Jocelyn Barker

Title: Jocelyn Barker and Surgical Video Analysis

Introduction

Jocelyn Barker is an inventor associated with San Jose, California, and Auris Health, Inc. Her patent work focuses on applying computer vision and machine learning to surgical video, including assessing surgical skill and tracking surgical tools. Barker is credited with 2 patents.

Latest Patents

Barker’s listed patents address two related challenges in analyzing surgical procedures through video.

“Video-Based Surgical Skill Assessment Using Tool Tracking” describes a method for classifying a surgeon’s technical skill based on detected tool motions. The system converts a sequence of tool movements into multi-channel feature representations, applies a one-dimensional convolution operation, and processes the resulting sequence with a transformer model. The model is trained to focus on the tool motions most relevant to the skill classification.

“Tracking Multiple Surgical Tools in a Surgical Video” describes a system for maintaining tracks of surgical tools across video frames. A deep-learning model detects objects in a current frame, and the system compares each detected object with established tracks using semantic and spatial similarity. These comparisons help assign objects to the appropriate tracks.

Career Highlights

Barker’s listed inventions bring together surgical video, tool-motion tracking, and machine-learning techniques. One patent applies these methods to skill assessment, while the other focuses on tracking multiple tools across video frames. Both aim to extract structured information from visual data recorded during surgery.

Collaborations

Barker’s listed coworkers include Mona Fathollahi Ghezelghieh and Mohammad Hasan Sarhan. Her work is associated with Auris Health, Inc.

Conclusion

Jocelyn Barker’s patent portfolio highlights the use of video analysis and machine learning to track surgical tools and assess surgical skill.