AI in healthtech is elevating the output of treatment

Dr Bharat Aggarwal, Principal Director – Radiology Services, Max Super Speciality Hospital, Saket explains the role of AI in healthcare

The role of Artificial Intelligence (AI) in medicine is evolving at a rapid pace. While artificial intelligence has been used in healthcare since the late 20th century (see figure1), the penetration of clinically relevant AI solutions being introduced in healthcare has multiplied manifold in the last decade.  This is exemplified by the number of FDA approved AI solutions in medicine – while a handful of products were approved every year by FDA in the 1990s, multiple products are approved on a monthly basis in 2024.

AI is omnipresent in healthcare, enabling improvement in patient engagement, productivity, efficiency and also impacting the quality of healthcare delivery on a day to day basis. A patient has numerous checkpoints in their journey for a medical problem. Smart “bots” are now being increasingly used by healthcare providers to simplify the appointments and inquiries that any healthcare facilities deal with on a daily basis. Patient scheduling – for doctor visits and investigations can now be managed real time ensuring both patient comfort and satisfaction.

The greatest impact in healthcare has been seen in image & pattern recognition applications in specialities like radiology, pathology, genomics, dermatology and even ophthalmology. AI is increasingly used in any field of medicine that is digitised – again radiology has seen a large impact of solutions extending from easy and accurate patient positioning, shortened image acquisition times, higher image resolutions and pattern based diagnostic aide solutions.

An area of concern in patient care has been the enormous data – both text & image based; that is generated for from an individual patient to an entire healthcare system. The use of AI, and more recently large language models is helping mine this data, filter it and make it more useable. For example, discharge summaries are now being created using AI models by scanning the entire record of a hospital admission, and ensuring that the summary includes a comprehensive and accurate information of all facets of the hospital visit.

Longitudinal healthcare data is being used to build predictive algorithms for identifying individuals at risk for certain diseases; and even to build “digital twins” for improved prevention and treatment of diseases for individuals, giving a large boost to personalised medicine. Such solutions are being built by capturing data from personal devices like wearables and sensors, investigations, hospital visits and other medical records that have the potential to recognise onset of critical and chronic ailments at an early and preventable stage, reducing morbidity and mortality in times to come.

Innovative ideas are being explored scientists and entrepreneurs to build solutions that are not limited to healthcare facilities like clinics and hospitals. Use of data acquisition from mobile phones and basic sensors are taking healthcare to remote and rural parts of the world. This not only helps patients to identify diseases early, but also enable paramedical and rural healthcare workers diagnose and treat patients in areas that do not have access to doctors.

While AI is becoming mora and more pervasive in our lives, including its applications in medicine, there continue to be questions about its authenticity and accuracy – the AI models are dependent largely on the training datasets that are used for development, and comprehensive, high quality & accurate data is necessary to build solutions. There are also legal issues and ethical issues like data privacy and social & ethnic sensitivies that need to be, and are being addressed by governments and professional organisations. One solution to this is to build AI models with a human in the loop where in the AI models, especially those impacting medical treatment are used as doctor assistants, rather than independent decision making entities.

We are living in a world of great transition and it is still early days for the use of AI in healthcare. The coming few years will see judicious and impactful of use of AI in healthcare.

 

artificial intelligence (AI)health techtechnology
Comments (0)
Add Comment