One paper selected for oral presentation at MICCAI 2017

We are pleased to announce that ICube will have some of its research featured as oral presentation at the international MICCAI conference on Medical Image Computing and Computer Assisted Interventions, held in Quebec City, Canada, in September 2017. The presentation will show some recent work from the research group CAMMA (http://camma.u-strasbg.fr/) demonstrating how Deep Learning can be used to automatically predict in real-time the remaining duration of a laparoscopic surgery using solely the laparoscopic video as input.

The proposed approach, based on recurrent neural networks, has been demonstrated on a large dataset of 120 Cholecystectomy procedures generated in collaboration with the IRCAD and IHU Strasbourg. It is shown to be particularly beneficial when surgery duration deviates from the average. Over a large number of surgeries, such assistance methods that monitor the surgical workflow have the potential to improve patient safety as well as to significantly reduce the clinical operative costs.

MICCAI 2017 has received 800 full paper submissions from which 255 papers were accepted based on a double blind review process. Only 29 of these papers have been selected for oral presentation.

The group CAMMA is part of the Control, Vision and Robotics team of the ICube laboratory.

Reference:
I. Aksamentov, A.P. Twinanda, D. Mutter, J. Marescaux, N. Padoy, Deep Neural Networks Predict Remaining Surgery Duration from Cholecystectomy Videos, Proceedings of Medical Image Computing and Computer-Assisted Intervention (MICCAI), to appear, 2017

Screenshot of the SurgFlow prototype for online and offline automatic surgical video analysis
©CAMMA, ICube laboratory

A video illustrating the prototype is available here:  LINK
Title : "Demonstration of the SurgFlow prototype for online and offline automatic surgical video analysis"