Intracavity surgical instrument action classification method based on kinematics appearance fusion and related device

By fusing kinematic and appearance features into intracavitary surgical videos and combining them with temporal smoothing using a hidden Markov model, accurate classification and stable identification of instrument movements are achieved. This supports graded risk warnings and surgical operation log recording, solving the problems of unstable identification and insufficient evaluation in existing technologies.

CN122265343AActive Publication Date: 2026-06-23ZHUHAI HENGQIN ALL-STAR MEDICAL TECHNOLOGY CO LTD
3 Cites 0 Cited by

Patent Information

Application Number
CN202610740926.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-05-27
Publication Date
2026-06-23
Estimated Expiration
2046-05-27

AI Technical Summary

Technical Problem

Existing methods for instrument motion recognition in intracavitary surgical videos suffer from problems such as insufficient single-modal information, easy confusion due to similar motion patterns, instability between frames, unstable classification output, and limited time-series smoothing capabilities, which cannot support downstream risk assessment.

Method used

A kinematic appearance fusion-based approach is adopted. By acquiring intracavitary video frame sequences and instrument tip trajectory data, and combining instance segmentation, kinematic feature extraction, appearance feature fusion, multi-class confidence prediction, and hidden Markov model temporal smoothing, the probability distribution of instrument action categories is output, triggering graded risk warnings and surgical operation log recording.

Benefits of technology

It achieves accurate classification of instrument movement types, solves the confusion caused by similar movement patterns, improves inter-frame stability, and outputs category probability distributions to support downstream risk assessment, thereby enhancing the safety and recording accuracy of surgical procedures.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides an intracavity surgical instrument action classification method based on kinematics appearance fusion and a related device. The method comprises the following steps: S1, acquiring an intracavity operation video frame sequence and an instrument tip movement track; s2, performing appearance feature extraction on the intracavity operation video frame sequence to obtain an appearance feature vector; s3, performing kinematics feature extraction on the movement track of the tip of the instrument to obtain a kinematics feature vector; s4, performing feature level fusion on the appearance feature vector and the kinematics feature vector to obtain a fused feature vector; s5, performing classification prediction based on the fusion feature vector to obtain action category probability distribution; and S6, performing time sequence smoothing on the action category probability distribution based on a hidden Markov model, and outputting a final instrument action category and confidence coefficient. The invention further provides a related device corresponding to the method. The related device comprises a device, electronic equipment, a computer readable storage medium and a computer program product.
Need to check novelty before this filing date? Find Prior Art