AI system
- System
-
Corsano CardioWatch 287-2 with the DETECT photoplethysmography cardiac-arrest detection and accelerometry fall-detection algorithms
- Model
-
DETECT, a rule-based algorithm that detects absent pulsations.
- Release
-
Exact release not established.
- Application developer
-
DETECT program, Radboud University Medical Center, with Corsano Health, The Hague (device and algorithm processing)
- Role
-
Not assessed
- Configuration
-
Researchers applied DETECT after the event to data from a CardioWatch 287-2. The algorithm did not run live or send an alert.
- Inputs and outputs
-
Not assessed
- Autonomy
-
Not assessed
- Workflow
-
Continuous wristband wear with 32 Hz photoplethysmography and accelerometry uploaded to a study cloud. A shockable arrest with clear loss of pulsations during cycling, in a patient whose ICD provided the reference and the rescue. Blinded after-the-fact processing by the developers' data scientist.
- Human oversight
-
Not assessed
- Relation to AI
-
The algorithm was run afterwards on exported data by a blinded Corsano data scientist. It correctly flagged the loss of pulse and the fall. It was not running live and sent no alert. The ICD shock is what restored circulation. The developers' own papers are the only full sources. They describe a rule-based detector of absent pulsations by signal amplitude and waveform morphology rather than a machine-learning model.
- AI contribution
-
The algorithm was run afterwards on exported data by a blinded Corsano data scientist. It correctly flagged the loss of pulse and the fall. It was not running live and sent no alert. The ICD shock is what restored circulation. The developers' own papers are the only full sources. They describe a rule-based detector of absent pulsations by signal amplitude and waveform morphology rather than a machine-learning model.
Benefits
- Outcome identifiers
-
Not assessed
- Benefit
-
detected cardiac arrest at 09:50:17 after loss of pulse
- Category
-
Improved Task Accuracy
- Status
-
Not assessed
- Beneficiary groups
-
People. Society.
- Beneficiaries
-
Not assessed
- Reach
-
Not assessed
- Distribution and access
-
Not assessed
- Duration
-
Not assessed
- Expected?
-
Yes
- Costs and harms
-
Not assessed
Measurement
- Studies
-
Not assessed
- Study design
-
Not assessed
- Measurement
-
Not assessed
- Setting and sample
-
Not assessed
- Measures
-
detected cardiac arrest at 09:50:17 after loss of pulse
- Comparison
-
the ICD's electrograms and the confirmed time of collapse (velocity 0 km/h), reviewed in post-processing
- Results
-
Result: Cardiac arrest was detected at 09:50:17, after a loss of pulse in the photoplethysmography signal.
- Additional value from AI
-
Not assessed
- Limitations
-
Not assessed
- Denominators
-
Not assessed
- Materials for replication
-
Not assessed
Money
- Kind of value
-
Not assessed
- Amount
-
Not assessed
- Currency and prices
-
Not assessed
- Period and unit
-
Not assessed
- Realization
-
Not assessed
- Recipients and payers
-
Not assessed
- Gross or net
-
Not assessed
- Included costs
-
Not assessed
- Comparison
-
Not assessed
- Method
-
Not assessed
- Return and payback
-
Not assessed
People and places
- Participants
-
Developer: DETECT program, Radboud University Medical Center, with Corsano Health, The Hague (device and algorithm processing). Operator: Radboud University Medical Center (DETECT-3 study).
- Jurisdictions
-
Benefit:
Netherlands (DETECT-3 study, Nijmegen; the cycling incident's location is not named). Developer: Netherlands. Operator: Netherlands. - Ownership and control
-
Not assessed
- Dependencies
-
Not assessed
- Cross-border connections
-
Not assessed
- Geopolitical implications
-
Not assessed
Follow-up
- Conditions for replication
-
Adequate photoplethysmography quality during motion (the authors note it is typically disturbed by motion). A live implementation with acceptable false-positive rates, which DETECT-3 is still measuring. An alerting chain to emergency services, which does not yet exist for this device.
- Replication attempts
-
2 follow-up entries. Benefit recurred: 2.
- Later findings
-
24 July 2026. The same group's later prospective validation in ICU patients with non-shockable arrest after withdrawal of life-sustaining treatment reported 100% sensitivity in the 23-patient test set with one false positive, and DETECT-1b (April 2026) externally validated the algorithm in 50 induced shockable arrests; no live alerting has been reported.
- Availability
-
Not assessed
- Institutional decision
-
Not assessed
- Grounds
-
Not assessed
- Action and result
-
Not assessed
- Success criterion
-
Not assessed
Evidence
- Source numbers
-
AIBD-0298-source-0001–0006
- Citations
-
Numbered references below.
- Language
-
Not assessed
- Source relationships
-
Developer, Independent
- Original evidence
-
Not assessed
- Access and versions
-
Not assessed
- Supported findings
-
Not assessed
- Evidence links
-
Not assessed
- Evidence strength
-
Not assessed
- Missing evidence
-
Not assessed
Grades
- Assessed benefits
-
People. Society.
- Depth
-
Not assessed
- Reasons
-
People: The patient was saved by his ICD. The wristband's detection was retrospective and produced no alert or treatment. Society: The authors argue a live system could act as a first witness for unwitnessed arrests. No deployment outcome exists yet.
- Observed or intended
-
Not assessed
- Rubric
-
Not assessed
- Assessor and date
-
Not assessed
- Alternative assessments
-
Not assessed
- Fill
-
The makers' claim
Entry details
- Kind
-
Not assessed
- Related entries
-
Not assessed
- Field
-
Emergency medicine / resuscitation. Research wristband for automated cardiac-arrest detection.
- Purpose
-
Built to detect out-of-hospital cardiac arrest from wrist photoplethysmography and accelerometry and, in future, alert emergency services. The retrospective detection demonstrates the first half of that purpose.
- Dates
-
Date basis unassessed: 2026 (precision unassessed). Updated: 21 October 2026.
- Stage
-
Not assessed
- Version
-
SHA-256: 0f34af61a9a1…
References
- Other identifiers
-
Not assessed
- Category mappings
-
Not assessed
- Brief
-
Not assessed
- Revisions
-
5 revisions. Latest: 21 October 2026.
Sources
-
Peer-reviewed study Developer
Edgar R, Jansen CE, Pol LR, Ebrahimkheil K, van Kaam RC, Ronner E, Brouwer MA, Beukema RJ, Cetinyurek-Yavuz A, Stas PC, Boersma E, Hoedemaekers CWE, van Royen N, Bonnes JL (2026). Automated cardiac arrest detection using wrist-derived photoplethysmography during withdrawal of life-sustaining treatment: a prospective clinical validation study. The Lancet Regional Health - Europe 67:101791 (PMC13427576). Corsano Health co-authors.
The previously developed rule-based algorithm was refined using two training cohorts and evaluated in a separate test cohort.
-
Peer-reviewed study Developer
Edgar R, Scholte NTB, Ebrahimkheil K, et al. (2026). Automated Cardiac Arrest Detection Using Wrist-Worn Photoplethysmography: External Validation in Patients With Induced Shockable Cardiac Arrest (DETECT-1b). Circulation: Arrhythmia and Electrophysiology 19(5):e014708 (PMC13182813). Corsano Health co-authors.
In the DETECT-1a study, we developed a wrist-derived photoplethysmography (PPG) algorithm for cardiac arrest detection using patient data from induced short‑lasting circulatory arrests, achieving 98% sensitivity.
-
Case report Developer
Edgar R, Jansen CE, Pol LR, Pisters R, van Royen N, Bonnes JL (2026). Out-of-hospital cardiac arrest detection by a wearable: the first real-life case. Resuscitation Plus 28:101245 (PMC12914388). Authors are the DETECT algorithm's developers.
The cardiac arrest detection algorithm of the CardioWatch correctly recognized the cardiac arrest event and subsequent collapse based on photoplethysmography and accelerometry data during post-processing.
While cycling, the patient experienced a ventricular fibrillation cardiac arrest that was successfully terminated by the ICD.
A data scientist at Corsano Health, who was blinded to the cardiac arrest event, ran the data through the photoplethysmography-based cardiac arrest detection algorithm.
-
Other Independent author
Veigl C, Billig S (2026). A digital witness for cardiac arrest: are wearables ready for real-world use? Resuscitation Plus 28:101246 (editorial accompanying the two case reports; PMC12914390).