Early Prediction of Hemodynamic Shock in Pediatric Intensive Care Units With Deep Learning on Thermal Videos

Abstract

Shock is one of the major killers in intensive care units, and early interventions can potentially reverse it. In this study, we advance a noncontact thermal imaging modality for continuous monitoring of hemodynamic shock working on 1,03,936 frames from 406 videos recorded longitudinally upon 22 pediatric patients. Deep learning was used to preprocess and extract the Center-to-Peripheral Difference (CPD) in temperature values from the videos. This time-series data along with the heart rate was finally analyzed using Long-Short Term Memory models to predict the shock status up to the next 6 h. Our models achieved the best area under the receiver operating characteristic curve of 0.81 ± 0.06 and area under the precision-recall curve of 0.78 ± 0.05 at 5 h, providing sufficient time to stabilize the patient. Our approach, thus, provides a reliable shock prediction using an automated decision pipeline that can provide better care and save lives.

Monitorización de la perfusión tisular en el niño con disfunción circulatoria aguda: revisión narrativa

“…En pediatría, existe poca evidencia que sustente el rendimiento de la gradiente de temperatura. Sin embargo, con los recientes avances en tecnología de imágenes térmicas e inteligencia artificial se han obtenido resultados en donde la gradiente de temperatura puede ser una herramienta útil en la detección e incluso predicción del shock en pacientes pediátricos 28,29 . Lamentablemente, esta tecnología no se encuentra disponible para el ámbito clínico.…”

Uses of infrared thermography in acute illness: a systematic review

“…Some studies found IRT as a non-ionising form of imaging beneficial, particularly those investigating musculoskeletal injuries in paediatric patients (32)(33)(34)(35). The use of a non-contact, continuous imaging modality was beneficial to studies applying IRT for monitoring purposes (20,26,(36)(37)(38).…”

Table_1.DOCX

“…The majority of authors identified their studies as exploratory or assessing feasibility (n = 16) (11, 12, 14-16, 19, 21, 22, 26, 27, 31, 32, 34, 35, 37, 39). Few were described as validation or diagnostic accuracy studies (n = 8) (13,17,18,20,24,33,36,38).…”

“…Some studies found IRT as a non-ionising form of imaging beneficial, particularly those investigating musculoskeletal injuries in paediatric patients (32)(33)(34)(35). The use of a non-contact, continuous imaging modality was beneficial to studies applying IRT for monitoring purposes (20,26,(36)(37)(38).…”

“…Anti-inflammatory use, peripheral vascular disease and ice pack application were identified as potential confounders that excluded some recruits from study participation. Two studies acknowledged limitations cause by inadequate resolution of the IRT device used (37,38).…”