Fighting breast cancer with AI-powered mobile, non-invasive solutions

This October, Google announced that a deep learning-based artificial intelligence (AI) algorithm called LYmph Node Assistant, or Lyna, that it had developed last year has reported a 99 per cent accuracy in detecting breast cancer. An even more heartening news is that similar cutting-edge technologies for cancer detection are already in use in India, where a few health-tech firms are employing them to test for breast cancer — a disease which has a 50 per cent fatality rate in the country. 

Breast cancer’s shockingly high fatality rate in India is due in large part to the fact that it is mostly detected at later stages. And the reason for the late diagnosis is that there is a stigma attached to getting one’s breasts examined. To combat this problem, a bunch of Indian innovators are using technologies such as thermal imaging, AI, machine learning,  big data and analytics to enable women to undergo the test in a private and non-invasive way.

“Our cultural fabric is such that cancer is seen as a taboo. Nobody freaks out if you have cholesterol. We want to reach a stage where every woman goes for regular screenings and it’s not a stigma anymore,” says Nidhi Mathur, chief operating officer of Niramai, a Bengaluru-based health-tech firm which has developed a patented AI and machine learning-based software to identify lumps in breasts.

Niramai, which stands for Non-Invasive Risk Assessment with Machine Intelligence, has come out with a diagnostic technique called Thermalytix which combines thermal imaging with artificial intelligence to detect breast cancer at an early stage. A high-resolution thermal sensing device and a patented software developed in-house scans for any thermal abnormality in breast tissue. 

Niramai’s software analyses this temperature distribution to generate a diagnostic report with a breast health score. Thermalytix takes 400,000 temperature values and analyses it using 120 abnormality indicators, thereby identifying high-risk patients who need a follow-up test. “Thermography can detect breast cancer in ways that are non-invasive, non-ionising and non-traumatising,” says H V Ram Prakash, head radiologist, at Vydehi Hospital in Bengaluru, which uses Niramai’s testing technique.

For patients, the procedure is a unique experience as it is non-touch and non-invasive. The thermal-sensing device is placed at a distance of three feet from the patient and even the technician who operates it cannot see her. “I believe Niramai will enable easy adoption for early cancer screening and will play a key role in improving outcomes in our ongoing battle against breast cancer,” says Kiran Mazumdar-Shaw, CMD, Biocon.

Geetha Manjunath, Niramai CEO and technology veteran, says that unlike mammography, which does not work on younger women who have dense breast tissue, thermal imaging technology can be used on women under 40. “Young women only go for an ultrasound after they find a lump during self-examination, which is already stage 2 or 3 of cancer, if it is malignant. But Niramai can address the problem at an early stage even for women below 40, thus increasing the rate of survival,” says Manjunath, who has a PhD from the Indian Institute of Science, Bengaluru, and who had sown the seeds of the start-up in 2016 after two of her cousins were diagnosed with cancer.

Niramai’s tests are currently being used in hospitals in Bengaluru, Mysuru, Mumbai, Chennai, Hyderabad, Pune, Dehradun and Bhubaneswar. But the goal is to reach the small pockets of the country, says Manjunath, as the technology is a tenth of the cost of a mammography. What’s more, the software and hardware can be set up within minutes at any camp or hospital. Backed by investors such as Pi Ventures, Binny Bansal, Axilor Ventures, and Ankur Capital, Niramai has so far tested around 5,000 women.

Health-tech innovator Mihir Shah’s iBreastExam is also combating the killer disease with an affordable and mobile solution. The hand-held device of iBreastExam uses patented ceramic sensors developed at Shah’s alma mater, Drexel University, to detect minute inconsistencies in breast tissues. Healthcare workers perform painless and radiation free breast examinations using a mobile app and get results within few minutes. 

While this test can cost between Rs 500 and Rs 1,000 in private hospitals, Shah’s company has been able to bring it down to between Rs 70 and Rs 280 by tying up with governments. 

Manjiri Bakre, another IISc alumna, also decided to use machine learning to fight the scourge of breast cancer after she lost a friend to the disease. “Chemotherapy is like an atom bomb that kills even the healthy hair cells and blood cells,” says Bakre. Her company OncoStem Diagnostics, is addressing the problem of over-treatment via chemotherapy through its CanAssist-Breast test.

CanAssist costs around Rs 60,000 and works in 70-80 per cent of patients who are diagnosed with early-stage breast cancer. The test is performed on the patient’s tumour tissue, and the test result is fed into the company’s machine learning-based statistical algorithm, which gives a code — a number between 1 and 100. If the number is below 15.5, the patient is deemed low-risk and not prone to metastasis. In that case, she could avoid the harmful chemotherapy procedure.

If AI and machine learning can be so effectively employed to fight breast cancer, the day not be far when artificial intelligence becomes our biggest benefactor when it comes to tackling deadly diseases.