FDA-approved Perioperative AI

Database accompanying
Melvin RL, Broyles MG, Duggan EW, John S, Smith AD and Berkowitz DE (2022) Artificial Intelligence in Perioperative Medicine: A Proposed Common Language With Applications to FDA-Approved Devices. Front. Digit. Health 4:872675. doi: 10.3389/fdgth.2022.872675
If you use this information in a research article, please cite the above article.

Name of device or algorithmShort description in source database*CategoryReason for categoryDescription
Biovitals Analytics Engine“Cardiac Monitor”OpaqueCollapses everything to one indexUses patients’ own baseline to detect changes 
RhythmAnalytics“Monitoring cardiac arrhythmias”Opaque Features are Fast Fourier Transforms fed to deep learningDeep learning to classify rhtym 
Guardian Connect System“Predicting blood glucose changes”Opaque Not giving explanationsPredicts glucose levels outside of range and gives predictive alerts.
eMurmer ID“Heart murmur detection”TranslucentShows the systolic and diastolic phases (and sizes on a graph) but not reason for conclusionDetermines if murmurs are innocent or pathological
physIQ Heart Rhythm and Respiratory Module“Detection of atrial fibrillation”Translucent Collapses everything to one indexUses patients’ own baseline to detect changes using SMB
DreaMed“Managing Type 1 diabetes”Translucent “Patient reporting” and list of factors, but no weights or indicators shownRecommends insulin doses
ECG App“Detection of atrial fibrillation”TransparentCan point out multiple irregular tachograms and show themWatch-based AF detection
FibriCheck“Cardiac Monitor”TransparentShows prediction and annotated output for clinician smartphone AF detection
Irregular Rhythm Notification Feature“Detection of atrial fibrillation”TransparentIf->then rules for tachycardia and bradycardia Smartphone irregular rhythm notification
WAVE Clinical Platform“Monitoring vital signs”TransparentGives reasons. Rule based https://www.excel-medical.com/predictive-algorithmsRemote vital sign monitoring and alerts.
EchoMD Automated Ejection Fraction Software“Echocardiogram analysis”TransparentTells when device is in “right” place (transparent) and shows “good” image. Explainable calculations.Helps place echo device. Calculates ejection fraction.
Caption Guidance“Software to assist medical professionals in the acquisition of cardiac ultrasound images.”TransparentSee EchoMD — integrated with CaptionGoes with EchoMD above
*Source database for FDA-approved AI algorithms is The Medical Futurist. See also Benjamens, S., Dhunnoo, P. & Meskó, B. The state of artificial intelligence-based FDA-approved medical devices and algorithms: an online database. npj Digit. Med. 3, 118 (2020). https://doi.org/10.1038/s41746-020-00324-0.