The All India Institute of Medical Sciences (AIIMS) Delhi, in collaboration with the Centre for Development of Advanced Computing, Pune, has launched an Artificial Intelligence (AI) platform – iOncology.ai. – designed for the early detection of breast and ovarian cancer.
While this indigenously developed technology has already been deployed in five district government hospitals, including Mathura and Faridabad, for validation, officials confirmed that in the future, the government plans to install the system in other medical centres with the assistance of National Information Centres and the Ministry of Information Technology.
Speaking about the latest addition, M. Srinivas, director, AIIMS, said that the Centre is focused on providing healthcare to ensure patient satisfaction, and to make available an ecosystem that is affordable, accessible, and has assured world-class quality.
Meanwhile, AIIMS noted that according to the Global Cancer Observatory estimates, there were 19.3 million cancer cases reported worldwide in 2020, with India ranking third, right behind China and the United States of America. A recent study published in the Lancet predicted that cancer cases in India would increase to 2.08 million, accounting for a rise of 57.5% by 2040.
In India, over 8 lakh deaths were caused due to cancer in the year 2022. Late detection of cancer remains the primary cause of deaths. It is globally estimated that out of 80% of cases reported late, only 20% survive. It added that out of 20% of cases detected at an early stage have a survival rate of 80% or higher.
“Early diagnosis is thus the key to reducing fatalities due to cancer. One of the key issues in manual diagnosis of cancer is reporting of false negatives, meaning falsely declaring a cancer patient as healthy. Use of AI significantly helps in reducing such false negatives,” Dr. Alok Thakkar, Professor and Head, National Cancer Institute (NCI), Jhajjar-AIIMS, said.
He added that at the core of this latest platform lies a sophisticated AI system capable of analysing complex medical data, including diagnosis, with unprecedented accuracy and efficiency.
AIIMS, in its release, explained that the iOncology.ai platform encapsulates deep learning models. As the volume of data, both clinical and radiological/ histopathological, increases, this platform has the capability of self-learning to constantly improve its outcome.
As breast and ovarian cancers are reported to be the most common cause of cancer amongst women in India, first application of iOncology.ai platform has been done by AIIMS in these two domains. Accordingly, the learning models used by the platform have been specially tailored for early detection of breast and ovarian cancer. Officials associated with the programme added that the present system has built a large data set of about half a million radiological and histopathological images. This data has been collected during examination and diagnosis of about 1,500 patient cases of breast and ovarian cancer over a period in AIIMS. The AI platform was successfully used by various departments in AIIMS itself also.
“Encouraged with its success, the AI platform has since been implemented at five district hospitals in the country,” Ashok Sharma, additional professor, Department of Biochemistry, National Cancer Institute (NCI), Jhajjar, said.
He added that the AI platform was showcased recently at the Med-hackathon event during Research Day celebrations at AIIMS on January 30. Over a hundred participants (clinicians as well as researchers) interacted with the iOncology.ai platform and showed keen interest in this innovative development. The platform’s user-friendly design and robust performance with high level of accuracy was highly appreciated by the visiting cancer experts during this event.
AIIMS said that it is looking forward to partnering with several other cancer hospitals and research centres in the country. “As this is an indigenously developed technology, it will be easy for AIIMS to make this platform available to these organisations,” it added.