Monitoring of medication intake using a camera system.

Area of Science:

• Computer Vision
• Biomedical Engineering
• Healthcare Technology

Background:

• Home care services require reliable methods for monitoring patient adherence to medication regimens.
• Ensuring proper medication intake is crucial for patient health outcomes and reducing healthcare costs.
• Existing monitoring systems may lack the accuracy or adaptability needed for diverse home environments.

Purpose of the Study:

• To develop and evaluate a computer vision system for automated medication intake monitoring in home care.
• To enhance patient adherence and safety through non-intrusive technological support.
• To assess the system's performance in recognizing medication intake events with cooperative users.

Main Methods:

• Utilizing color and shape-based features for detecting body parts (skin, face, hands) and medication bottles.
• Employing color histograms, Hu moments, and edge detection for robust object tracking.
• Implementing Petri networks and event recognition algorithms for accurate identification of medication intake actions.

Main Results:

• The system demonstrates an accuracy exceeding 75% in monitoring medication intake.
• Successful detection of medication intake was achieved across various cooperative user scenarios.
• The color and shape-based approach proved effective in differentiating key objects like faces, hands, and medicine bottles.

Conclusions:

• The proposed computer vision system offers a promising solution for monitoring medication intake in home care settings.
• The method's accuracy and ability to function in cooperative scenarios highlight its potential for improving medication adherence.
• Further research could explore system adaptability for non-cooperative users and integration with broader telehealth platforms.