Where artificial intelligence stands in the development of electrochemical sensors for healthcare applications-A review

Abstract

The electrochemical sensor (E-sensors) market trends have identified the biomedical applications as a significant market growth with impact on personalized therapy. Given the wide variability among individuals, a key point is to acknowledge that the assays in biological samples are still limited to laboratory setup. While slight changes in the raw experimental data are beyond human capability to process, some issues related to the design of sensors identification, matrix interference, and prediction tasks can be assisted by AI tools. However, the data delivered by E-sensors for Machine Learning (ML) is not common in literature, but since the measurements can be done in real time and can identify trends and patterns, while keeping human-driven decisions in the loop, this topic is invaluable. In this work, a critical analysis of the AI-assisted sensors was performed regarding the specific tasks that can be solved by AI tools. The data flow from the design of the concept to the final results was presented related to the elaboration of E-sensors. Additionally, wearable sensors designed for biomedical applications were critically reviewed from the perspective of AI highlighting the limitations on this topic and what does the “promising” statement mean in this context.