Jay Singh Toor of Toronto is an Orthopedic Surgeon and the Founder of Primus Healthcare Solutions. Jay Singh Toor’s company is focused on generating efficiency in healthcare via the novel application of business principles including machine learning and predictive analytics. Below, Dr. Toor gives an introduction to how machine learning is being applied in medicine.
Jay Singh Toor of Toronto explains that machine learning is essentially the use of Artificial Intelligence (AI) such that information can be “learned” directly from data without reliance on an equation. It is an important component of the field of data science.
What does this mean for our health? Machine learning is making its way into our healthcare systems, as AI can detect, diagnose, and even treat serious illnesses, thus improving the efficacy of treatment and overall patient outcomes.
AI Development a Focal Goal
Jay Singh Toor reports that in
their recent survey, Deloitte found that 88% of business leaders are planning to increase spending in the field of AI in the coming year. And the healthcare sector is no exception.
By inputting vast amounts of data into hospital systems, the sophisticated AI technology will be able to identify patterns, calculate data and ultimately learn from previous results to perform patient diagnostics and recommend treatment programs.
Unsurprisingly, there is much debate surrounding the ethics of machine learning within such a sensitive field as healthcare, and the FDA has
proposed regulatory frameworks to regulate machine learning software as a medical device.
Where can Machine Learning Help?
Despite us hearing more about it in recent times, machine learning is not an entirely new phenomenon, according to Jay Singh Toor of Toronto. Since the 1960s, experts have been developing neural networking; namely,
creating artificial neural networks that replicate those of the brain, to predict brain patterns and therefore future brain activity.
However, recent advances in technology mean that machine learning is now being put into practice in healthcare systems worldwide.
A prime example of a way it will have been used to great effect has been the recent Covid 19 pandemic. The input of data of thousands, if not millions, of patients all over the world presenting with similar symptoms will quickly have given AI useful data to process.
Jay Singh Toor of Toronto explains that by consolidating this data and using software algorithms, machines that had learned a great deal about the illness from the data would be able to quickly come up with patterns and trends in the virus that would otherwise have taken doctors much longer to spot.
This can lead to faster disease recognition, more effective treatments, and ultimately better outcomes for patients suffering from the illness.
Furthermore, huge quantities of data can lead to predictions for the future. Sophisticated AI programs will be able to determine the pattern of a disease’s spread in the body, and in the case of viruses, among entire populations.
Dr. Jay Singh Toor’s Work
Dr. Toor’s work has focused on the application of business tools such as optimization to improve the efficiency with which health care organizations such as hospitals operate. Some of his recent work includes the implementation of surgical inventory optimization to reduce unnecessary costs (
https://www.sciencedirect.com/science/article/pii/S1553725021002440). Jay Singh Toor states that machine learning provides an opportunity to “supercharge” such optimizations.
Being able to predict hospital resource utilization will result in better optimization. For example, many hospitals attempt to divert cases away from expensive after-hours OR time via daytime dedicated orthopedic trauma rooms. However, there is some hesitation associated with this scheduling practice as
orthopedic trauma presentation is temporally variable. This may lead to unused OR time that would have otherwise been used for elective surgery. Jay Singh Toor believes that machine learning may be leveraged to both predict the presentation of orthopedic trauma and therefore reduce this uncertainty associated with dedicated orthopedic trauma rooms.
Future Directions
Effectively integrating established business and operations management tools can generate tremendous benefits to health systems striving to minimize expenses by maximizing efficiency. Jay Singh Toor states that although attempts have been made to apply such techniques in the past, AI and machine learning can offer improved model performance to help overcome the barriers hampering broader adoption.
