Smart Microfluidics: Synergy of Machine Learning and Microfluidics in the Development of Medical Diagnostics for Chronic and Emerging Infectious Diseases

The COVID-19 pandemic, emerging/re-emerging infections as well as other non-communicable chronic diseases, highlight the necessity of smart microfluidic point-of-care diagnostic (POC) devices and systems in developing nations as risk factors for infections, severe disease manifestations and poor clinical outcomes are highly represented in these countries. These POC devices are also becoming vital as analytical procedures executable outside of conventional laboratory settings are seen as the future of healthcare delivery. Microfluidics have grown into a revolutionary system to miniaturize chemical and biological experimentation, including disease detection and diagnosis utilizing μPads/paper-based microfluidic devices, polymer-based microfluidic devices and 3-dimensional printed microfluidic devices. Through the development of droplet digital PCR, single-cell RNA sequencing, and next-generation sequencing, microfluidics in their analogous forms have been the leading contributor to the technical advancements in medicine. Microfluidics and machine-learning-based algorithms complement each other with the possibility of scientific exploration, induced by the framework's robustness, as preliminary studies have documented significant achievements in biomedicine, such as sorting, microencapsulation, and automated detection. Despite these milestones and potential applications, the complexity of microfluidic system design, fabrication, and operation has prevented widespread adoption. As previous studies focused on microfluidic devices that can handle molecular diagnostic procedures, researchers must integrate these components with other microsystem processes like data acquisition, data processing, power supply, fluid control, and sample pretreatment to overcome the barriers to smart microfluidic commercialization.