OC-061 IMAGE-BASED DEEP LEARNING MODELS (DLMS) IDENTIFY ABDOMINAL WALL RECONSTRUCTION (AWR) PATIENTS RECEIVING PREOPERATIVE BOTULINUM TOXIN A (BTA) WHO REQUIRE COMPONENT SEPARATION TECHNIQUE (CST)

Abstract Aim Preoperative BTA facilitates muscle/fascia elongation and fascial closure, which increases the durability of hernia repair may reduce requirement for CST. The aim this study was to develop image-based DLMs that predict whether patient will require CST if injected with prior surgery. Methods An institutional database used identify AWR patients who received preoperative had CT imaging. Axial cuts hernias were rendered train a DLM. primary outcome ROC predicting DLM tested on scans from pre-identified subset underwent did not receive BTA. Results There 116 met criteria (4,580 images). Of these patients, 69 (59.5%) required (2884 images); 47 (40.5%) undergo (1,696 0.78 (Figure 1); accuracy, sensitivity, specificity 0.79, 0.86, 0.68, respectively. 98 in test set; 57 (58.1%) predicted 41 (41.9%) not. For an M1 component, 21.1% be spared versus 46.8% those M2-M5 component (p=0.04). Figure 1 Model Performance Prediction Conclusions Image-based accurately receiving When applied set, very likely CST, is consistent previously reported data, further validating model.