Researchers at Tomsk-based SibMed University (Siberian State Medical University) have developed new hybrid artificial intelligence (AI) technology to control multiple cyber-physical systems across medical disciplines.

At the core of the development are AI-driven systems to support physicians’ decision-making process which are based on data and on knowledge.

As modern computing and communication devices are getting increasingly compact and affordable, the creation of an array of sensor- and actuator-enabled cyber-physical systems is no longer a future plan. Sensors and other slave gadgets help collect tons of data on processes that take place in physical and biological objects.

What the SibMed team focused on was the development of decision-making support systems for doctors which could operate where datasets available are very limited and where both data and empirical and theoretical knowledge are required to do the job.

“Most AI systems for medical purposes serve the disciplines with huge amounts of information collected, such as X-ray diagnostics images, text data in electronic medical records, etc. Unfortunately, such systems cover but a fraction of what medicine requires to meet its needs,” said Ivan Tolmachov of SibMed’s Bionic Digital Platforms program.

The developers have already tested the new AI solution in assessing biomechanical movement parameters in locomotor disability rehab cases and in optimizing premature babies’ nutrition profile with unremitting blood sugar level monitoring.

With the technology, a broad range of diagnostic and therapeutic equipment could be coalesced in a single automatically controlled platform to capture and follow all processes that occur in the human body in real time.