OC-035 AUTOMATION OF HERNIA VOLUMES - SAC DETECTION AND HERNIA VOLUMETRY USING COMPUTER VISION AND DEEP LEARNING - PUSHING FRONTIERS

Abstract Aim To identify the sac and defect using deep learning algorithms to compute hernial volume in an automated fashion. Materials methods 70 ventral hernia CT scans were used. Slices with annotated computer vision tools. 80% of data was used for training validation. 20% testing. We a combination 3D reconstruction train models. The model output compared manual annotation by experts. Results testing showed that our novel algorithm able tp estimate accuracy 90%. mean IOU detection 88% estimation 85%. Conclusion A supervised volumetry will increase volumetric indices shorten time. In most parts world, is not calculated radiologist surgeon left it rudimentary means such as visual estimation. Automated play major role standardising reports. believe this pivotal step creating actionable radiology reports help assess complexity plan surgery.