A Survey on Non-Invasive Computing: Neurological-Hematological Framework for Early Infection and Stroke Detection with Future Directions

Document Type

Conference Proceeding

Publication Date

2025

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Source Publication

IEEE 49th Annual Computers, Software, and Applications Conference (COMPSAC)

Source ISSN

2836-3795

Original Item ID

DOI: 10.1109/COMPSAC65507.2025.00110

Abstract

This paper introduces a novel, non-invasive tool for building a holistic neurological-hematological diagnostic framework that can be used for early detection of infection and stroke. We proposed and developed a multimodal sensing approach, where an ear canal–based acoustic system can be used to identify brain activities and a smartphone-based facial video analysis platform to estimate white blood cell (WBC) and hemoglobin (Hb) levels. Our ear-based EEG methodology achieved a remarkable 96% classification accuracy from the low-frequency level (>30hz), with regression models with mean R2 scores above 0.96 across all EEG bands. For hematological diagnostics, the system predicted WBC counts with a mean squared error (MSE) of 0.79 and Hb levels with a mean absolute percentage error (MAPE) of 8.24% using optimized support vector regression. These results demonstrate the feasibility of real-time, point-of-care physiological in-clinic, or remote monitoring without an invasive approach. Future work will focus on motion-artifact mitigation and large-scale validation. By bridging neurology and hematology through accessible AI-driven technologies, this work lays the foundation for next-generation mHealth-based non-invasive diagnostic tools.

Comments

Published as part of the proceedings of the IEEE 49th Annual Computers, Software, and Applications Conference (COMPSAC), 2025: 824-833. DOI.

Link to Full Text

Share

COinS