Skip to content

Machine Learning Challenge

Are You up for a Machine Learning Challenge This Summer?

Surgical Science is hosting a endoscopic vision challenge:  Synthetic Data for Instrument Segmentation in Surgery, as part of the 2023 MICCAI meeting.

Challenge Description:

Participants will build machine learning models to identify which pixels in an image belong to the robotic instruments. This application of AI is called “image segmentation”.

Surgical Science is generating synthetic data from our FlexVR robotic surgery simulator. The competition participants will use the data to train their machine learning models to solve this instrument segmentation problem. The machine learning models will be submitted to Surgical Science for evaluation. Surgical Science will use an additional “test” data set that would be kept private to evaluate the performance of the participant’s model.

The evaluation results will be presented at MICCAI 2023 (26th International Conference on Medical Image Computing and Computer Assisted Intervention), October 8th, 2023.

Synthetic data will be critical in the surgical intervention’s domain, as medical devices across the anatomy spectrum are becoming AI-powered. The Surgical Science line of simulators presents a great opportunity to leverage state-of-the-art graphics and physics simulation to generate automated labeled data for building the next generation AI solutions for medical devices.

The synthetic data challenge at MICCAI 2023 is a step towards enabling this!

The challenge is LIVE.  Check it out now and register!

Skip to navigation