The prospects of AI in radiology
The panel discussion on AI in radiology at 73rd IRAI conference focussed on how radiologists can enhance their practice using AI applications.
Panellists in this discussion Dr Deepak Takhtani, Dr Harsh Mahajan, Dr Shailesh Lunawat, Dr Jan Kimpen, Dr Amit Kharat, led by Dr Deepak Patkar, delved into understanding the opportunities associated with AI, addressed the fear surrounding it and spoke about the opportunities that lie ahead. Experts on the panel started by examining the events in which digital technologies are driving exponential growth of health data. The panellists spoke about the many challenges and major opportunities. They looked at various AI’s clinical applications available to Indian radiologists and examined workflows associated with it.
Dr Jan Kimpen said, “AI will help turn large amounts of data into actionable insights to support and empower people. Integration of AI is critical across the full end-to-end flow in radiology notably as an enabler, but also as a foundation to the integration.” Further, talk about the need for radiologist to integrate AI into their workflows, Dr Amit Kharat pointed out that around two-third of the world does not have access to X-ray. Therefore, AI can be that assistant which ensures and improves access to radiological services across the world.
The panellists also spoke about licensing of AI products in radiology and what should be the pre-requisites for the same.
Dr Deepak Takhtani and Dr Harsh Mahajan addressed queries related to what makes the connection between radiology and AI so special compared to the rest of the specialities in medicine?
Dr Takhtani pointed out that there are many challenges in the way to make AI truly understood and applied. It is still a long way to go for organisations, radiologists and the imaging community to crack the code in AI but the journey is exciting and has immense scope, he asserted.
During the discussion, a pertinent query about the commoditisation of radiology services was raised. Some radiologists felt that there is a risk that technological innovations are commoditising imaging as technical services and could undervalue the expertise of radiologists. To that, Dr Harsh Mahajan agreed that commoditisation of radiology is already happening, and with AI, the commoditisation will increase. Having said that the community needs to understand that this makes for a strong business case where tapping the untapped areas is a must and that by doing so even the patients will be at an advantage.
Further on, Dr Amit Kharat expressed his views on the significance of getting involved with patients and ensure that AI enhances patient experiences. “It is important to get involved with patients. Look at clinical radiology and AI can assist the radiologist to do so,” he added.
The panel then discussed the RoI when it comes to investing in AI. Experts on the panel referred to burnout faced by many AI companies and unanimously said that right now the RoI cannot be immediately recovered. The current focus is to make the machine learn. But it will be a long-term goal.
At the end of the discussion, the panel spoke about the ethical and medico-legal aspects involved. The question of who owns the data kept lingering and the panellists were hopeful of finding answers to these questions soon.