IBM Watson is able to synthesize vast amounts of structured and unstructured data from textbooks, published research reports, clinical reports. It then combines this information with patient’s data file, attributes and available clinical information to suggest the most suited treatment plan, empowering the clinicians to make better decisions and deliver better quality care to patients, way faster than the current set up. In fact we at Manipal have also partnered with IBM Watson.

AI as Augmented Intelligence automates the basic tasks in care giving through workflows. The workflows can be triggered by alarms, other workflows or by the caregivers. Here; we can see example of nurses who can find veins easier with augmented reality. Accuvein is using AR technology to make vein puncture easier and less traumatic. Many times, we miss the vein on the first prick and the numbers are actually bad when we are doing for elders and children. Accuvein uses augmented reality by using a handheld scanner that projects over skin and shows nurses and doctors where veins are in the patients’ bodies. Such technologies could assist healthcare professionals and extend their skills.

AI, as in Automated Intelligence, is based on doctors no longer to be in loop and the control is handed over to the machine (as in driver-less cars). This situation is very risky. The solution is often ambiguous requiring discretion and judgment from the clinicians. But there are some situations where automated intelligence is helpful like in some cases hospitals are using wearable technology to improve patient outcomes while is also reducing expenses. Problems that are caught early on can be treated quickly, which usually translates to fewer expenses. The results have been impressive so far. In California, a Medicare company that uses wireless glucose monitors has reduced amputation rates for diabetic patients while spending below the average national rate per patient by 18 percent.

Some other areas where AI can support healthcare is drug development. AI and machine learning can offer cost effective solution by having virtual search. Atomwise did it for Ebola virus the company used deep learning technology (AI) and supercomputers to assess the potential efficiency of candidate drugs and predict with a high degree of accuracy, greatly reducing the time necessary for the development, testing, and dissemination of new medicines. Another area is Precision Medicine, in this machine learning system models fundamental aspects of molecular biology and therefore, can be applied to any disease segment and any area. This may include biomarker discovery, interpretation of genetic variations etc.

Also AI can help as Cognitive Radiology Assistant, which means a tool with advanced multimodal analytics, superior reasoning capability and powered by a wide range of clinical knowledge offers precise clinical decision making in radiology and cardiology. The best example here is The Medical Sieve by IBM and it filters essential clinical and diagnostic imaging information to drive anomaly driven summaries and recommendations that dramatically reduces the load off the radiologists and clinicians without compromising the quality of diagnosis.

Challenges

There are always some challenges in implementation of any technology and in AI for healthcare; the major challenge is standardization of data across organizations and also to work in collaborative way. The other challenge is to have large data sets for developing neural learning algorithms and once developed we need thorough validation of outcomes and need to cover various geographies and races for reliable outcomes.

It is also observed that AI is capable of handling structured data like lab test values, but when it comes to unstructured data like radiography images it require inferences. In my opinion, AI is helpful in healthcare and redefining healthcare on daily basis and if leveraged wisely, AI coupled with IoT could be the answer to solve several healthcare challenges that the world is facing today. The world of “big data” is so huge that we will need “artificial intelligence” to keep a track of it, which means faster, accurate clinical decisions. If we don’t agree than we need to think twice!