Early Prediction of Hemodynamic Shock in the Intensive Care Units with Deep Learning on Thermal Videos: A Retrospective Longitudinal Study (Preprint)

Abstract

BACKGROUND Shock is one of the major killers in Intensive Care Units and early interventions can potentially reverse it. In this study, we advance a non-contact thermal imaging modality for continuous monitoring and prediction of hemodynamic shock in advance. OBJECTIVE We aim to monitor and predict the advent of hemodynamic shock 6 hours in advance using an automated non-contact thermal imaging decision pipeline. METHODS Thermal Videos were captured in a Pediatric ICU-setting along with vitals time-series data. Deep-learning-based body-part segmentation models were trained to extract the Center-to-Peripheral temperature value difference from the videos. Extracted time-series data along with heart rate was finally analyzed using Long-Short Term Memory models to predict the shock status up to the next 6 hours. RESULTS 103,936 frames from 406 non-contact thermal videos were recorded longitudinally upon 22 patients. Our models were able to predict the shock well till 6 hours of lead time using thermal information and achieved the best area under the receiver operating characteristics curve of 0.81±0.06 and area under the precision-recall curve of 0.78±0.05 at 5 hours, providing sufficient time to stabilize the patient. CONCLUSIONS Our approach leverages thermal imaging as a non-invasive and non-contact modality to continuously monitor hemodynamic shock, and thus, provides a reliable shock prediction using an automated decision pipeline that can provide better care and save lives. CLINICALTRIAL None

Heart Rate Assessment in a Pediatric ICU with Non-Contact Infrared Thermography and Machine Learning

“…Our current study includes 124 recording sessions from 65 UHID patients. The details of the data procurement procedure are described in the Supplementary methods S1 Section in Vanshika et al [5] supplementary files.…”

“…The thermal camera is placed at an appropriate distance from the patient in order to avoid any direct contact or disturbance during routine check-ups as stated in Vanshika et al [5]. The camera orientation is done in a way that captures the full body of the infant in a standard color-scaling.…”

“…Thus, a non-invasive, contactless method that is able to monitor vitals like heart rate is highly recommendable. These technologies are gaining momentum for their rise in application in many complex diseases such as breast cancer detection, hemodynamic shock, and skin disorders [4,5].…”