Back

A Review of Point-of-Care Devices for Blood-Testing Towards AI-driven Remote Digital Care, Precision Healthcare and Predictive Medicine

Gu, J.; Zenil, H.

2025-12-15 health informatics
10.64898/2025.12.13.25340658 medRxiv
Show abstract

Point-of-care (POC) blood testing enables rapid, decentralized diagnostics with transformative promise, yet its innovation landscape remains poorly mapped. To this end, we focused on features that we believe are key to make progress in areas of precision healthcare and predictive medicine, such as longitudinal data collection and data analytics integration. While no review can be complete, this work attempts to address this gap by analyzing 86 POC blood testing devices worldwide and proposing a unified framework to compare them across technology principles, diagnostic breadth, usability, regulatory pathway, deployment feasibility (via a custom index), and data/AI integration. Electrochemical biosensors were the single largest platform (29.1%), strongly associated with glucose testing ({chi}2=237.8, p<0.001), while spectroscopic and microfluidic systems remained niche due to higher costs and specialized requirements. Regulatory approval skewed toward moderate risk (44.2% FDA II; 27.4% IVDR C), while approval times lengthened with risk class (e.g., IVDR D {approx}540 days). A trade-off was observed between usability and panel breadth: tools for home or low-resource settings emphasize simplicity and affordability, whereas clinical systems expand diagnostic range at higher complexity and cost. Deployment feasibility scores favored handhelds, while benchtops were penalized by workflow and capital demands, and microfluidics by consumables. Innovation clusters in North America, Europe, and East Asia reinforce global leadership and disparities.

Published in Computer Methods and Programs in Biomedicine Update · not in our set (fewer than 10 published preprints to learn from) · training set

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.