Automatic identification of artifacts in electrodermal activity data
Open accessAbstract
Recently, wearable devices have allowed for long term, ambulatory measurement of electrodermal activity (EDA). Despite the fact that ambulatory recording can be noisy, and recording artifacts can easily be mistaken for a physiological response during analysis, to date there is no automatic method for detecting artifacts. This paper describes the development of a machine learning algorithm for automatically detecting EDA artifacts, and provides an empirical evaluation of classification performance. We have encoded our results into a freely available web-based tool for artifact and peak detection.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
92% of collected SC data were classified as clean data using an artifact detection algorithm.
“In addition, 92% of the collected SC data were classified as clean data using an artifact detection algorithm [36].”
From Identifying Objective Physiological Markers and Modifiable Behaviors for Self-Reported Stress and Mental Health Status Using Wearable Sensors and Mobile Phones: Observational Study (2018) · cites this paper as [36]
The Results (IV.A; page unavailable) describe an artifact-detection algorithm that labels portions of electrodermal activity data clean or as artifacts. Neither that section nor the Methods (III.B; page unavailable) reports that 92% of collected skin-conductance data were classified as clean; searches for “92” and the clean-data proportion found no such result. (LLM, conf=0.85)
Experimental — accuracy not yet measuredThe automated classifier separated clean epochs and noisy epochs of SC data for further analysis.
“We developed scripts to download the data from sensors and check sensor readings automatically for quality using a previously developed and tested automated classifier [36]. This classifier separated clean epochs and noisy epochs of SC data for further analysis.”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
The citation lists haven’t been fetched yet
References 16 · Cited by 279
Research works printed but not linked: 8
- [5]Hedman EB In-situ measurement of Electrodermal Activity during Occupational Therapy PhD thesis MIT 2010
- [9]Poh M-Z Continuous assessment of epileptic seizures with wrist-worn biosensors PhD thesis MIT 2011
- [10]Reinhardt T Salivary cortisol, heart rate, electrodermal activity and subjective stress responses to the mannheim multicomponent stress test (mmst) Psychiatry research 198 1 106 111 2012 22397919
- [11]Sano A Discriminating high vs low academic performance, self-reported sleep quality, stress level, and mental health using personality traits, wearable sensors and mobile phones Body Sensor Networks (BSN) (to appear) 2015 PMC5431072 28516162
- [12]Sano A Picard R Stress recognition using wearable sensors and mobile phones ACII 671 676 IEEE 2013
- [13]Storm H The development of a software program for analyzing spontaneous and externally elicited skin conductance changes in infants and adults Clin Neurophysiology 111 10 1889 1898 2000 11018507
- [14]Wilhelm FH Roth WT Taking the laboratory to the skies: Ambulatory assessment of self-report, autonomic, and respiratory responses in flying phobia Psychophysiology 35 5 596 606 1998 9715103
- [15]Xu Y Wavelet transform domain filters: a spatially selective noise filtration technique Image Processing 3 6 747 758 1994 18296244
- Authors
- Taylor, S., Jaques, N., Chen, W., Fedor, S., Sano, A., & Picard, R.
- Affiliations
- Massachusetts Institute of Technology (US)
- Venue
- 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
- Publisher
- IEEE
- Published
- 2015-08
- Volume
- 2015
- Pages
- 1934-1937
- ISSN
- 2375-7477 (ISSN-L)
- PMC
- PMC5413200
- OpenAlex
- W2294683276
- Cited by
- 279
- References
- 16
- Citations by year
- 2026: 13 · 2025: 26 · 2024: 28 · 2023: 38 · 2022: 51 · 2021: 32 · 2020: 35 · 2019: 26 · 2018: 18 · 2017: 6 · 2016: 4 · 2015: 2
- FWCI
- 13.35
- Type
- conference paper
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Locations
- Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (publishedVersion)
- PubMed (publishedVersion)
- Europe PMC (PubMed Central) (acceptedVersion)
- PubMed Central (acceptedVersion)
- DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc-sa)
- Funders
- National Institute of General Medical Sciences (R01GM105018)
- Topics
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Emotion and Mood Recognition
- Health Sciences › Medicine › Cardiology and Cardiovascular Medicine › Heart Rate Variability and Autonomic Control
- Health Sciences › Medicine › Cardiology and Cardiovascular Medicine › ECG Monitoring and Analysis
- Keywords
- electrodermal activity, artifact detection, wearable devices, ambulatory measurement, machine learning classification, peak detection, web-based tool
- MeSH terms
- Machine Learning, Algorithms, Galvanic Skin Response, Humans, Signal Processing, Computer-Assisted
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.
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
