NIT Rourkela researchers develop AI-enabled autofocusing technology for microscopic imaging

Researchers from the Department of Biotechnology and Medical Engineering at National Institute of Technology Rourkela have developed an AI-enabled autofocusing technology aimed at improving microscopic imaging for biomedical diagnostic applications. The technology is designed to produce rapid, accurate, and repeatable results with minimal human intervention.

The technology was developed in collaboration with Glowvista Instruments, a startup incubated at NIT Rourkela’s incubator centre (FTBI). The research team has secured a patent titled, “A Method for Autofocusing in Optofluidic Microsystems and Processes” for the technology. The patent number is 589270 and the application number is 202431080016.

Microscopy technology is used in healthcare for diagnosing cells, tissues, microorganisms, and biological structures that cannot be viewed with the naked eye. It is also used in the detection of diseases such as cancer, pathology-related conditions, malaria, and tuberculosis. In research, microscopy supports drug development and point-of-care diagnosis through visual analysis.

Conventional microscopy systems depend on manual adjustments, which can be time-consuming and prone to human error. These limitations can result in inconsistent findings, inaccurate diagnoses, and treatment delays, particularly in complex biological samples and emergency diagnostic settings.

To address these limitations, the NIT Rourkela research team developed an optofluidic digital microscopy platform integrating deep learning technology with an optical imaging system and automated motion control. The system continuously analyses microscopic images in real time and automatically adjusts focus using an intelligent feedback mechanism.

The research team included Prof. Earu Banoth, Dr Shaik Ahmadsaidulu, Amol Lalchand Salve, and Padmanaban Selvakumar.

Developed at a cost of Rs 1.20 lakh, the system demonstrated accurate results in laboratory-scale testing for detecting Acute Lymphoblastic Leukemia (blood cancer), malaria, and Complete Blood Cells counts through blood cell classification using 5-class and 7-class categorisation.

Speaking about the technology, Prof. Banoth said, “Our target is to develop a simple handheld system that should work as effectively as imported automated microscopy technology with precise information. Further, the system should be extended for various applications, unlike Flow Cytometers and Imaging Flow Cytometers.”

The developed system includes AI-powered autofocus with real-time image processing, automated motion control for focus adjustment, imaging support for biological and micro-scale samples, cloud-enabled learning, and user-focused operation designed to improve repeatability and efficiency.

The system has been developed with research grants from the Anusandhan National Research Foundation, the Department of Science and Technology, and the Department of Biotechnology.

The technology has potential applications in biomedical diagnostics, disease detection, digital pathology, tissue imaging, AI-assisted microscopy, automated imaging, point-of-care healthcare devices, microfluidic analysis, biofluid monitoring, biomedical and life sciences research, laboratory automation, and portable diagnostic systems.

Speaking about the next phase of the project, Prof. Banoth said, “We are extending our work to develop complete ground truth data as well as scale up the work for deployment at various locations for field testing and to obtain feedback from diagnostic centres and research laboratories. The data will be used for further approvals towards developing a market-ready product, and we look forward to its market launch. At present, we are looking for funding support from both the research and startup sides to scale up the work.”

The development aligns with the Government of India’s Make in India initiative and reflects NIT Rourkela’s focus on indigenous research-driven healthcare technologies.