Healthcare is breaking out of silos, becoming increasingly connected, but also increasingly complex. While it poses an opportunity for medical professionals to learn more than before, the enormous amount of medical information can overwhelm the decision-making processes. This is where artificial intelligence (AI) systems come in.
Here’s how it works: Data from different sources are fed into a system. With the use of algorithms, programs and systems to simulate human intelligence, AI can analyse big data sets. These data sets can be from clinician’s notes, reports from a patient’s file, medical research publications, and clinical trial outcomes. In a matter of seconds, it can create an actionable gist of all, for a particular clinician or a group to review.
Ajit Narayanan, CTO, mfine, Bengaluru explains, “Typical applications in healthcare fall in the categories of diagnosis processes, treatment plan development, drug development, personalised medicine, patient monitoring, process automation, robotic surgery and clinical trials.”
How it helps
Take the pharmaceutical industry, where technology is being used to curtail lead time. “AI and machine learning use data sets and algorithms to model various scenarios in the drug-discovery process. This modelling is used to arrive at a predictive hypothesis much faster than the traditional trial-and-error method, which results in much shorter time for creating proof of concepts,” says Suhas Tamras, Global Head, Medical Devices and Healthcare Practice, Tata Elxsi, Bengaluru.
AI in healthcare can help to leverage technology to deploy efficient, impactful interventions at exactly the right moment in a patient’s care. “As patients demand more from their providers, and the volume of available data continues to increase at a staggering rate, artificial intelligence can provide insights into diagnostics, care processes, treatment variability, and patient outcomes,” says Dr Garima Anandani, Head (Clinical Operations), QI Spine Clinic, Mumbai, that uses a knowledge management system to diagnose, manage, and prevent pain in the back and neck. These insights could mean the difference between over-prescription, over-testing, and over-treatment.
Some important aspects of computational systems are that they can perform repetitive tasks without feeling bored or fatigued and can analyse billions of data bytes in a matter of seconds. “Now, if we train such systems to seek specific information in such huge piles of data and throw up results, it can be done easily. Further, we can grade and qualify such information. It may not be as intuitive as a human, but it can be precise,” says Arindam Haldar, CEO, SRL Diagnostics, Gurugram.
AI systems can bring in better standardisation of processes, and therefore subjectivity in interpreting information will be reduced. There are AI systems already in operation that have demonstrated the ability to identify sight-threatening conditions with equal accuracy to human ophthalmologists. There are those which can scan a human chest X-ray faster than a medical professional can, and detect a small tumour or an early onset of pneumonia. Some trained neural networks can interpret pathology images of tumours at a success rate of detection upwards of 90%, compared to an expert pathologist.
AI-assisted robotic surgery, where robots are able to analyse pre-op medical data and guide a surgeon’s instrument during surgery, ensures patients develop fewer complications than otherwise.
The biggest impact of using AI is that it can significantly improve efficiency, while reducing wastage and costs. In a resource-constrained environment with a low doctor-patient ratio, AI has the potential to deliver remote medicine and create virtual access in an effective manner. This does not mean the doctor will be eliminated, but that it will be an aid for the doctor.
“Doctors with AI can treat and monitor patients across geographies. Healthcare organisations can plan policies, guidelines, strategies and infrastructure to address health needs in a precise manner, thereby optimising the resources and delivery. The State can prioritise plans, budgetary allotments with greater understanding and in optimal utilisation,” says Hanumantha Rao Chitipothu, Co-Founder and CEO, HealthSignz, Bengaluru that calls itself a health intelligence platform using AI.
It can also be preventive in nature. “The auto diagnosis tool is an AI-powered tool that runs on actual diagnostic tests results and provides probable risks factors. Additionally, it also suggests users change their lifestyle by recommending diet and exercise plans. Further investigations (if any) and a list of repetitive examinations for regular health tracking and monitoring are also suggested,” says Deepak Sahni, Founder and CEO of Healthians.com, Gurugram that offers testing at home.