Skip to content

Latest commit

 

History

History
21 lines (11 loc) · 950 Bytes

README.md

File metadata and controls

21 lines (11 loc) · 950 Bytes

OpNet

opnet_problem statement

Project Goal

• Develop an object segmentation algorithm to automatically identify challenging anatomical structures to facilitate surgeons.

opnet_obj_segmentation

Solution

We developed OpNet, a convolutional neural network trained on images mined from YouTube to identify the borders of blood vessels and nerves.

opnet_schematic

Deliverables: • We presented this solution in the Sling Health Incubator 2021 Demo Day and provide the code in GitHub.

opnet_model_accuracy