"Perfusion Assessment of Healthy and Injured Hands Using Video-Based Deep Learning Models".

Background: Assessing in-field hand trauma is challenging, and inaccurate perfusion assessment can substantially impact the patient and health system. Technology that enhances perfusion assessment could improve in-field triage. We present non-contact, video-based deep learning methods to classify perfused and ischemic fingers in control and acute trauma settings.

Methods: We obtained iPhone video from two cohorts of subjects. The first group were control participants, some of whom were evaluated during cycles of tourniquet-induced ischemia. The second group were acutely injured patients in our emergency department(ED). For both groups, imaging photoplethysmography (iPPG) waveforms were extracted using a deep learning model, after which the waveform's spectrogram was classified as either perfused or ischemic using a ResNet-18 classifier. This was then compared to clinical ground-truth labels.

Results: We captured videos of 48 controls including 14 evaluated during tourniquet-induced ischemia, and 15 acutely injured patients. Over five-fold cross-validation of control subjects, our algorithms correctly classified ischemic finger regions with a sensitivity of 72%, a positive predictive value (PPV) of 74%, and an accuracy of 90%. We then tested on videos of acutely injured patients, without controlling hand pose, skin cleanliness, or other variables, and achieved a sensitivity of 33%, a PPV of 24%, and an accuracy of 77%.

Conclusions: Under controlled settings, deep learning methods for perfusion classification performed well. In hospital settings - with uncontrolled lighting, hand pose, and injuries - classification performance degraded. This technology is promising but additional approaches that account for acute trauma-related variables are needed for clinical applicability as a triage tool.