Real-time And Accurate Estimation Of Surgical Hemoglobin Loss Using Deep Learning-based Medical Sponges Image Analysis

This study aimed to explore a novel surgical Hb loss estimation method using deep learning-based medical sponges image analysis. Whole blood samples were collected from the department of laboratory medicine, Renmin Hospital of Wuhan University, and patient Hb measurements were recorded. The Hb concentration was selected, ranging from 50 to 170 g/L, and the images were randomly divided 7:3 into training and testing groups. The training group was used to develop the predictive model, and the testing group was used to evaluate the model performance.