KOZHIKODE: What if a microscope could do more than magnify what lies beneath its lens? What if, as a medical professional scans a slide, artificial intelligence could quietly pick out suspicious cells, flag areas that deserve attention, and present the findings instantly on a screen?
That possibility is now taking shape at the National Institute of Technology Calicut, where researchers, working alongside medical experts, have developed an AI-enabled smart microscopy system designed to turn conventional microscopes into intelligent medical imaging platforms.
The technology is aimed at tackling one of the less visible challenges in pathology — the time, concentration, and visual effort required to examine microscopic samples manually. Instead of requiring doctors and pathologists to spend long periods peering through microscope eyepieces, the system captures images or video from the microscope and analyses them in real time, highlighting cells and other features of interest.
At the heart of the system is a compact combination of a digital camera, a lightweight AI processor and an LCD. The camera captures the microscopic field, while an AI model processes the images locally. The results are then displayed on the screen with highlighted regions of interest and visual annotations, allowing medical professionals to examine the material more conveniently.
One of the most significant features is that the system does not depend on an internet connection or cloud-based computing. The AI processing takes place directly on the device, enabling rapid, low-latency analysis while keeping the equipment compact and relatively cost-effective.
This edge-AI approach could prove particularly valuable in healthcare settings where sophisticated computing infrastructure is not readily available. While the technology could be deployed in major hospitals and high-volume laboratories, its compact design also raises the possibility of use in primary healthcare centres and facilities in resource-constrained settings.
Dr Jayaraj P B, associate professor, NIT Calicut, who led the team, talking about the invention, said, “Our team first focused on urine cytology screening, particularly the identification of cells and microscopic features relevant to urinary bladder cancer screening”.
But the researchers see possibilities far beyond urine samples. With suitable training, the underlying imaging and AI architecture could be adapted for cervical cytology, blood smears, histopathology slides and a range of other cytology and pathology applications, he added.
Dr Jayaraj further said, “By bringing AI closer to the microscope itself, our team is trying to make medical image analysis faster, less laborious and more consistent — while keeping the final clinical judgement firmly in the hands of medical professionals.”
The project brings together expertise from engineering, artificial intelligence and medicine. It was supported under the Scheme for Promotion of Academic and Research Collaboration (SPARC) of the Ministry of Education, Government of India. The technology is now being positioned for a journey beyond the laboratory. Its commercialisation is planned through HealForge AI Private Limited, a company incubated at the Technology Business Incubator (TBI), NIT Calicut. The objective is to translate the research prototype into a deployable AI-assisted microscopy solution for clinical and laboratory use.
The inventors behind the system are Arathy Menon N P, research scholar, NIT Calicut; Dr Jayaraj P B, associate professor, NIT Calicut; Dr Pournami P N, associate professor, NIT Calicut; Dr Simi S, assistant professor, department of pathology, Government Medical College, Kozhikode; and Dr Swathy Shanker, assistant professor, department of pathology, KMCT Medical College, Kozhikode.