News Room
CII Media Releases
AI development process is a continuous & never ending cycle
Mar 24, 2021

There is an inevitable rise of digitization in life-sciences – and Artificial Intelligence based solutions are augmenting the capabilities of the healthcare workers. AI based diagnostics is estimated at $2.9 bln by 2025 with a CAGR of 36% between 2019-2025 (IDTech-X report 2020-2030). To explore the role of artificial intelligence in the lifesciences sector CII organised a workshop on AI in Healthcare: An Impending Revolution on 24th March on a Virtual Platform.


Speaking at the session Dr Ashwini K Aggarwal, Director, Applied Materials India said, “ Application of AI in lifesciences is increasing in delivery of healthcare towards a favourable patient outcome. AI tools are seen as increasing productivity and efficiency of healthcare professionals. We need to explore how we can create an enabling technology landscape through appropriate AI intervention.”


Dr. Prof Rakesh Lodha, Division of Pulmonology, Intensive Care, Tuberculosis, All India Institute of Medical Sciences (AIIMS) mentioned that there are different components of healthcare which are preventive, promotive and curative. Idea is to see how we can use the power of computational analytics. Areas of AI application are in diagnostics, radiology and pathology, addressing resource & manpower limitation and interpretation of data and in interpretation of slides in pathology. AI is also needed in a complex ICU setting where multiple organ systems have been affected and early warning system to predict serious infections and hemodynamic instability. The challenges of AI application are accuracy of models, generalisability, demonstration of benefit in clinically relevant outcomes, quality of input data may vary, fitting into the workflow and it should be explainable.”


Dr Sujoy Kar Chief Medical Information Officer Apollo Hospitals mentioned the important yards of conceptual highlights in building clinical AI algorithms which are, “ Ideation, data management analysis & modelling, ethics & regulatory aspects, validation, publication/presentation, API integration, consortium networks, commercial platform.  He mentioned, “We need to understand where bias can set up in the system.” From the ethics imperative in AI he explained the EASE model that comprise of ethical considerations i.e. the principles that govern the ethical aspects of developing, validating and implementing AI and digital health, adoption i.e. how are clinical AI and digital health implemented, suitability i.e. the aspects that deal with what is suitable for a clinical AI program and explainability i.e. how we make the clinical AI models and components of digital health much easier for understanding and implementation.”


Putting forth the industry perspective Ms Geetha M Data & AI Excellence and Capability Lead Philips Innovation Campus mentioned of the quadruple aim of AI that includes lower cost of care, improved patient outcomes, staff experience and improved patient experience. The main challenges that AI in healthcare faces are large quantity of data with appropriate patient privacy protection, curated and clean databases, trust and transparency, combining data and knowledge driven learning, understanding of the ethical context, review cycles with clinicians, new ways of working and fragmented healthcare IT infrastructure. She mentioned that the AI development process is a continuous and never ending cycle that includes gathering of data, labelling, development, testing, deployment, receiving feedback and monitoring.”


Putting forth the start up perspective, Dr Vidur Mahajan Head (R&D) Centre for Advanced Research in Imaging, Neuroscience & Genomics said, “ AI is the future of medicine and is helping in reducing cost, improving quality & increasing access to healthcare. AI is a huge market which can reach Rs 45 billion dollar by 2026. We have developed a comprehensive AI research portfolio through deep collaboration across the industry that include brain/head, chest, breast, automated bone age, spine, liver and prostate segmentation. However he mentioned that in reality we do not see AI being applied to these areas which are mainly due to challenges related to techno-clinical trust and bias in AI and integration of AI into hospital IT systems is expensive and difficult.”


Artificial Intelligence is having a big impact on the life science sector. Diagnosing and treating patients is becoming more effective and pharma companies are relying on the up-and-coming technology to speed up their drug discovery projects. AI algorithms are processing images faster and results are comparable/ more efficient vs humans though current value is primarily driven as a decision support system. Applications are ranging from Cancer, Cardio-vascular, Retinal, Respiratory to Neuro-degenerative diseases.


24 March 2021

New Delhi

Email to a friend   Print
Related Information