Hospital Entity Tracking With Articulated Pose Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing clinical workflows in medical facilities face challenges in automating the analysis of complex and cluttered video feeds due to the limitations of conventional image processing techniques, and deep learning algorithms like YOLO struggle to provide semantic understanding of events in hospital environments.
Innovation Solution
The use of articulated models with keypoints and affinity fields, combined with deep neural networks and graph-fitting processes, allows for accurate modeling and tracking of entities in medical facilities, enabling semantic understanding and automation of workflows.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual monitoring methods are used, then device complexity is low, but measurement precision and reliability of monitoring are insufficient
Solution Approach 1:
The patent introduces an intermediary processing system that receives data from multiple sources (sensors, medical devices, wearables), standardizes it, and presents unified monitoring information to clinicians. This intermediary layer enables precise multi-parameter monitoring without requiring clinicians to directly manage complex data integration, thus improving measurement precision while managing device complexity.
Solution Approach 2:
The monitoring system is designed to universally handle multiple types of medical parameters (vital signs, lab results, imaging data) from various sources through a single integrated platform. This multi-functional approach allows precise monitoring of diverse parameters without requiring separate specialized systems for each parameter type.
2Reliability
If comprehensive monitoring of multiple parameters is implemented, then reliability of monitoring is improved, but device complexity increases
Solution Approach 1:
The patent segments the monitoring system into distinct functional modules: data acquisition from multiple sources, data standardization and integration, analysis and interpretation, and presentation to clinicians. Each module handles specific tasks independently, enabling reliable comprehensive monitoring while managing complexity through modular architecture that allows independent development and maintenance of each component.
Solution Approach 2:
An intermediary processing layer standardizes and integrates data from diverse sources before presenting unified information to clinicians. This mediator handles the complexity of multi-parameter integration, enabling reliable comprehensive monitoring without exposing clinicians to system complexity.
3Productivity
If rapid data processing is implemented, then productivity is improved, but use of energy increases
Solution Approach 1:
The system implements periodic processing cycles where data is collected, processed, and updated at optimized intervals rather than continuously. This periodic action maintains high productivity by providing timely updates while reducing energy consumption by avoiding constant high-power processing operations.
Solution Approach 2:
The monitoring system automatically processes and prioritizes data without requiring continuous manual intervention. Critical parameters are processed with higher priority and frequency, while non-critical parameters use lower processing power, enabling the system to self-manage energy allocation based on clinical needs.
Data Source
Figure 1
Figure 2
Figure 3a~3c
AI summary
A computer implemented method for use in monitoring a first entity in a medical facility comprises: i) obtaining an image of the medical facility, ii) using a machine learning process to fit a first articulated model to the first entity in the image, wherein the first articulated model comprises keypoints corresponding to joints and affinity fields that indicate links between the keypoints. The method further comprises iii) determining a location or posture of the first entity in the medical facility from relative locations of fitted keypoints of the first articulated model in the image.