Pilot study: a low-cost, point-of-care microfluidic chip for smartphone-integrated colorimetric detection using convolutional neural networks
Abstract Point-of-care testing is essential for individuals with various diseases, as it allows for timely and effective management, preventing complications and promoting overall health. Colorimetric detection is a valuable tool for disease diagnosis due to its simplicity, affordability, and abilit...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-07-01
|
| Series: | Discover Applied Sciences |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s42452-025-06859-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Point-of-care testing is essential for individuals with various diseases, as it allows for timely and effective management, preventing complications and promoting overall health. Colorimetric detection is a valuable tool for disease diagnosis due to its simplicity, affordability, and ability to provide rapid, qualitative results at the point of care. In this endeavour, we have designed a novel, affordable point-of-care diagnostic device utilizing microfluidic principles, a smartphone camera, and established laboratory colorimetric methods for accurate colorimetry. Our proposed microfluidic device comprises of layers of adhesive poly-vinyl films with micro-wells precision-cut using a cutting printer and stacked on a glass slide. As part of this pilot study, we employed the glucose-oxidase/peroxidase reaction on the microfluidic platform as a demonstration, successfully achieving enzymatic glucose determination. The resulting coloured complex, formed by phenol and 4-aminoantipyrine in the presence of hydrogen peroxide generated during glucose oxidation, is captured at various glucose concentrations using a smartphone camera. Raw images are processed and utilized as input data for a 2-D convolutional neural network (CNN) deep learning classifier, demonstrating an impressive 97% overall accuracy against new images. The glucose predictions done by CNN are compared with ISO 15197:2013/2015 gold standard norms. Furthermore, the classifier exhibits outstanding precision, recall, and F1 score of 96.6%, 96.2%, and 96%, respectively, as validated through our study, showcasing its exceptional predictive capability. The developed CNN model can be successfully used as pre-trained model for future glucose concentration predictions. |
|---|---|
| ISSN: | 3004-9261 |