A young woman steps nervously into a makeshift screening tent. Inside, a small tripod-mounted thermal camera waits. Asked to place her hands behind her head, she sits still for a few minutes while the device maps temperature points across her chest. Within moments, Thermalytix, an AI-powered thermal imaging system, generates a simple triage report indicating whether further mammography is needed. Developed by Niramai, Thermalytix uses AI trained on large datasets of thermal and radiological images to detect early signs of breast cancer through a non-invasive, radiation-free scan. The technology has already screened over 3.5 lakh women across India, emerging as one of the most promising tools among a new wave of startups making early cancer detection more accessible and affordable.
India’s breast cancer incidence has multiplied dramatically over the past three decades, with new cases rising nearly fivefold since 1990—a surge that mirrors an alarming global trend—says a new global analysis published in The Lancet Oncology in March 2026. The analysis also highlighted that India logged nearly 2.03 lakh new breast cancer cases in 2023, marking a 477.8 per cent rise compared to 1990. In the same period, breast cancer deaths reached 1.02 lakh, a 352.3 per cent increase. Little wonder doctors are looking for a solution that would help them diagnose at the earliest. AI screening seems to be offerng that solution.
Dr Geetha Manjunath, Founder CEO of Niramai, believes early detection can help India fight the growing cancer epidemic. “If detected at an early stage, the treatment cost can be less than 10 per cent of what it would cost to treat stage three or stage four cancer.” However, multiple barriers such as economic, health resources, sociocultural, and disease apprehension, make early detection difficult. Manjunath says AI-enabled solutions like Thermalytix can reduce these barriers. A 2026 study by Lund University shows AI could help detect breast cancer earlier and lower the number of women diagnosed with the disease years after screening. The study tracked women and evaluated for MASAI (Mammography Screening with Artificial Intelligence) and found that mammography supported by AI identified more cancers during routine screening and reduced breast cancer diagnoses over the following two years by about 12 per cent. Another analysis, published in The Lancet Digital Health in 2025, reported a 29 per cent increase in cancer detection without a rise in false positives. Breast cancer is only one part of a larger shift. Across India, startups and research labs are experimenting with AI tools for detecting other cancers as well.
For instance, Periwinkle Technologies has developed Smart Scope CX, a portable, non-invasive handheld device designed for rapid, point-of-care cervical cancer screening. Similarly, Qure.ai offers qXR, an AI-driven tool that analyses chest X-rays to identify lung nodules that could indicate lung cancer. Researchers at the Indian Institute of Science, Bengaluru, are also working on Aarogya Aarohan, an AI-based system aimed at detecting early signs of oral cancer.
Veena Moktali, co-founder, Periwinkle believes these screening devices can complement the traditional pap smear or HPV test. “Both these tests are infrastructure dependent and require a number of expert technicians who can operate various machinery to get these samples. Moreover, the invasive procedure associated with these tests keep them beyond the reach of a lot of women in India—especially women from rural India.” Periwinkle has screened over 4.5 lakh women at grass roots level. Of these, close to 7,000 women, Moktali claims, have been identified at pre-cancerous stages. As per analysis published in Science Direct, the Southeast Asia Region ranks second, contributing 29.59 per cent of global cervical cancer cases and 34.12 per cent of deaths, with India alone reporting 1,27,526 new cases and 79,906 deaths in 2022.
One of the reasons behind high cervical case incidence and deaths is poor screening coverage and use of subjective screening tests. Indian government does conduct free screening tests for cervical cancer for women aged 30-65, focusing on Visual Inspection with Acetic Acid (VIA). But studies have shown that VIA has a varied sensitivity and lower specificity in detecting precancerous lesions due to its high subjectivity, as its results are highly dependent on interpretation skills and the examiner’s level of training and expertise.
In India, cancer is often detected much later, resulting in poor prognosis. Looking for cancers before any symptoms or “cancer screening” is a powerful strategy for prevention, early diagnosis, and better management. However, cancer screening numbers in India remain very low. For instance, the percentages of women who have ever undergone cervical, breast, and oral cavity screening were 1.9 per cent, 0.9 per cent, and 0.9 per cent, respectively, while only 1.2 per cent of men participated in oral cavity screening.
This is the gap many AI-based screening tools hope to fill. But, oncologists urge caution. According to many, the promise of these AI-enabled pre-cancer screenings must be backed by stronger evidence. “Each screening test has a sensitivity and a specificity, and they have to reach a level where cost-effectiveness suggests that yes, it is a cost-effective tool that is better than our previously existing tools, or that it offers an advantage in lead time. In that respect, most AI-enabled solutions haven’t yet proven that advantage over the existing ones,” says Dr Kanika Sood Sharma, director, radiation oncology at Dharamshila Narayana Hospital. Senior oncologist Shipra Gupta believes AI screening solutions are more helpful with analysing or simplifying the test reports, especially working like a co-pilot of an experienced oncologist.
Munjunath believes, “The biggest hurdle in large-scale adoption of these solutions is collaboration and partnerships with organisations that have access to large populations. A policy-level decision by the government would make a huge difference.” Oncologist Sharma echoes this view, stressing the need for clear government guidelines and policy support for large-scale adoption. “At present, we are screening only about one percent of our population,” she says. “If screening has to reach the grassroots, regulatory bodies and policymakers must first evaluate the credibility of these technologies and then integrate them into community programmes.”
Whether AI-powered tools ultimately transform cancer screening or remain support tools for doctors will depend on stronger evidence advocating for their efficacy, wider adoption by the medical fraternity and policy intervention by the government.
For the young woman who walked into the screening tent, the test lasted only a few minutes. But for entrepreneurs, researchers and doctors across India, a larger question remains: Will AI help millions more detect cancer earlier; or is it just another disruption?
AI or Nay
Beyond the efficiency, and AI bias, oncologist Kanika Sood Sharma believes absence of any post screening follow-up or comprehensive guidelines raises questions around many of these tests. “For instance, what if a self-assessment AI tool reports presence of malignant growth, but the traditional test does not indicate any of such growth? In such cases, what should a patient rely on becomes a serious question.” In 2026, the Central Drugs Standard Control Organisation (CDSCO) issued a notification stating, software that uses AI to detect or diagnose cancer by analysing X-rays, CT scans, and other medical images will now be regulated as Class C medical devices. The government is already using AI screening solutions for TB detection under Cough Against TB (CATB) and for diabetic retinopathy screening using MadhuNetrAL solution. Though the inclusion of AI enabled devices is a welcome step, oncologists suggest comprehensive guidelines, approvals and serious policy intervention approvals are needed for these innovations to be accepted at a larger scale.