A Chinese man has his face scanned by a self-service machine supported by face recognition technology of Alipay, the online payment service of Chinese e-commerce giant Alibaba's Ant Financial, to pay fees at a hospital in Dongyang city, Jinhua city, east China's Zhejiang province, on July 24, 2019.

Information technology continues to transform many fields of activity through increasingly sophisticated digital and visual data analysis and correlation work on a massive scale – and determining how useful information can emerge. Readers may recall my September 26 article in Asia Times, “AI assists human intelligence – it doesn’t replace it.” Here we turn our attention to medicine, where information technologies have a growing impact on the delivery and quality of medical care.

I recently met with Dr Steven Safyer, chief executive of Montefiore Medicine, which is the university hospital group for the Albert Einstein College of Medicine of Yeshiva University in New York City, where I have served as a trustee. Montefiore is a remarkable example of growing success in the health-service field, with profitable annual revenues of about US$6 billion.

We discussed the major initiatives that have allowed this success. Dr Safyer stressed the importance of their investment in computer systems and information technology in enabling the successful growth from a local organization in the Bronx, New York, into a major metropolitan region organization with close to 30 locations including clinics and hospitals.

What is noteworthy is that the medical-service revenues are largely from government Medicare and Medicare reimbursements that are modest at best and make profitable hospital operations challenging. Montefiore managed this challenge to profitability while maintaining quality and building a number of world-class specialty care centers, with a focus on technology. As a result, quality medical services are delivered at affordable cost. Safyer noted that such investment was not possible for local hospitals, which hindered their ability to operate profitably. As a result of its successful operational model, Monetefiore absorbed a number of loss-making regional hospitals.

Information technology was implemented in stages. First came computer systems to track and record care delivery digitally, replacing the handwritten records infamous because of the notoriously poor handwriting of so many doctors. Then came analysis and data management to improve the efficiency and quality of medical care.

With digital records tracking patient care, an enormous database has been accumulating, dealing with the historical care patterns of many thousands of patients. As a result of using sophisticated data analytics and artificial intelligence software, it became practical to analyze historical patient data to study the unfolding of a disease and and its treatment. Using such data, physicians can tailor disease treatment on the basis of large amounts of historical information that would not otherwise be available.

When I discussed this topic with Dr Seven Rudolph of Mount Sinai Hospital in New York, he pointed out that sophisticated image-analysis software using machine-learning techniques is increasingly used to improve the speed and quality of analysis of images from X-ray, MRI (magnetic resonance imaging) and other tools in ophthalmology as they relate to patient disease. In fact, subtle image patterns can be discovered and compared to historical data of known provenance to give radiologists new insights and clues to patient problems.

In another area, and longer term, IT research worldwide is correlating patients’ genomic data with their medical history. I recently visited the Weizmann Institute in Israel, one of the world’s leading research centers, where I had the opportunity to learn about this work. Its program consists of working with 2 million patient history records from a health maintenance organization (HMO) and analyzing the medical history in light of the patients’ genomic data.

One of the objectives of this research is to determine correlation information of patient symptoms and disease history and their medical care and determine what pattern of diagnostic information led to disease conditions. For example, identification of symptoms at an early stage could prepare people for early treatment at a stage when the disease can be most easily controlled.

We have just touched here on the surface, because new AI techniques are being applied to more and more health-care-related activities. Combining these new tools, a new generation of practitioners is being trained that is learning to be much more efficient in prescribing treatments and reducing wasteful procedures based on guesswork. We all remember the wisdom of age that comes to physicians with extensive experience. Now the lessons learned from experience on a large scale are becoming available to practitioners to reduce errors and improve care.

Leave a comment

Your email address will not be published. Required fields are marked *