AI-based point-of-care device for total bilirubin quantification

• Published in: Research on biomedical engineering Vol. 41; no. 3
• Main Authors: Benedetti, Rafael, Gevaerd, Ava, Carvalho, Henrique Cunha, Figueredo, Marcus Vinicius Mazega, Barros, Frieda Saicla
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.09.2025
• Subjects: Biomaterials; Biomedical Engineering and Bioengineering; Biomedical Engineering/Biotechnology; Engineering; Original Article; Jaundice; AI; Bilirubin; Point-of-care
• ISSN: 2446-4732; 2446-4740
• DOI: 10.1007/s42600-025-00425-3

Purpose This work aimed to create a medical device for quantifying total bilirubin in human samples, using artificial intelligence as a diagnostic tool. Methods The device uses the principle of chromatography for sample flow, enabling the assessment of bilirubin concentration in human serum samples from both adults and neonates. With the integration of artificial intelligence, the device reading and result characterization were achieved through a dedicated reading instrument. The device’s performance was evaluated using the CLSI EP15-A3 as a testing metric to assess the results’ accuracy. Controls and samples were used to enable the clinical tests and the development of AI standardization. Results Following performance studies, the device demonstrated linearity from 0.7 to 20 mg/dL of total bilirubin. Precision studies indicated an intra-run variation of 7.6% and an intra-laboratory variation of 6.9%. Clinical performance data established a Pearson correlation (r) of 0.996 and a t -test of 0.8 (significant value, p  > 0.05). Stability studies concluded that the product is valid for 12 months post-manufacture, achieving the established performance with rapid and accurate results in under 10 min. Conclusion The development of this point-of-care device for bilirubin detection brings a practical and innovative solution. With reliable performance comparable to the biochemical reference method, the device integrates advanced technologies while requiring fewer resources and less training. This approach expands healthcare access, especially in remote areas, promoting efficient and accessible diagnostics through portable technology.