AI in medicine: A supplement, not a substitute

CHENNAI: Artificial Intelligence (AI) has led the way towards predictive healthcare. While it reduces the duration of diagnostic procedures by accurate and easier reports, it also assists doctors with complex procedures. However, with limited medical evidence on the application of AI in various specialities, its implementation in providing timely and efficient healthcare services to the patients needs to be explored more.

Supervised tech

Doctors say that AI has had a transformational impact on the healthcare sector in terms of diagnosis, treatment and surgeries in the recent decade. For example, AI algorithms can identify potential drug candidates more efficiently than traditional methods thus enabling new drugs to the market while reducing costs.

“Though ethical concerns on the safety of patient data and privacy remains a challenge, robotic surgeries in renal health, for instance, have made complex and challenging surgeries effective, and hastened quick recovery. Advanced systems and technologies have resulted in greater and better outcomes in surgeries with AI assisting doctors and technicians in treating critical cases,” says Dr Venkat Subramanium, executive director, Asian Institute of Nephrology and Urology.

He added that though factors like empathy and concern are missing from AI tools, human-control was possible especially in managing, regulating and operating such devices/tools without losing the core benefit of the process/surgery etc.

Better analysis

AI and predictive analytics help us to understand more about the factors that influence our health even when we are in robust health. It’s not just about predicting when we’d get the flu or the medical conditions we’ve inherited, but linking our environmental factors – place of birth, eating habits, work, and the air pollution levels in the neighbourhood we live and work – to our holistic health.

A hospital is no longer the place that covers a broad range of diseases; instead, it focuses care on the acutely ill and implementing highly complex procedures, while less urgent cases are monitored and treated via smaller hubs and spokes such as retail clinics, specialist treatment clinics and even people’s homes.

“These locations are connected to a single digital infrastructure. Centralised command centres analyse clinical and location data to monitor supply and demand across the network in real time. AI can also spot patients at risk of deterioration. This network can also remove bottlenecks in the system and ensure that patients and healthcare professionals are directed to where they can best be cared for or where they are most needed,” explained Chandrasekhar Nair, co-founder and CTO, Molbio Diagnostics.

He added that AI has already made significant inroads in imaging-based diagnosis. Chest X-rays are being read with AI and even cough vocalisation with AI can diagnose diseases. AI enables reading of blood smears to provide instant CBC results.

“All these efforts are transforming the way we identify diseases and advise precision therapy. This will have a tremendous impact in healthcare delivery, especially in resource limited settings. The first AI driven drug discovery candidate has entered clinical trials. This promises a new era in drug discovery too,” he said.

Application in specialities

Various hospitals have been using new AI tools in specialities such as reproductive medicine, radiology, neurology, pulmonology, urology, orthopaedics, critical care, trauma, emergency medicine, etc, and also in hospital administration.

Dr Iyappan Ponnuswamy, medical director and chief radiologist, Kauvery Hospital, said: “AI has been revolutionising healthcare to a great extent with precision diagnosis, improved patient care especially in critical care and emergency, and helping in faster turn-around times in report interpretations, digitisation of data, etc.”

The healthcare industry, on the whole, realised the true relevance and value of AI, perhaps, when COVID-19 hit. “AI proved to be useful in triaging COVID patients based on severity of lung involvement on CT scans, which was done using an AI tool called Predible at the hospital. This cut-short the time taken by emergency physicians to make split-second decisions during the second wave,” he recalled.

Another area of usage was in measuring the volume of certain structures in the brain which was difficult to measure manually. “An example is an AI tool called Neuroshield. We use it in our hospital. Neuroshield uses algorithms that help diagnosing many diseases including age-related cognitive decline, mood and psychotic disorders, and other neuro-psychiatric conditions. UroLogiQ is being used for analysing CT scans of the abdomen to analyse stones in the kidney and ureter along with their associated complications,” explained Dr Iyappan.

He said that ChatGPT-based applications for clinical decision-making were being tested. “It has potential to analyse patients’ medical records and provide guidance on diagnosis, investigations and treatment. It’s also being evaluated as a symptom checker,” he pointed out.

Predictive healthcare

The role of predictive healthcare, especially in reproductive medicine, has significantly increased, as witnessed by its applications across the globe.

Doctors say that India has seen many developments in reproductive health and the new research offers promising benefits for patient care and standardisation of medical care. Dr R Sujatha, Head-Embryology at Nova IVF Fertility, says, “In the last few years, a large part of medical research presented at international conferences consists of developmental projects in AI. The technology is very useful in IVF procedures and reproductive medicine. AI can pick up the right sperm and make a careful selection that can comparatively result in a good embryo. But we need medical evidence from many years to establish this system. We need to see if they’re as accurately applicable in India and local settings to see how beneficial it is.”

Limitations

While AI can be used as a guidance tool for clinicians, it has not been authorised it to make clinical decisions. Despite the fast-pace development of AI in healthcare, medical professionals say that AI’s application cannot become a substitute to the human resources we have. Chandrasekhar Nair, co-founder & CTO at Molbio Diagnostics, opined that AI must complement and not replace the expertise and judgment of healthcare professionals. “Human oversight and ethical considerations should always be incorporated to ensure patient safety and confidentiality,” he pointed out.

Concurring with him, Dr Sujatha added that the new AI advancements would benefit patients especially if it can be made affordable for the middle class. “However, it’s completely a choice of healthcare providers – based on the type of AI being put to use, to levy the cost of these technologies on patients or bear them as part of patient care,” she explained. “Digitisation of records and standardisation of procedures would also need to be done to link medical history and update them with AI.”