CHENNAI: What can a selfie do other than fetching a few likes on Facebook and the odd follower on Instagram? Very soon, it could help find whether you have breast cancer. As part of a new initiative to diagnose breast cancer non-invasively, researchers from IIT-Madras, Paterson Cancer Centre (PCC) and NIEIT would also examine whether a selfie of the breast taken by a smartphone with an infra-red clip on camera could act as a simple diagnostic exercise.
“At the moment, we are using very high intensity infra-red thermal cameras to capture images of the breast. This thermo mammogram records the temperature differences inside the breast that can differentiate between normal tissue and cancerous cells. Thus, the software will be able to read the image and tell whether there is indication of a tumour or not,” said Professor V Kamakoti, Department of Computer Science and Engineering, IIT-M. Though smartphones do not come equipped with IR cameras at the moment, clip-on cams are available in the market with apps that could scan and share images.
Despite the fact that several higher intensity diagnostic mammograms like MR and nuclear-isotope based scans are being used currently, all of them involve the injection of nuclear dyes or contrast into the body. This is primarily why the research team has built the new thermal mammographic unit at Paterson Cancer Centre at Vadapalani. “At the moment it’s intrusive, takes a long time, requires an expert to be present and it is not portable. I was quite shocked that women needed to go through such a process in the 21st century,” said Balaji Lakshmanan, founder of imakeRobots, which has constructed the robotic arm that swivels the camera around in this new device.
All the patient needs to do is lie down with the breast suspended through an opening on the examination table. “The robotic swivel that has the IR camera takes images rapidly at gaps of 15 degrees until a 360 degree panoramic heat-image of the breast is captured. The operator is behind the screen and can move the robot without making the patient feel uncomfortable about baring their breasts,” added the IIT alumnus. The hardware support is from National Institute of Electronics and IT while the software that studies the images will be developed at IIT-M.
Explaining how their first goal was to build a repository of thermal images to get the software to “safely” diagnose which breasts were safe and which ones had masses, Professor Kamakoti said. “Today in the country there are only 450 cancer centres or so and if 100,000 scans are done, an expert can’t see every one. It is statistically impossible. So we want a software to do the detection. This way, we can filter out about 80-90 per cent of the patients who are safe.”
“IIT-M has a data sciences lab with many students and professors working there. There is also an image processing group there. It’s going to be a decades-long journey and we look to people like NIELIT to commercialise it,” he added.
The key to the indigenous project is that besides being cost-effective, crores of women could get more than a yearly diagnosis done without worrying about the expenditure.
“In India, 37 per 1 lakh population are vulnerable for breast cancer. This means that about 4.6 lakh women have it across the country. If this system succeeds then even slow growing tumours can be detected as women can get multiple checks in a year as opposed to one expensive mammogram, once a year,” said Dr S Vijayaraghavan, Chief Clinical Oncologist, Paterson Cancer Centre.