Vivek Prakash, Director, SW Engineering, NetApp reiterates that a strong and effective mechanism to evaluate AI based techniques for clinical accuracy is critical
With advancement in science and technology, healthcare has improved dramatically over the last century. Average life expectancy has improved significantly. While there are many factors contributing to this, the use of technology in healthcare has played a major role in improving quality of life and increasing life expectancy. Some breakthrough technology interventions happened as early as 1958 when the first ever pacemaker was implanted in a person so that the heart pulsates artificially! Invention of MRI (Magnetic Resonance Imaging) machines to non-invasively diagnose health anomalies and Laser for vision correction are yet other examples of early disruption on technology in healthcare.
Today technology can be leveraged for not only better diagnosis but also improving the accuracy of diagnosis and efficacy of treatments. Although we have a long way ahead of us with intervention of AI and robotics in medicine. The ability to collect data, process it and give a clear output, distinguishes AI from other technologies to impact healthcare.
Artificial Intelligence (AI) is predicted to impact healthcare by facilitating:
Access to good healthcare starkly tilts in favour of the urban population across the world. Individuals in rural areas travel substantial distances to get good care. According to a report by KPMG and the Organisation of Pharmaceutical Producers of India, 75 per cent of dispensaries, 60 per cent of hospitals and 80 per cent of doctors serve 28 per cent of the Indian population. This 28 per cent lives in urban areas. AI can help bridge this gap.
Smartphone ownership and usage is high even in rural areas. AI based healthcare chatbots can help improve the rural healthcare accessibility gap. Chatbots can provide assessment and recommend initial steps, like immediate medication or tests to be taken. This saves time for the individual and enables healthcare practitioners to serve more patients.
According to the World Health Organisation, globally 2.6 million deaths happen annually due to medical errors. So far technology has given healthcare practitioners and doctors more data to make decisions. The interpretation of that data has traditionally been left to humans. AI steps in with its ability to aid healthcare practitioners with intelligent diagnoses. For example, a radiologist will look at an individual’s image and based on experience will analyse what is abnormal. AI can help in speeding up this process, bringing in more precision and consistency.
Patients are compelled to take multiple opinions before they zero in on the treatment for a disease. There are variations in different doctors’ analyses and lines of treatment. A lot depends on how experienced a particular doctor is. This experience is based on learnings built over the years of being a practitioner, having learnt by handling multiple similar cases and being appraised of the latest research and findings. AI can help condense that experience so that diagnoses are more predictable, precise and consistent.
For AI to be effective, it needs data to learn from. Data is the fuel behind the intelligence and predictive ability of AI algorithms. For better diagnoses AI algorithms will need to understand a patient’s symptoms and medical history, co-relate the symptoms with potential similar cases and check the recorded analysis of experts. The biggest challenge here is the varied data that needs to be pulled in from multiple sources. This means reaching out to multiple EHRs (electronic health records) from different hospitals or health centres and analysing images. Non standardisation of EHR data and lack of interoperability of this data, are critical challenges that need to be overcome in order to make faster progress in this space.
Future will be hybrid cloud
AI solutions in healthcare and other fields can make an impact because of the modern compute and data services infrastructure which is available. While technology is leading the way in providing the necessary compute power for AI algorithms, without having a sound data strategy it will be difficult for organisations to scale and win in the field of AI. Hybrid cloud data management services like data redundancy, data security, data mobility, and data availability play a pivotal role in enabling the right data to be available at the right place and time, whether it is residing in on-site data centres, or the cloud.
However, standardised and quality data will be the key
The prospects of AI being used in healthcare are exciting. While AI based diagnosis has made strides in radiology and is considered on par with human-like capabilities, comprehensive clinical diagnosis based on AI still has a long way to go before it becomes mainstream. A strong and effective mechanism to evaluate AI based techniques for clinical accuracy is critical. With hundreds of startups focused on exploring different aspects of AI based healthcare services, the future of health tech looks promising. Over next few years, we will see industry bodies and governments working closely with tech enterprises in order to overcome the data standardisation and regulation challenges, thus paving the way for AI in the healthcare space.