Dr. Ankur Gupta, Associate Professor in the Department of Mechanical Engineering at IIT Jodhpur and one of the study’s authors said that smartphones offer seamless integration with various technologies and platforms. The ability to connect the smartphone-based spot detection framework to a broader network or database can facilitate remote monitoring, data storage, and result sharing. This connectivity could be crucial for healthcare professionals or researchers.
One significant limitation of PADs is their reliance on specific light conditions for operation. However, the system devised by the researchers overcomes this hurdle entirely, enabling the PAD to function and transmit information to smartphones under almost any lighting circumstance.
The researchers utilized artificial glucose samples and processed various images of these colored samples using a machine learning application to create the smartphone app. This ensured that the color intensity detected by the PAD remained unaffected by varying light conditions and smartphone camera types. Consequently, the PAD can seamlessly connect to smartphones equipped with diverse camera optics.
Dr. Ankur Gupta emphasizes the extensive potential of this research by saying that this study showcases that the developed system is capable of conducting initial disease screening at the user’s end. By incorporating machine learning techniques, the platform can deliver reliable and accurate results, thus facilitating the evaluation of result accuracy for enhanced initial healthcare screening and disease diagnosis. This module has the potential to be adaptable in detecting various diseases through the provision of sample data for training and testing purposes.
In a forward-looking application, the team is currently developing a method for simultaneously detecting glucose, uric acid, and lactate using different color indicators as distinct codes in non-blood samples. Although the current focus is on glucose detection, this approach holds promise for screening and diagnosing other diseases.
While the fundamental concept remains unchanged, adjustments are necessary to accommodate different target analytes, enzymes, and indicators. Subsequently captured images can undergo additional training for machine learning analysis, enabling diagnostic information to be directly accessible on the user’s smartphone.