Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{179484,
        author = {Darshan B and Malatesh KR and Shashikanta Modak and Dr. Sreelatha R},
        title = {Patient Activity Analysis from Image and Video using YOLOv8 in Healthcare Settings},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {11},
        number = {12},
        pages = {7580-7591},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=179484},
        abstract = {This paper presents a computer vision-based system for patient activity analysis in healthcare environments that utilises the YOLOv8 object-detection model. The proposed system analyzes static images and pre-recorded video feeds to detect key patient states and events, such as ’Fallen Patient, Seizure, and ’Patient Sleeping, by classifying visual data with a fine-tuned YOLOv8n model. The methodology integrates a comprehensive pipeline encompassing data input (image/video), frame preprocessing, YOLOv8 inference, event classification, and subsequent visualisation. YOLOv8 was trained on a custom- labelled dataset with 12 classes, derived from publicly available hospital scene images (CC BY 4.0). Experimental results indicate that the system achieved approximately 0.80 for both precision and recall, an F1-score of approximately 0.80, and a mean Average Precision (mAP@0.5) of approximately 0.86 on the designated test data. The system exhibited processing speeds suitable for deployment on GPUs, CPUs, and edge devices. The architecture is designed to support deployment on resource- constrained hardware (leveraging the yolov8n.pt model variant) and is capable of producing annotated visual output. These findings demonstrate the feasibility of employing YOLOv8 for noninvasive patient activity analysis from images and videos, offering a robust approach to enhance patient safety assessment and situational awareness in ICUs, eldercare facilities, and home- based healthcare contexts.},
        keywords = {Computer Vision, Patient Monitoring, YOLOv8, Object Detection, Healthcare, Fall Detection, Activity Recognition, Ethical AI, Tele-ICU, Eldercare, Deep Learning},
        month = {May},
        }