Cancer remains one of the biggest health issues in India, with as many as 7 lakh new cases of the disease emerging every year.
If the data from National Institute of Cancer Prevention and Research (NICPR) is anything to go by, women in the country are at a greater risk zone, with one in every 22 women in urban areas prone to develop breast cancer during her lifetime.
While medical experts look for a curable solution, a USD 107 billion cancer care industry has spawned across the globe as per IMS Health report, 2015.
Several startups have emerged in India that claim to bring cost effective treatment for the poor. One such startup is the Hyderabad-based Onward Health.
The venture is grabbing eyeballs with Onward Assist, their AI-based software which they claim helps predict the risk for breast and cervical cancer.
Founded in 2017 by Hyderabad based engineers Dinesh Koka and Vikas Ramchandra, infant venture managed walked away with the Jury’s Choice award at the recently concluded LetsVenture's angel funding seminar, LetsIgnite, at Bangalore for the product.
The makers of Onward Assist insist that using machine intelligence and computer vision techniques, their product helps in the early diagnosis, patient management process and even critical decision making.
The computer vision-based algorithm zeroes in on the affected part and then finds similar regions using image comparison and generates a risk score.
Speaking to The Passage, Dinesh Koka said, “The idea was to make something to help in cancer treatment, primarily around decision making. Initially we were looking at multiple types of data that include mammography, histopathology and EMR data. And later we focused on mammography and histopathology, with attention to breast cancer and cervical cancer.”
Meet Onward Assist
The startup claims that their algorithm has been successful in reducing unnecessary biopsies significantly, when used in conjunction with a radiologist reading. They specifically choose breast cancer as their focus because there is a high incidence of it and there is a huge healthcare cost to be borne by patients and as well as the hospital.
“The tool developed helps in decision-making for a radiologist, where it looks at the image and provides inputs on where the tumor could be. The tool looks at the region and does a contouring of the tumor and gives a risk score – which helps deciding whether it’s benign or malignant tumor. This is where it makes a difference and unnecessary painful bioscopies can be reduced, by improving the accuracy of the mammography reporting. Also, the time taken to decide by pathologist will be cut down by 50 percent, that now takes 40-45 mins,” said Koka.
The machine learning tool can help histopathologists improve their productivity, report on more number of cases per day with more accuracy and hence reduce cost of reporting. The tool can also extend the diagnostics to low resource settings in partnership with pathologists and radiologists, which would also lead to inherent cost benefits for pathologists, Koka explained.
The startup receives incubation at CIE (Centre of Innovation and Entrepreneurship) in IIIT Hyderabad. Since August last year, they spent the next few months on initial discussion with doctors and other healthcare experts and making the product. It was only this year when they hit the market for clinical tie ups and partnerships with heal hubs, including cancer institutes.
So far they have a total of six partners and talks with a few more are in the pipeline, the founders say. Out of the six, some are leading cancer institutes and some are histopathology diagnostics centres. Their next focus is to start talking to larger set of customers and deploy the first set of tools in commercial partnership. The team has a target valuation of INR 9-10 crores (USD 1 million approximately).
With a vast chunk of population outside insurance covers, the cost of cancer treatment in Indo-China eludes the average pocket.
As mentioned in ‘Cost of Treatment for Cancer: Experiences of Patients in Public Hospitals in India’ report, published in Asian Pacific Journal of Cancer Prevention, 2013, the total costs of treating a typical cancer case may cost INR 10,00,000 (USD 14,000) with each chemotherapy sitting costing INR 90,000 (US 1,300)”
“There is a worldwide shortage of pathologists and our tool, Onward Assist, will really help speed up the process of diagnosis. The accuracy of this cannot be commented on yet, as it is still developing, but, I think an application like this will surely be accepted by the doctors,” an expert on histopathology working with the Onward Health team, said on the condition of anonymity (as dictated by the terms of an MoU between her organisation and the team.)
Not a lonely path
Other Indian Startups that are working on cancer research include Bangalore-based Niramai, that is working on an artificial intelligence pain-free breast cancer screening solution; AIndra Systems that uses AI-based computer visions for screening cervical cancer and UE Lifesciences, who have set up a handheld breast cancer screening device for use across rural and urban India.
The last word
Dr Manish Jain, Director of Medical Oncology at Ruby Hall Clinic’s Kamal Nayan Bajaj Cancer Centre, Pune, does not see an imminent miracle happening.
“Artificial Intelligence in diagnostics is like Google speaking of driverless cars now. It may be helpful in the future but right now it’s too premature and the accuracy of AI (in cancer treatment or diagnostics) cannot be trusted.”
However, Dr Jain is quick to add that since AI might be able to generate a more precise diagnosis, the cost of treatment may be brought down heavily.
“(Using AI) Since you will know (the location, and extent of the from the beginning itself, a targeted treatment can be planned which will bring down the cost.”