Geetha Manjunath: Championing early detection of breast cancer

Traditional mammography, while effective in older women, isn’t as useful in screening younger women due to several factors, including breast density. It’s also a method that involves some discomfort and is invasive in terms of a person’s privacy.

Further, in middle- and low-income countries, the screening rates themselves are abysmal, Manjunath points out. In India, it’s as low as 1.3 percent, whereas in Europe it’s as high as 80 percent. And nearly 70 percent of breast cancer deaths are seen in the poorer countries due to late detection.

“We have developed a new method of detecting cancer where we just measure the temperature variations on the chest using a thermal sensor and convert the temperature map to a cancer health report automatically using AI,” Manjunath tells Forbes India.

It’s like a changing room experience, she explains. A woman enters, closes the door and removes her clothes by herself. While she’s seated in front of the imaging device, the operating technician remains outside. The process requires about five minutes of cool air to be blown first and then the imaging itself takes about two minutes. Overall, in about 10 minutes, Thermalytix will come up with its report.

View the full list here: 

The results flag only those women who need further diagnostics. This approach not only lowers costs to about 20 percent of a standard mammogram but also increases accuracy by 30 percent, particularly for women under 45, who are often missed by conventional screening.

In addition to the hardware devices—the higher-end versions are installed in some 200 hospitals and diagnostics centres today—the magic of Thermalytix is a suite of patented machine learning algorithms trained on datasets gathered through clinical studies in India, Europe, Africa and a smaller study in the US as well.

Also read: Ganesh Ramakrishnan's BharatGen Vision: Inclusive, diverse, self-reliant

The AI analyses subtle heat patterns—often invisible to the naked eye—produced by increased blood flow and metabolic activity in cancerous tissues. The cloud-hosted analytics engine delivers a risk score and report, which can be reviewed by clinicians for further action.

A portable version that can be carried in a backpack is also available for government rural outreach programmes, for example, or for health care-focussed NGOs.

Niramai has screened more than 280,000 women thus far. And Thermalytix has regulatory clearances in both the US and Europe. Niramai aims to deepen its presence in existing markets, expand to 35 to 40 countries, and scale up manufacturing through local partnerships. Manjunath is also working to expand the application of her AI-thermal imaging platform to diagnose other diseases and body parts, including whole-body screening and fever detection.

“A common learning for all entrepreneurs in deep medical technology is that the go-to-market is really hard,” says Manish Singhal, founding managing partner at pi Ventures, an early backer of Niramai, and a pioneer in deeptech VC investing in India.  Especially when there is a novel technology, and in spite of FDA clearance, practitioners often have a bias of not wanting to use a new product, Singhal says.

“Today there is momentum for deeptech in India, and all such startups stand to reap benefits from it,” he says. The government at the highest levels is taking note of the entrepreneurial efforts in deeptech, and linkages are beginning to grow with the defence sector, which will provide a critical fillip.

At Niramai, “that 200 hospitals have adopted Thermalytix and the product has touched so many women shows that there’s a degree of product-market fit”, he says. “From here on, how do we scale this 10x, and how quickly can we do that?”