrical connectivity can change with the assembly route. The ‘Morphable 3D radio frequency electronic devices’ section describes shape-dependent functions, including a circuit that switches between a filter and an inductor. These examples support retained electrical operation and added reconfigurable functions, but not the blanket claims that conductivity is always maintained or that 2D electronics cannot offer those functions. (LLM, conf=0.85)

  cites [70] sourceExperimental — accuracy not yet measured
• Partly supported

  In “Inspired by Wind-Dispersed Seeds” (page unavailable), the paper describes converting planar electronic precursors into 3D fliers and reports electrical properties consistent with devices made on planar wafers. It also demonstrates functional 3D fliers, but does not establish the broader claim that 3D flexible electronics offer more sophisticated functions unavailable to 2D electronics. (LLM, conf=0.84)

  cites [71] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [72] source

In the paper

“Prosthetics can be controlled by neural electrophysiological signals (will be further discussed in section 4) and mechanical signals such as gestures[189,190] or body movements.[191]”

What the sources say

• ?Not checked — no full text of the sourcecites [189] source
• Partly supported

  The Results and Discussion section reports that sensors detecting finger bending and gestures transmitted signals to control a robotic hand. It does not identify that hand as a prosthesis. This citation is attached to the gesture-control part of the sentence, not the separate claims about neural signals or body movements; the gesture mechanism is supported, but its application to prosthetics is not established by this paper. (LLM, conf=0.84)

  cites [190] sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section (“Wearable closed-loop HMI system”; page unavailable) shows that sensors detect shoulder and hand movements and use those signals to control a robotic hand. The paper calls it a robotic hand with prosthetic skin, not a prosthesis used by a person, so it supports movement-based control but not that application specifically. Citation [191] is attached to “body movements,” not to the separate claim about neural electrophysiological signals. (LLM, conf=0.82)

  cites [191] sourceExperimental — accuracy not yet measured

In the paper

“Compared with resistive sensors, capacitive sensors have excellent linearity with lower power consumption.[119]”

What the sources say

• Partly supported

  The Results and Discussion (p. 5) state that the capacitive sensor had better linearity than the resistive sensor, and Figure 2 (p. 3) reports a linear capacitive response. Our copy does not report lower power consumption for the capacitive sensor; a search for power consumption found no supporting passage. (LLM, conf=0.87)

  cites [119] sourceExperimental — accuracy not yet measured

In the paper

“Furthermore, the transparency based on polymers, or nanomesh materials, also contributes to greater aesthetic properties of flexible devices.[77,78]”

What the sources say

• ?Not checked — no full text of the sourcecites [77] source
• Partly supported

  The Abstract and Introduction (pages not provided) describe transparent polymer-based elastomers for flexible electronics and explain that a lack of transparency hinders optical applications. They do not say transparency improves a device’s aesthetic properties or discuss nanomesh materials. (LLM, conf=0.83)

  cites [78] sourceExperimental — accuracy not yet measured

In the paper

“Thermocouples can generate a potential difference under different temperature between different materials based on the thermoelectric (Seebeck) phenomenon.[216]”

What the sources say

• Supported

  The Introduction states that thermocouples use the Seebeck effect and generate a voltage proportional to the temperature difference between their measurement point and open ends. It also describes their two arms as being made of conductors, semiconductors, or a combination, supporting the claim about different materials. (LLM, conf=0.94)

  cites [216] source

In the paper

“In medical rehabilitation, temperature sensing is commonly used to monitor the body, which can be a reference for health.[225,226]”

What the sources say

• Partly supported

  The Results section on continuous exercise monitoring describes axillary temperature as a marker of health and performance and demonstrates its measurement during exercise. The Discussion identifies rehabilitation as a possible application for full-body monitoring, but does not say temperature sensing is commonly used in rehabilitation. (LLM, conf=0.85)

  cites [225] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [226] source

In the paper

“For measuring spatial temperature gradients, an organic active matrix with polymer PTC sensor pads was fabricated on a PI substrate, which exhibits a bending insensitivity and high spatial resolution (Figure 7A and 7B).[103]”

What the sources say

• Supported

  The Results describe a 12 × 12 organic active matrix with polymer PTC sensor pads fabricated on a polyimide substrate and report temperature-gradient mapping at about 5-mm resolution. The Significance section calls the approach high-spatial-resolution, while the Results report very little change in the sensor’s thermal response after bending. The bending test was on the sensor film, rather than explicitly on the complete active-matrix sheet. (LLM, conf=0.87)

  cites [103] source

In the paper

“As a heat source and sensor, the human hand can effectively perceive the heat dissipation ability for fluid flow rate sensing[223] and identifying the thermal characteristics of surface materials.[224]”

What the sources say

• ?Not checked — no full text of the sourcecites [223] source
• Supported

  Citation [224] is attached to identifying the thermal characteristics of surface materials, not to fluid-flow-rate sensing [223]. The Background and Related Research section explains that a finger’s temperature creates heat flow to or from a contacted material, which its nerves detect, and that this reveals thermal conductivity and diffusivity useful for material identification. (LLM, conf=0.92)

  cites [224] source

In the paper

“Whereas the forward voltage in PN junction temperature sensors can vary with temperature, converting the heat into electrical signals in the diodes and transistors.”

What the sources say

• Partly supported

  The ‘Characterization of SiNR temperature sensors’ section (page unavailable) describes p–n-diode sensors calibrated by plotting voltage at a fixed current against temperature. This supports the temperature-dependent voltage claim for diodes, but the cited passages do not establish it for transistors. (LLM, conf=0.85)

  cites [106] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [217] source

In the paper

“Researchers demonstrated the sensor’s functions by using a robotic hand equipped with e-skin to grab and measure the temperature of different cups. (Figure 7D)”

What the sources say

• Partly supported

  In “Intrinsically stretchable rubbery sensors for robotic skins” (pp. 5–6), the paper describes temperature sensors on a robotic hand touching hot and cold cups and records their responses. It says the hand touched the cups, but does not establish that it grabbed them. (LLM, conf=0.84)

  cites [222] sourceExperimental — accuracy not yet measured

In the paper

“Long-term monitoring of body temperature by using flexible temperature sensors is convenient and can effectively prevent medical abnormalities such as foot ulcers.[227]”

What the sources say

• Partly supported

  The Introduction (page unavailable) describes wearable sensors intended to monitor body, skin, and ulcer temperature, and says a temperature rise can warn of ulcer onset. But the Introduction says continuous wearable ulcer-temperature monitoring was not yet available, while the Conclusion (page unavailable) says the sensors still needed clinical validation. The paper therefore does not establish that convenient long-term monitoring with these sensors effectively prevents foot ulcers. (LLM, conf=0.88)

  cites [227] sourceExperimental — accuracy not yet measured

In the paper

“Sweat detection can respond in real-time to changes in the volume and rate of sweat loss,[241,242] as well as the concentration of various ions and metabolites in the body.[45,243–245]”

What the sources say

• Partly supported

  The Introduction (page unavailable) describes real-time sweat-biomarker sensing, and the Results section on aquatic and dryland athletes reports measurements of sweat volume, rate, and chloride concentration. It does not demonstrate real-time measurement of changes in all these quantities: the Methods say volume and chloride readings were taken after exercise. The Results discuss metabolite testing as a potential application, not a demonstrated measurement. (LLM, conf=0.84)

  cites [241] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [242] source
• ?Not checked — no full text of the sourcecites [45] source
• Supported

  The citation is most relevant to the sentence’s claim about analyte concentrations, rather than the separately cited claim about sweat volume and rate. The Introduction (page unavailable) says wearable sensors can analyze sweat electrolytes and metabolites, while the Results describe a sensor that responds to changes in sodium concentration. (LLM, conf=0.86)

  cites [243] source
• Partly supported

  The Introduction describes real-time measurement of local sweat volume and rate, and the Results section on smartphone-based analysis describes measuring sweat volume and chloride concentration. The cited paper does not establish detection of various metabolites or measurement of ion and metabolite concentrations in the body; its demonstrated chemical measurement is sweat chloride. (LLM, conf=0.87)

  cites [244] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [245] source

In the paper

“Thus, adding chemical sensing capability to prostheses can prevent users from being exposed to harmful chemicals, such as organophosphate pesticide residues in agricultural products, and subsequently reduce the potential risk of harm.[239]”

What the sources say

• Partly supported

  The Introduction (page unavailable) says organophosphate pesticide residues in agricultural products pose a health threat and that remote robotic sampling can avoid placing people in unsafe environments. The study equips robotic manipulators with chemical sensors, not prostheses worn by users, so it does not establish that adding sensing to prostheses prevents users’ exposure or reduces their risk of harm. (LLM, conf=0.88)

  cites [239] sourceExperimental — accuracy not yet measured

In the paper

“In addition to monitoring patient fatigue, sweat sensing can also be used for psychological applications by monitoring the level of stress hormones such as cortisol and norepinephrine.[249,250]”

What the sources say

• Partly supported

  The Introduction describes sensing cortisol in sweat for stress monitoring, and the Conclusion proposes applications in mental-health monitoring while noting that the study tested only physical stress stimuli. The cited paper does not support monitoring norepinephrine or patient fatigue. (LLM, conf=0.88)

  cites [249] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [250] source

In the paper

“Although chemical sensing is not sufficiently developed for prosthesis, in the field of health monitoring, sweat chemical sensing (including electrochemical and optical modalities) is well-studied and highly relevant to rehabilitation exercises.[240]”

What the sources say

• Partly supported

  The State-of-the-Art Wearable Sweat Sensors section (page unavailable) describes electrochemical and optical sweat sensing and its use in fitness and health monitoring. It does not assess whether chemical sensing is sufficiently developed for prostheses or establish that sweat sensing is highly relevant to rehabilitation exercises; its exercise discussion concerns fitness monitoring. (LLM, conf=0.87)

  cites [240] sourceExperimental — accuracy not yet measured

In the paper

“In recent years, with the popularity of smartphone cameras, the quantitative measurement of color is highly attractive for in-home monitoring of optical sensors.[237]”

What the sources say

• Partly supported

  The Results (Hybrid, battery-free, skin-mounted system for sweat sensing; page unavailable) say a smartphone camera captures images for colorimetric analysis, and Materials and Methods (Data collection and analyses; page unavailable) describes using those images to determine sweat chloride and pH. The paper does not establish that smartphone-camera popularity makes quantitative color measurement attractive for in-home monitoring of optical sensors generally. (LLM, conf=0.84)

  cites [237] sourceExperimental — accuracy not yet measured

In the paper

“Motor intention originates from the primary motor cortex of the brain and is transmitted from the central nervous system to the peripheral nervous system in the form of electrophysiological signals, which are translated into mechanical contractions of the muscles.[269]”

What the sources say

• Partly supported

  The Abstract reports that the interface derived motor commands from neuronal activity in primary motor cortex (M1). It does not establish that motor intention originates there or describe signals traveling from the central to the peripheral nervous system and producing muscle contractions; the interface instead bypassed the body. No page number is provided in our copy. (LLM, conf=0.84)

  cites [269] sourceExperimental — accuracy not yet measured

In the paper

“Due to new sensing mechanisms, e-skin can exhibit greater functionality beyond the human skin, such as proximity[259,260] and magnetic field sensing,[261,262] which can be seen as complements to the natural skin capabilities, allowing HMI devices to obtain richer information about the external environment.”

What the sources say

• Supported

  The Introduction describes a deformable electronic-skin sensor that detects a nearby finger using projected electric fields and explicitly calls this proximity sensing. The citing sentence attaches separate citations to magnetic-field sensing, so that capability is not attributed to this paper. (LLM, conf=0.92)

  cites [259] source
• ?Not checked — no full text of the sourcecites [260] source
• Supported

  The Introduction describes an e-skin with magnetic functionality that provides complementary information channels and enables touchless interaction with virtual objects. It also explains that its magnetic field sensors function as proximity sensors, supporting the cited example of sensing beyond natural touch. (LLM, conf=0.91)

  cites [261] source
• ?Not checked — no full text of the sourcecites [262] source

In the paper

“PPG uses optical differences between the vascular and other tissues or oxygenated and deoxygenated hemoglobin to measure heart rate,[263] blood oxygen,[267] and other in-depth cardiovascular information.[268]”

What the sources say

• Partly supported

  The citation [268] is attached to “other in-depth cardiovascular information.” In the Ultraflexible organic photodetectors section (page not specified), the paper says reflected light tracks blood-vessel volume pulsation, but it reports calculating pulse rate and does not identify other in-depth cardiovascular measurements. (LLM, conf=0.78)

  cites [263] sourceExperimental — accuracy not yet measured
• Supported

  The citation [267] is attached to the blood-oxygen part of the sentence. The Abstract states that pulse oximetry uses PPG at two wavelengths to measure oxygen saturation by distinguishing oxyhemoglobin from deoxyhemoglobin; the Analytical Models section describes calculating saturation from those optical signals. (LLM, conf=0.93)

  cites [267] source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [268]

In the paper

“Subcutaneous invasive electrophysiological signals can be derived through brain interfaces,[270–272] using electrocorticography (ECoG)[273,274] on the cortical surface as well as local field potentials (LFP)[275,276] extracted from electrodes inserted into the cortex.”

What the sources say

• Supported

  The Abstract (page not provided) describes ECoG as recorded by subdural electrodes on the cortical surface and LFP as recorded by a small electrode in the brain. This supports the cited distinction between the two recording locations. (LLM, conf=0.88)

  cites [270] source
• ?Not checked — no full text of the sourcecites [271] source
• Supported

  The review’s opening section describes implanted intracortical electrode arrays used to read out neural signals through brain–machine interfaces. That supports the claim at this citation’s placement, [270–272], which follows “brain interfaces”; the ECoG and LFP details carry separate citations. (LLM, conf=0.90)

  cites [272] source
• ?Not checked — no full text of the sourcecites [273] source
• ?Not checked — no full text of the sourcecites [274] source
• Supported

  The citation to this paper is attached to the brain-interface portion of the sentence; ECoG and LFP have separate citations. The Methods describe an intracortical microelectrode array implanted in motor cortex and neural signals recorded through the interface, supporting the cited example of invasive signal acquisition. (LLM, conf=0.91)

  cites [275] source
• ?Not checked — no full text of the sourcecites [276] source

In the paper

“Common wearable techniques include electroencephalography (EEG),[281,282] surface electromyography (sEMG),[283,284] and electrooculography (EOG),[285] which capture the electrical activity of the brain, muscles, and eye movements respectively.”

What the sources say

• Supported

  This citation is attached to EEG, so the sEMG and EOG portions are not attributed to this paper. The Introduction describes EEG as a noninvasive neural signal, and Materials and Methods (Frequency-Domain Electrical Source Imaging; page unavailable) relates scalp-recorded EEG to cortical current density, supporting the description of EEG as capturing brain electrical activity. (LLM, conf=0.91)

  cites [281] source
• ?Not checked — no full text of the sourcecites [282] source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [283]
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [284]
• ?Not checked — no full text of the sourcecites [285] source

In the paper

“In contrast, ECoG electrodes are generally placed on the surface of the cerebral cortex and can collect signals without penetrating the tissue, with a larger recording area with relatively less damage to the brain.[270]”

What the sources say

• Partly supported

  The passage labeled Abstract describes ECoG as recording with subdural grid electrodes on the cortical surface, in contrast to electrodes placed in the brain for LFP recordings. This supports surface recording without penetrating brain tissue, but our copy does not establish a larger recording area or relatively less brain damage. (LLM, conf=0.87)

  cites [270] sourceExperimental — accuracy not yet measured

In the paper

“Recently, there has been a focus on wireless transmission of data and energy for flexible implants, which can reduce messy wires and improve user mobility and social interaction.[325]”

What the sources say

• Partly supported

  The “Limitations of tethered systems” section reports that, compared with untethered mice, mice with fiber tethers had reduced mobility and social interaction. The paper also discusses flexible components for wireless implants, but the cited behavioral comparison concerns animals; it does not establish improved mobility or social interaction for human implant users. (LLM, conf=0.84)

  cites [325] sourceExperimental — accuracy not yet measured

In the paper

“Compared to commercially available rigid electrodes based on silicon wafers and rigid packaging, flexible electrodes are also more biocompatible (details described in section 2) and can adapt to the complex tissue geometries of the human body,[293–295] such as sulci in the cerebral cortex,[296,297] bundled peripheral nerves,[298] and stretching motions on the skin.[299]”

What the sources say

• Partly supported

  The Introduction says that subcellular electrode designs such as carbon fiber arrays reduce mechanical mismatch and inflammatory impact, which supports part of the biocompatibility rationale. It does not establish a comparison between flexible and rigid electrodes or say that flexible electrodes conform to complex tissue geometries. No page number is available in our copy. (LLM, conf=0.74)

  cites [293] sourceExperimental — accuracy not yet measured
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [294]
• Partly supported

  The Results (p. 3) say the soft electrodes can be adapted to the anatomy of cortical, spinal, and peripheral-nerve tissues, and the technology description (p. 4) calls their materials biocompatible. The paper does not establish that they are more biocompatible than rigid silicon-wafer electrodes, nor does it demonstrate adaptation to human anatomy. (LLM, conf=0.88)

  cites [295] sourceExperimental — accuracy not yet measured
• Partly supported

  The Abstract (page not provided) says flexible, non-penetrating subdural arrays minimize irritation and injury to brain tissue compared with penetrating arrays. Electrode array fabrication and testing (page not provided) shows that this flexible device can fold to access cortical sulci. The paper does not establish a comparison with commercially available silicon-wafer electrodes and rigid packaging, or show that flexible electrodes adapt to complex tissue geometries throughout the human body. (LLM, conf=0.86)

  cites [296] sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction says flexible neural probes comply with brain tissue, and the Results (Long-term chronic stability of Neurotassels; page unavailable) report less neuronal cell loss around Neurotassels than around silicon probes in mice. This supports a brain-implant biocompatibility advantage, but the paper does not establish that flexible electrodes adapt to complex geometries throughout the human body. (LLM, conf=0.88)

  cites [297] sourceExperimental — accuracy not yet measured
• Supported

  The citation is attached specifically to peripheral nerves as an example of tissue geometry that a flexible electrode can accommodate. In Results > Peripheral nerve interface design and fabrication (page unavailable), the paper describes a low-stiffness thin-film array and a nanoclip tailored to nerve dimensions. It supports that peripheral-nerve example; the broader biocompatibility comparison is associated with other citations in the sentence. (LLM, conf=0.89)

  cites [298] source

In the paper

“Apart from the conductive functional materials, the insulative packaging materials are also a critical part of sensing recorders. Many popular insulating soft materials have been used for packaging sensing electrodes,[318] such as PI,[319,320] PDMS,[321,322] hydrogel,[323,324] etc., as they have suitable mechanical, dielectric and biological properties.”

What the sources say

• ?Not checked — no full text of the sourcecites [318] source
• Supported

  The Introduction (p. 2) identifies several flexible materials used for neural probes and describes polyimide’s insulation, mechanical, and biocompatibility properties. The Flexible polyimide probes section (p. 3) reports applying a second polyimide layer to insulate the probes, directly illustrating the cited point about insulating materials used around sensing electrodes. (LLM, conf=0.88)

  cites [319] source
• ?Not checked — no full text of the sourcecites [320] source
• ?Not checked — no full text of the sourcecites [321] source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [322]

In the paper

“Various electrodes, with different materials, shapes, and characteristics, can be places on or in the human brain, peripheral nerves, and muscles to record ECoG,[300] LFP,[301] ENG[302] and EMG[303] respectively.”

What the sources say

• Supported

  Citation [300] is attached to ECoG, not to the separate LFP, ENG, or EMG examples. The paper’s Abstract describes intracranial ECoG electrodes used in clinical epilepsy monitoring and reports that its own electrode arrays recorded ECoG from the cortical surface in animal models. (LLM, conf=0.90)

  cites [300] source
• ?Not checked — no full text of the sourcecites [301] source
• Partly supported

  The Methods describe electrodes implanted around rat sciatic nerves and electrophysiological recording of evoked compound action potentials. This supports peripheral-nerve monitoring, the portion linked to citation [302], but the recording electrodes were placed in the rat hind paw; the paper does not establish ENG recording from electrodes on or in human peripheral nerves. (LLM, conf=0.86)

  cites [302] sourceExperimental — accuracy not yet measured

In the paper

“As invasive brain monitoring electrodes, ECoG and LFPs play an important role in examining motor intention and elucidating the fundamental pathogenesis of various neurological disorders, such as epilepsy and Parkinson’s disease.[330,331]”

What the sources say

• Partly supported

  The paper describes using ECoG to study motor movement and imagery, supporting the motor-intention portion of the claim. The Methods describe epilepsy as the reason participants had electrodes implanted, not as a disorder whose pathogenesis was investigated. The paper does not establish the claimed role for LFPs or investigate Parkinson’s disease pathogenesis. (LLM, conf=0.88)

  cites [330] sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction describes cortical and subcortical field potentials as probes of neural synchronization, and the Results use those recordings to decode motor states in people with Parkinson’s disease. The paper does not establish that they examine motor intention or elucidate the fundamental pathogenesis of epilepsy; epilepsy is mentioned in the Introduction as a condition treated with implanted stimulation devices. (LLM, conf=0.86)

  cites [331] sourceExperimental — accuracy not yet measured

In the paper

“With further evolution of processing technology, these materials can be fabricated into ultra-thin and ultra-fine forms, thus enabling high-density integration, as well as endowing some flexibility.[310,311]”

What the sources say

• Supported

  The Design and characteristics section (page not provided) describes fabricated NeuE structures about 0.9 μm thick, with interconnect widths as small as 1 μm. It also says higher-resolution fabrication could increase channel integration without sacrificing flexibility, supporting the attributed connection between finer fabrication, integration and flexibility. (LLM, conf=0.91)

  cites [310] source
• Supported

  The Results and Discussion describe lithographic processing of an ultrathin silicon-nanomembrane structure in a flexible electronic platform. They also report a 64-site active matrix and explain how small features and the dual-sided design enable high resolution and effective use of area for electronics. The claim is supported for the silicon-based materials studied here. (LLM, conf=0.86)

  cites [311] source

In the paper

“Unlike static neural implants mentioned above, in vivo ECG recording equipment on the surface of the heart is subjected to dynamic movement.[375–378]”

What the sources say

• Partly supported

  The Biological modulation section (page unavailable) describes flexible devices contacting a rat heart’s surface while the heart contracted, supporting the dynamic-movement aspect. But Methods: Isolated heart (page unavailable) says the heart was removed for these ECG experiments; the paper does not establish an in vivo cardiac recording setup or the comparison with static neural implants. (LLM, conf=0.85)

  cites [375] sourceExperimental — accuracy not yet measured
• Supported

  The Abstract describes cardiac recording devices designed for the curved, moving surfaces of beating hearts. The Results (Cardiac mapping in animal heart models) also state that the platform was demonstrated in vivo. The comparison with static neural implants is framing in the citing sentence, rather than a finding attributed to this source. (LLM, conf=0.88)

  cites [376] source
• Partly supported

  In “Assembly of a fully viscoelastic device and in vivo validation” (page not provided), the paper reports that an electrode array recorded ECG on a mouse heart and remained in place as the heart continued to beat. This supports the dynamic-movement point, but the paper does not characterize neural implants as static or establish the stated contrast. (LLM, conf=0.84)

  cites [377] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [378] source

In the paper

“As materials and manufacturing processes evolve, more micro-wires are becoming flexible alternatives to microneedle arrays, which are able to remain relatively stationary as the brain moves, improving signal stability while reducing tissue damage.[339–342]”

What the sources say

• Partly supported

  The Introduction (p. 3) says the fine carbon fibers are much more compliant than conventional tungsten microwires and may reduce biological response; the Abstract (p. 1) likewise describes reduced inflammation as a promise of the technology. The paper does not show that the electrodes remain stationary as the brain moves or that this improves signal stability. Its in vivo recording demonstration was acute, so it does not establish reduced tissue damage over time. (LLM, conf=0.86)

  cites [339] sourceExperimental — accuracy not yet measured
• Partly supported

  The Methods describe chronically implanted microwire arrays, and the Results report stable recordings over recording sessions and viable single-unit recordings 18 months after implantation (page unavailable in our copy). The Discussion contrasts slow microwire implantation with the Utah array. The paper does not establish that the microwires flex to remain stationary as the brain moves, or that this mechanism improves signal stability or reduces tissue damage; its reference to design flexibility concerns which microwires the circuit board can accommodate. (LLM, conf=0.85)

  cites [340] sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction (page unavailable) describes thin, flexible polymer probes as an alternative to rigid electrode arrays and says their flexibility should improve biocompatibility. The Threads section describes a design intended to minimize tissue displacement. The paper does not say these probes remain relatively stationary as the brain moves or that this improves signal stability; it also does not demonstrate reduced tissue damage. (LLM, conf=0.86)

  cites [341] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [342] source

In the paper

“In contrast, surface electrodes generally are more flexible,[364,369] whose morphology in the limited space between this epimysium and the skin is either 1D or 2D thin electrode layers[370,371] or very tiny electrodes (millimeter or even sub-millimeter level).[369,372]”

What the sources say

• ?Not checked — no full text of the sourcecites [364] source
• Partly supported

  The Epimysial recordings section (p. 4) describes a flexible, 20 µm-thick electrode substrate with submillimeter detection sites, supporting this paper as an example of a thin, flexible electrode. It does not establish that surface electrodes generally are more flexible than other electrodes; its devices are epimysial electrodes mounted on muscle. (LLM, conf=0.84)

  cites [369] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [370] source
• Partly supported

  The Results describe a flexible printed circuit board with 0.2 × 0.2 mm recording pads that contact nerve or muscle, and a mote placed on a muscle surface beneath the replaced skin. This supports the cited example of a tiny surface-contact electrode, but does not establish that surface electrodes are generally more flexible than other electrodes. The 1D/2D morphology clause cites other sources, not this paper. (LLM, conf=0.84)

  cites [371] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [372] source

In the paper

“Nevertheless, ENG signals have a low signal-to-noise ratio and limited stability.[363]”

What the sources say

• Partly supported

  The Materials and methods section (Figure 2 description; page unavailable) says LIFE signals commonly have quite low signal-to-noise ratios. The Discussion (page unavailable) describes the interface as quite stable during the trials and says long-term stability still needs investigation, so it does not establish the claimed limited stability. (LLM, conf=0.88)

  cites [363] sourceExperimental — accuracy not yet measured

In the paper

“The bending stiffness of this implant is comparable to that of a neuron’s axon, enabling biocompatible high-resolution LFP recording.[310]”

What the sources say

• Partly supported

  The Design and characteristics section states that the NeuE interconnect’s calculated bending stiffness is comparable to a neuron’s axon. The Summary and outlook reports negligible immune response, and Stable single-unit recording following implantation reports stable recordings of individual neurons. Our copy does not establish high-resolution local field potential (LFP) recording; its recording results concern single-unit spikes. (LLM, conf=0.88)

  cites [310] sourceExperimental — accuracy not yet measured

In the paper

“In vivo Vagus nerve stimulation was carried out on a rabbit to demonstrate the validity of the device (Figure 8F).”

What the sources say

• Supported

  The Results section (“Vagus nerve stimulation”) explicitly says that in vivo experiments on a rabbit were carried out to demonstrate the twining electrodes’ practical biomedical implications. It also reports electrical stimulation of the rabbit’s right vagus nerve with the electrode and a resulting decrease in heart rate. (LLM, conf=0.94)

  cites [358] source

In the paper

“While in vitro and on-body artificial sensory neurons can communicate to batches of afferent cells using optical methods, they currently lack the ability to selectively connect with individual neurons (Figure 10A).[426,427]”

What the sources say

• ?Could not verify

  The Background and Methodology > Sensory feedback system passages describe prosthetic-hand feedback through auditory cues, not optical communication with afferent cells. Searches for artificial sensory neurons, optical methods, and selective connections to individual neurons found no passage addressing the claim in our copy; page numbers were unavailable. (LLM, conf=0.76)

  cites [426] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [427] source

In the paper

“The problem with inaccurate communication channels in medical robotics is that it goes against visual cues from the user leading to cognitive confusion,[428] which is the number one reason why 40% of amputees reject wearing a prosthetic device.[19]”

What the sources say

• ?Could not verify

  The citation to this paper follows the claim that cognitive confusion is the leading reason 40% of amputees reject prostheses. The Introduction discusses the lack of sensory feedback, and Materials and Methods describes experiments with one amputee, but neither reports a rejection rate or identifies cognitive confusion as a reason for rejection (searched: 40%, rejection, abandonment, visual cues). (LLM, conf=0.78)

  cites [428] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [19] source

In the paper

“CNNs have additionally been shown to detect Alzheimer’s disease in MRI images.[445]”

What the sources say

• ?Could not verify

  The Introduction discusses convolutional neural networks only as classifiers for hand-gesture recognition. The Method and Conclusions describe an ANN using surface EMG signals, not MRI images. Searches for Alzheimer’s disease and MRI found no passage addressing the attributed claim. (LLM, conf=0.76)

  cites [445] sourceExperimental — accuracy not yet measured

In the paper

“It is worth noting, most single tendon drive does not have the ability to actuate bidirectionally (no”

What the sources say

• ?Could not verify

  The Introduction (page unavailable) describes electro-ionic soft actuators as artificial muscles, and the Results (page unavailable) discuss their bending performance. Searches for single-tendon drives, bidirectional actuation, multiple cables, and agonist–antagonist muscles found no passage addressing the claimed limitation or cable requirement. (LLM, conf=0.75)

  cites [486] sourceExperimental — accuracy not yet measured
• ?Could not verify

  The Methods section describes a string attached to the motor’s rotor to measure torque, but does not discuss single-tendon drives. The Results and discussion passages concern dielectric-elastomer actuation, not whether tendon drives require multiple cables for bidirectional motion or recovery. Searches for tendon, cable, bidirectional actuation, and agonist–antagonist muscles found no passage addressing the claim; page numbers were not provided. (LLM, conf=0.76)

  cites [487] sourceExperimental — accuracy not yet measured

In the paper

“Specifically, these skin-like thermal actuators can be used for prosthetic applications,[106,502] AR/VR feedback thermal stimulators,[191,503,504] and thermal therapy in rehabilitation.[505–507]”

What the sources say

• Supported

  This citation is attached to “prosthetic applications,” not the separately cited AR/VR or rehabilitation uses. The section on SiNR mechanical sensors with site-specific sensitivity describes stretchable thermal actuator arrays for prosthetic devices and artificial skin, including heaters that can maintain body temperature. (LLM, conf=0.93)

  cites [106] source
• ?Could not verify

  The Introduction and Results describe conformal thermal actuators used to heat skin at low levels and measure thermal transport; the Introduction says the experimental heating was below the human sensory threshold. The Discussion mentions dermatological studies and monitoring, but our copy does not address prosthetic use, AR/VR thermal feedback, or rehabilitation therapy (searched: prosthetic applications, feedback, rehabilitation, and therapeutic heating). (LLM, conf=0.73)

  cites [502] sourceExperimental — accuracy not yet measured
• Supported

  This source is cited for prosthetic applications, not the separately cited AR/VR and rehabilitation uses. In Results, “Wearable closed-loop HMI system,” the paper describes a temperature sensor on a prosthetic robotic hand controlling a soft thermal stimulator on human skin. (LLM, conf=0.93)

  cites [191] source
• ?Not checked — no full text of the sourcecites [503] source
• ?Not checked — no full text of the sourcecites [504] source
• ?Not checked — no full text of the sourcecites [505] source
• ?Not checked — no full text of the sourcecites [506] source
• ?Not checked — no full text of the sourcecites [507] source

In the paper

“The name is a generic term, because the “tendon” can be powered by many sources, such as shape memory alloys (SMAs),[479] SMPs,[480] and thermo-responsive polymers[481].”

What the sources say

• Supported

  Citation [481] is attached specifically to the thermo-responsive polymer example, not to the preceding shape-memory examples. The paper’s Abstract (page unavailable) describes polymer fiber muscles controlled by heat and a demonstration of joint manipulation, supporting their use as a thermally powered tendon-like actuator. (LLM, conf=0.90)

  cites [479] source
• ?Not checked — no full text of the sourcecites [480] source
• ?Not checked — no full text of the sourcecites [481] source

In the paper

“Broadly speaking, there are two types of pneumatic actuators: PAM-pneumatic artificial muscles (known as McKibben actuators) for linear actuation[471–473] and FEA- fluidic elastomer actuators, which are composed of low durometer rubber and driven by relatively low air pressure (3–8 psi) for bending.[474–476]”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [471]
• Supported

  The citation to this paper is attached to the FEA description. Section 2.1.4 describes FEAs as low-durometer rubber actuators driven at 3–8 psi, focuses on bending designs, and says they can be powered pneumatically. The preceding PAM claim has separate citations. (LLM, conf=0.92)

  cites [472] source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [473]
• ?Not checked — no full text of the sourcecites [474] source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [475]
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [476]

In the paper

“These two actuation methods can be applied to gloves, exoskeletons, and foot orthoses, providing a new paradigm for rehabilitation devices.[459]”

What the sources say

• ?Could not verify

  Section 2.3 describes an electrically heated nitrile glove that provides thermal feedback, not pneumatic or tendon actuation. Searches for pneumatic and tendon actuation, exoskeletons, foot orthoses, and rehabilitation found no passage addressing the attributed claim. The copy provides no page number for Section 2.3. (LLM, conf=0.78)

  cites [459] sourceExperimental — accuracy not yet measured

In the paper

“In contrast to soft pneumatic actuators that demand gas supplements and tendon driven actuators that require motors, EAP actuators are activated directly through electrical signals.[488,489]”

What the sources say

• ?Not checked — no full text of the sourcecites [488] source
• Partly supported

  The Introduction (page unavailable) says traditional pneumatic or hydraulic actuators require external compressors or pumps to supply pressurized fluid, and describes electrically responsive actuators. It does not discuss tendon-driven actuators or establish that they require motors; its electrically powered examples also do not substantiate the statement about EAP actuators as a class. (LLM, conf=0.86)

  cites [489] sourceExperimental — accuracy not yet measured

In the paper

“Endowing e-skins with”

What the sources say

• ?Could not verify

  The Introduction describes conformal thermal sensors for measuring skin temperature and thermal transport, while the Methods study healthy volunteers. Neither those passages nor the additional searches for amputees, prostheses, and social touch address whether temperature-enabled e-skins improve social interaction or prosthesis acceptance. (LLM, conf=0.76)

  cites [502] sourceExperimental — accuracy not yet measured

In the paper

“Common substrate materials include PI and hydrogel, which can remain chemically inert in vivo while also providing sufficient dielectric strength to prevent discharge breakdown.”

What the sources say

• ?Could not verify

  The Discussion sections on specificity, invasiveness, and heating risks discuss implantable optical devices and electrode insulation, but not PI or hydrogel substrates. Searches for those materials, chemical inertness, dielectric strength, and discharge breakdown found no passage addressing the claim in our copy. (LLM, conf=0.76)

  cites [515] sourceExperimental — accuracy not yet measured

In the paper

“Many studies have shown that feedback stimulation has the potential to not only regain tactile and thermal sensation in amputees,[106,518] but also improve their accuracy and confidence in manipulating the prosthesis, such as grasping objects.[157,455]”

What the sources say

• Supported

  The Abstract describes a prosthetic skin with tactile and temperature sensors and nerve-stimulation electrodes, and says it offers mechanical and thermal skin-like perception. The section on relaying sensory signals to peripheral nerves identifies enabling amputees to feel external stimuli as the goal. This supports the cited potential to restore sensation; the separate claim about manipulation accuracy and confidence is attached to other citations. (LLM, conf=0.84)

  cites [106] source
• ?Not checked — no full text of the sourcecites [518] source
• Supported

  The functional-performance results report that sensory feedback improved cherry-plucking success and significantly increased the subject’s self-reported confidence. The Discussion likewise reports improved fine-grasp performance and confidence. These passages support the prosthesis-manipulation clause; the separately cited thermal-sensation clause is not attributed to this source here. (LLM, conf=0.90)

  cites [157] source
• ?Not checked — no full text of the sourcecites [455] source

In the paper

“Stimulation holds tremendous promise for prosthetic information feedback[86,515] and rehabilitation.[329,516,517]”

What the sources say

• ?Not checked — no full text of the sourcecites [86] source
• Supported

  The Discussion (page unavailable) identifies prosthetic feedback about limb position and force as an application for closed-loop optogenetic stimulation. It also envisions therapeutic uses, including restoring function for people with foot drop or hand tremor. These passages support the claim as a statement of promise, not demonstrated clinical benefit. (LLM, conf=0.90)

  cites [515] source
• ?Not checked — no full text of the sourcecites [329] source
• ?Not checked — no full text of the sourcecites [516] source
• ?Not checked — no full text of the sourcecites [517] source

In the paper

“Optogenetic stimulation is based on opsins that are sensitive to light signals and only expressed in certain types of cells.”

What the sources say

• Supported

  The Extended Data Figure 2 caption reports that the channelrhodopsin construct was targeted to PV-positive interneurons and that its expression was observed only in PV-positive cells. The Methods’ optogenetic stimulation protocol describes delivering light to manipulate those neurons. In this study, the cell-type restriction was achieved through Cre-dependent targeting. (LLM, conf=0.88)

  cites [555] source

In the paper

“By genetically engineering these proteins into target neurons and controlling the light around them, these neurons can be activated or inhibited, which is particularly important for sensory and motor recovery and rehabilitation.[555]”

What the sources say

• Partly supported

  The Methods (AAV vectors; Optogenetic stimulation protocol, page unavailable) describe expressing channelrhodopsin-2 in targeted mouse neurons and manipulating them with light. The paper supports light-driven activation, but does not establish light-driven inhibition or a role in sensory and motor recovery or rehabilitation; its Discussion instead concerns amyloid reduction in mouse models. (LLM, conf=0.87)

  cites [555] sourceExperimental — accuracy not yet measured

In the paper

“Some Utah electrodes also use flexible substrates to improve their adaptability to complex surface.[544]”

What the sources say

• ?Could not verify

  The Materials and methods section describes an ECoG grid made from a silicone sheet with platinum electrodes, but does not identify it as a Utah electrode or say the sheet improves adaptability to complex surfaces. Searches for Utah electrodes and flexible substrates found no passage addressing that claim. (LLM, conf=0.75)

  cites [544] sourceExperimental — accuracy not yet measured

In the paper

“Although the second type, thin film flexible electrical stimulators, are available in non-inserted way on the cortical surface to reduce damage to brain tissue, the array density of such devices is much smaller than the first, and they are more suitable for stimulation of relatively large brain areas.[545,546]”

What the sources say

• Partly supported

  Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites [545] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [546] source

In the paper

“Many studies have delved into the relationship between optogenetic stimulation and neurological disorders and control, such as Parkinson's,[557] Alzheimer's,[558] spinal cord injuries,[559] and somatosensory nociception,[560,561] which have indistinguishable relationship with prosthetic control, feedback, and human movement rehabilitation.”

What the sources say

• ?Not checked — no full text of the sourcecites [557] source
• Supported

  Split into 2 atoms — 2 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites [558] source
• Partly supported

  Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites [559] sourceExperimental — accuracy not yet measured

In the paper

“Furthermore, stimulation of the spine can suppress pain and restore movement in paralyzed patients.[549]”

What the sources say

• Partly supported

  The Results and discussion section (‘Stimulation of the distal stump to alleviate muscular atrophy’; page unavailable) reports improved motor function after stimulation of injured sciatic nerves in rats. It does not establish that spinal cord stimulation suppresses pain or restores movement in paralyzed patients. (LLM, conf=0.82)

  cites [549] sourceExperimental — accuracy not yet measured

In the paper

“Furthermore, optogenetic stimulation can excite nerve cells and inhibit them, whereas electrical stimulation only excites nerves.[556]”

What the sources say

• Partly supported

  The Introduction (page unavailable) says light activation of channelrhodopsin-2 can create action potentials in neurons, supporting the excitation part. The passages do not establish that optogenetic stimulation can inhibit nerve cells or that electrical stimulation only excites nerves. (LLM, conf=0.85)

  cites [556] sourceExperimental — accuracy not yet measured

In the paper

“In addition, the combination of hydrogels and conductive polymers is a hot area of research since the former has mechanical properties similar to those of nerves and can provide a hydrophilic interface that reduces the adhesion of non-specific proteins.[58,539–542]”

What the sources say

• ?Not checked — no full text of the sourcecites [58] source
• ?Not checked — no full text of the sourcecites [539] source
• Partly supported

  The Introduction describes conducting-polymer hydrogels as promising because their high water content resembles tissue, and the Discussion says the resulting gels can be tuned to mimic biological tissue at soft interfaces such as the brain. The paper does not specifically establish a match to nerve mechanics or say that a hydrophilic interface reduces nonspecific protein adhesion; searches for protein adsorption and adhesion found no supporting passage. Page numbers are unavailable in our copy. (LLM, conf=0.86)

  cites [540] sourceExperimental — accuracy not yet measured
• ?Could not verify

  The Introduction describes carbon-fiber arrays on polyimide, and the Discussion and Conclusion addresses their use as nerve electrodes. Searches for hydrogels, conductive polymers, hydrophilic interfaces, and non-specific protein adhesion found no passage addressing the attributed claim. (LLM, conf=0.75)

  cites [541] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [542] source

In the paper

“Afferents and efferent projections from the spinal grey matter also carry somatosensory and motorial information.[515]”

What the sources say

• ?Could not verify

  The Introduction discusses stimulation of peripheral nerves and motor units, while the Results describe nerve stimulation and muscle-state feedback. Searches for afferent and efferent projections and spinal grey matter found no passage in our copy describing the claimed sensory and motor pathways. (LLM, conf=0.73)

  cites [515] sourceExperimental — accuracy not yet measured

In the paper

“Based on NFC data and power transmission, the device could be implanted over the mouse spinal cord in a minimally invasive way without impeding movement of the mouse (Figure 15J).”

What the sources say

• Partly supported

  The Discussion (page unavailable) confirms NFC powering and communication. The Results and implantation methods (pages unavailable) describe a subcutaneous base station over the thoracic vertebral column, a cuff around the sciatic nerve, and no detected gait or coordination impairment compared with sham surgery. The paper does not report implanting the cuff over the spinal cord; it discusses spinal-cord placement only as a possible future adaptation. (LLM, conf=0.88)

  cites [569] sourceExperimental — accuracy not yet measured

In the paper

“In addition to haptic feedback, electrical stimulation is also attractive pain relief and sensory-motor control for prosthetics and orthotics.[599,601]”

What the sources say

• Partly supported

  The Introduction (page unavailable) says that TENS electrodes are used for pain relief and that electrotactile stimulation can provide sensory feedback relevant to controlling prosthetic devices. It does not address orthotics or establish electrical stimulation as a method of sensory-motor control for them. (LLM, conf=0.86)

  cites [599] sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section (“Fully integrated ET system for augmented tactile sensations”; page unavailable) says the electro-tactile system could be applied to an artificial limb to give its user more information during operation. That supports a potential sensory-feedback application for prosthetics, but our copy does not describe pain relief, orthotics, or sensory-motor control beyond that proposed feedback. (LLM, conf=0.83)

  cites [601] sourceExperimental — accuracy not yet measured

In the paper

“This radio frequency-based optogenetics device and its successful operation for up to four weeks represent a novel stride into closed-loop wireless optogenetic modulation for expediting the adoption of optogenetics technologies in clinical applications (Figure 15H).[566]”

What the sources say

• Partly supported

  The Results describe an RF-powered, single-µ-ILED spinal device and report robust operation for more than four weeks, although average in-vivo survival was nearly three weeks. The device layout does not include a photodetector, and the Discussion presents sensing and wireless data transmission as possible future adaptations, not demonstrated closed-loop operation. The paper does not establish clinical adoption. (LLM, conf=0.90)

  cites [566] sourceExperimental — accuracy not yet measured

In the paper

“To solve the issue of insufficient stimulation force, multi-layer stacking structure,[592] or hydraulic pressure to amplify the force have been proposed.[63,593,594]”

What the sources say

• ?Could not verify

  The cited paper discusses pneumatic actuators and notes that increasing air pressure increases fingertip force (5.1.1 Normal Indentation, p. 9). Its discussion of actuator choice (4.1 Form Factor, p. 6) does not address hydraulic pressure as a force-amplification method. Searches for hydraulic force amplification and multi-layer stacking found no passage supporting those proposed solutions. (LLM, conf=0.73)

  cites [592] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [63] source
• ?Could not verify

  The Introduction and Conclusion describe amplification of vibration in a polymer-film touch button, not amplification of stimulation force. Searches for hydraulic pressure and force-amplifying multilayer stacks found no supporting passage; the multilayer actuator listed in the References concerns low-voltage operation. (LLM, conf=0.76)

  cites [593] sourceExperimental — accuracy not yet measured
• ?Could not verify

  The device description and Methods describe a thin PZT actuator-and-sensor array, while Supplementary Figure S7 describes pressure induced by an actuator’s expansion. Searches for hydraulic force amplification and multilayer stacking found no passage proposing either approach to address insufficient stimulation force. The copy therefore does not verify the claim attributed to this source. (LLM, conf=0.73)

  cites [594] sourceExperimental — accuracy not yet measured

In the paper

“Optogenetics was initially applied in rodent models to control neural circuits in the brain, but recent research efforts have shown great promise in modulating the activity of the spinal cord and peripheral nervous system.[556]”

What the sources say

• Partly supported

  The Results and conclusion (page numbers unavailable) show that light stimulation of ChR2-expressing spinal neurons in injured rats restored diaphragmatic activity, supporting the spinal-cord portion of the claim. The paper does not establish that optogenetics was initially applied to brain circuits in rodents or show modulation of peripheral nervous system activity; diaphragmatic muscle activity is an outcome, not evidence of direct peripheral nervous system modulation. (LLM, conf=0.86)

  cites [556] sourceExperimental — accuracy not yet measured

In the paper

“Moreover, in this demonstration, the facts that the parameters of the applied current on the skin determines the tactile stimulation comfort, and stimulation with a large current will induce muscle contractions were also verified.[599]”

What the sources say

• Partly supported

  In the Application demonstration > Electrical stimulation section (page unavailable), the authors increased current to identify comfortable thresholds and stopped when stimulation became uncomfortable or caused muscle twitching. This supports a relationship between current amplitude and comfort, and shows twitching can occur, but it does not establish that a large current will always cause muscle contractions or test the effects of multiple current parameters on comfort. (LLM, conf=0.85)

  cites [599] sourceExperimental — accuracy not yet measured

In the paper

“The result illustrates that subjects with force feedback were able to stop grasping at any desired level of gentle touch, while those without force feedback cannot.”

What the sources say

• ?Could not verify

  The Application demonstration section measures electrotactile perception and comfort thresholds, while the Discussion and conclusion describes a proof of concept for reusable electrodes. Neither reports a grasping experiment comparing subjects with and without force feedback; searches for stopping grasping at a desired gentle-touch level and for a feedback-versus-no-feedback comparison found no such result in our copy. (LLM, conf=0.75)

  cites [599] sourceExperimental — accuracy not yet measured

In the paper

“Electrical stimulation roughly includes two types, one with voltage control, which is capable of reducing skin burns feeling during stimulation, and the other with current control, which is relatively insensitive to differences in the feature of the skin interface.[360,599,600]”

What the sources say

• Partly supported

  The Electro-tactile stimulation section describes voltage- and current-regulated stimulation and says changes in interface impedance and load might not affect the current under current regulation. It says voltage regulation can minimize the chance of skin burns, not the feeling of burning during stimulation. Its two-type distinction is also specific to electro-tactile stimulation rather than electrical stimulation generally. (LLM, conf=0.86)

  cites [360] sourceExperimental — accuracy not yet measured
• Partly supported

  The Application demonstration > Electrical stimulation section (page unavailable) reports using a constant-current stimulator, supporting the existence of current-controlled stimulation. It does not compare current and voltage control, say that voltage control reduces skin-burn sensations, or establish that current control is relatively insensitive to skin-interface differences. (LLM, conf=0.78)

  cites [599] sourceExperimental — accuracy not yet measured
• ?Could not verify

  The Results section on network hubs and the Discussion address brain-network findings in children with epilepsy, not electrical stimulation. Searches for voltage control, current control, skin burns, and skin-interface features found no passage addressing the claim in our copy; no page numbers were provided. (LLM, conf=0.78)

  cites [600] sourceExperimental — accuracy not yet measured

In the paper

“For example, the large conformal attachment of prosthetic skin,[149] sweat trap between the wearable sensor and the skin,[75] and fixation of neural implants.[420]”

What the sources say

• Partly supported

  The cited paper supports the part of the sentence concerning large-area, conformable attachment: Related Work (p. 1) discusses skin modules covering robot surfaces, and “Mounting and Covering Robot Surfaces With Robot Skin” (p. 5) describes fastening skin patches to a robot. It does not establish the claimed problem for prosthetic skin attached at a human-device interface. The sweat-trap and neural-implant examples have separate citations and are not judged here. (LLM, conf=0.83)

  cites [149] sourceExperimental — accuracy not yet measured
• Supported

  The citation is attached to the sweat-trapping example, not the prosthetic-skin or neural-implant examples. The Introduction says sweat accumulates at the interface between electronic skin and human skin, impairing long-term sensing and adhesion; the Discussion describes sweat-permeable channels as a solution. (LLM, conf=0.93)

  cites [75] sourceExperimental — accuracy not yet measured
• ?Could not verify

  The citation [420] is attached to the fixation of neural implants, not the preceding examples of prosthetic skin or wearable sensors. The Introduction (p. 2) discusses neural coding as a communication channel, and the Results (p. 14) discuss synthetic tests of that method. A search for implant fixation and related terms found no passage in our copy addressing the claimed interface problem. (LLM, conf=0.76)

  cites [420] sourceExperimental — accuracy not yet measured

In the paper

“Promisingly, some prospective systematic works have been reported.[76,153]”

What the sources say

• Supported

  The Applications on amputees section describes an original study using standardized hand-function tests to compare a soft neuroprosthetic hand with a rigid one in two participants. That systematic experimental evaluation supports citing the paper as an example of the works reported; the sentence does not attribute a specific result to it. No page number is available in our copy. (LLM, conf=0.84)

  cites [76] source
• ?Not checked — no full text of the sourcecites [153] source

In the paper

“biofuel cells,[615] piezoelectric,[122] triboelectric,[127] solar[616] and other self-powered methods are prominent representatives of power supply for low-energy devices, truly enabling the user to move freely.”

What the sources say

• ?Not checked — no full text of the sourcecites [615] source
• ?Not checked — no full text of the sourcecites [122] source
• Partly supported

  The Results (Structure and working mechanism; page unavailable) show that the paper’s textile sensor generates an electrical signal through triboelectrification, supporting the citation’s triboelectric example. The paper does not show that this sensor supplies power to another low-energy device or that it enables unrestricted movement; its Results note that motion can affect pulse signals. (LLM, conf=0.84)

  cites [127] sourceExperimental — accuracy not yet measured

In the paper

“The on-body test of this system showed excellent results: after 15 minutes of training, the amputee was able to use the flexible prosthetic system to perform a range of precise gripping and social tasks, demonstrating tremendous potential for future applications.[76]”

What the sources say

• Supported

  In “Applications on amputees,” the paper reports that a subject with transradial amputation mastered the soft hand’s functions after 15 minutes of training. The same section describes precise object-handling tasks and interactions such as shaking hands; the Conclusion states the system’s potential for future applications. (LLM, conf=0.93)

  cites [76] source

In the paper

“As an exemplary case, a soft EMG controlled pneumatic neuroprosthetic with ionic capacitive sensors and flexible stimulators was developed to reduce manufacturing costs and achieve lightweight actuation.”

What the sources say

• Partly supported

  The Abstract and Conclusion describe a lightweight, potentially low-cost, pneumatically actuated hand controlled by EMG signals. The Design section describes ionic hydrogel-elastomer capacitive touch sensors and electrical stimulators, but does not say the stimulators are flexible. (LLM, conf=0.88)

  cites [76] sourceExperimental — accuracy not yet measured

In the paper

“Currently, some integrated work can enhance space utilization,[599] reduce manufacturing costs,[155] and achieve in-situ information sensing, processing, and actuating simultaneously for more accurate therapy.[548,605]”

What the sources say

• Partly supported

  The Introduction describes the paper’s electrode as low-cost, and “Environmental and financial cost” reports a $0.45 material cost per electrode and potential savings from reuse. It does not establish that integration reduces manufacturing costs; its comparison concerns long-term use versus disposable electrodes, and it excludes equipment costs. The paper does not report simultaneous in-situ sensing, processing, and actuation for therapy. (LLM, conf=0.82)

  cites [599] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [155] source
• ?Not checked — no full text of the sourcecites [548] source
• ?Not checked — no full text of the sourcecites [605] source