63-year-old Nagesh was in agony. Nursing an injury on his left foot for nearly two weeks with no medical intervention, the pain had become unbearable that morning. The foot throbbed as Nagesh wondered what to do. He slowly hobbled towards his neighbour’s house and requested to be ferried to the nearest PHC, almost 70 kms away. Upon reaching the PHC nearly two hours after, Nagesh was delirious from the pain. The doctor on call realized that the elderly man’s foot infection had reached the bone and he could possibly have to undergo an amputation. If only Nagesh had a way to attend to his injury when it was minor, he wouldn’t be an amputee.

The story of Nagesh is tragically common in many of India’s villages and towns. Access to healthcare, availability of specialists and high costs impede the quality of life in the fringe areas of the country, and limit last mile delivery of healthcare services to those in dire need. Chronic pain, skin diseases and other ailments that could be detected and treated well, if caught in early stages, continue to plague those who live in remote towns and villages. By the time they seek medical help, it is too late.

Biomedical engineer Geetanjali Radhakrishnan was all too aware of the plight of people like Nagesh. Having grown up along the Western Ghats, Radhakrishnan knew how challenging it was for villagers and rural inhabitants to seek timely medical help who had to live with conditions like diabetes, cardiac diseases, skin infections and burn injuries among others. “I always wanted to build something that could enable early screening and diagnostics in low-resource settings.” As she experimented with different ideas, her paths crossed with Dr. Ayush Gupta, who worked at DY Patil Hospital in Pune as a dermatologist and associate professor. They met a farmer near Pune, who unfortunately suffered from a similar condition to that of Nagesh. Radhakrishnan understood the dire need for point-of-care diagnostics to assess wounds and skin conditions with precision and speed, so they could seek timely help.

In December 2015, Radhakrishnan started Adiuvo Diagnostics, a techno-social enterprise, aimed at developing platform technologies in opto-electronics to effectively aid in early disease detection, specifically customized for low-resource settings. The Chennai-based startup has developed a novel imaging device that captures multispectral auto fluorescence images and uses machine learning algorithms on the obtained spectral images to aid in label-free, rapid diagnostics. This technology has wider application areas from pathogen detection to skin parameter assessment to cancer diagnostics.

Multispectral auto-fluorescence imaging is a novel technique that leverages the tissues' and microbiome's native fluorescence and aids in label free detection. Biomarkers like NADH, Flavin, Porphyrin, Di-tyrosine linkages, Chitin, collagen etc. have characteristic emission when excited with various wavelength of light sources. Analysis of these spectral images at various emission bands and applying machine learning/deep learning techniques on these images can be effectively used in label free detection from pathogens to antibiotic susceptibility testing to cancer diagnostics.

While conducting field tests, Radhakrishnan and her team realized that their datasets of skin conditions were not suitable as they failed to diagnose Indian skin tones. This is when the team decided to deploy AI to effectively remove the noise generated from different skin colours and to compensate for the gaps in data. This is being done through their product Illuminate, which can deliver non-invasive, label-free, microbial detection and classification of wounds in real-time. Radhakrishnan adds that they have already filed a patent for this application. This invention is of utmost importance, as it addresses the high costs involved with procuring a camera for multi-spectral imaging – usually ranging from Rs.15 lakh to Rs. 1 crore. Since then, Adiuvo Diagnostics has conducted trials with GMERS, Valsad – Gujarat; DY Patil Hospital, Pune – Maharashtra and Prof. M Viswanathan Diabetes Research Centre in Chennai – Tamilnadu, which helped them develop a Minimum Viable Product (MVP) by 2019 following which they received approval from CDSCO for the product.

(Illuminate by Adiuvo Diagnostics being used at a government hospital in India.)

Adiuvo Diagnostics is supported by BIRAC, Menterra, Venture Center, Artha and Villgro. In 2020, the startup won the Qualcomm Design in India Challenge. The support from Qualcomm was especially useful since Qualcomm chips allowed for customisations, provided the necessary processor, imaging modalities and peripherals needed to manufacture at scale.

While the startup was well on track to start rolling out the product in hospitals across India, the COVID19 pandemic abruptly halted their plans but Radhakrishnan says the past year and half has been a time of great learnings, as they further improvised the product and are hopeful that they can launch the product in hospitals very soon.