Medical Video Highlight Extraction for Faster Selective Uploads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Long medical videos captured during surgeries consume significant time and memory resources for uploading to servers, making it inefficient to upload only necessary scenes.
Innovation Solution
An information processing system that acquires medical videos, sets highlight scenes for priority upload, generates a video screen with images arranged chronologically, and allows user-driven scene changes for efficient upload to storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the entire long medical video is uploaded to the server, then all video content is preserved, but upload time and memory consumption increase significantly
Solution Approach 1:
The system extracts and identifies highlight scenes from the long medical video based on scene change detection and importance criteria. Only these extracted highlight portions are uploaded to the server, while the rest of the video remains local. This extraction principle directly resolves the contradiction by preserving only the most important video content for upload, significantly reducing upload time while maintaining reliability for critical medical information.
Solution Approach 2:
The medical video is segmented into multiple scenes based on scene change detection algorithms. The system divides the continuous video stream into discrete scene segments, evaluates each segment's importance, and selectively uploads only the highlight segments. This segmentation approach allows the system to maintain complete video content locally for reference while uploading only essential portions to the server, thereby reducing upload time without compromising content preservation.
2Reliability
If the entire long medical video is uploaded to the server, then all video content is preserved, but memory capacity of the storage is consumed
Solution Approach 1:
The system extracts highlight scenes from the complete medical video based on scene change detection and importance evaluation. Only these extracted highlight portions are transferred to and stored on the server, while the full video remains in local storage. This extraction principle directly addresses the memory capacity issue by eliminating the need to upload and store redundant non-highlight video content on the server, thereby preserving memory resources while maintaining reliability through local retention of the complete video.
3Productivity
If scene change detection is performed to identify highlight scenes, then upload efficiency is improved, but processing complexity increases
Solution Approach 1:
The system implements self-service through automated scene change detection and highlight scene identification algorithms that operate without manual intervention. The processing unit automatically analyzes the medical video, detects scene changes based on visual or audio cues, identifies highlight portions according to predetermined criteria, and prepares them for upload. This automation resolves the contradiction by improving upload efficiency through intelligent processing while managing complexity through algorithmic self-service rather than requiring complex manual workflows or additional hardware systems.
Data Source
AI summary
The present technology relates to an information processing system, an information processing method, and a program that enable only a necessary scene of a medical video to be easily uploaded to a server device.A medical video captured by a medical image capturing device is acquired, a highlight scene that is a candidate to be preferentially uploaded to a storage is set on the basis of the medical video, a video of a first screen representing a range of images included in the highlight scene is generated on an arrangement screen in which images of frames at regular intervals in the medical video or images representative for each scene are arranged in chronological order, and the highlight scene is changed on the basis of a user operation.