What does the future hold for artificial intelligence in healthcare?
Can you imagine a future where babies will use smart clothes to track their every move? It may sound like some science fiction, but a pantsuit piloted in Helsinki, Copenhagen and Pisa does just that.
The “baby coverall” (MAIJU) looks like a typical baby garment, but there is a fundamental difference – it is full of sensors that assess a child’s development.
“MAIJU offers the first of its kind quantitative assessment of a baby’s motor abilities through the ages from lying down to walking smoothly,” says Professor Sampsa Vanhatalo, Project Manager at the University of Helsinki. “Such an amount has not been possible anywhere, not even in hospitals. Here we bring the solution to homes, providing the only ecologically relevant context for engine evaluation. ”
The Old House describes the path from the wishful thinking of a solution to a possible clinical implementation as a “windy road.”
“There is no shortage of dreams or technology, but we lack relevant and adequate clinical problem statements, ecologically and contextually relevant data sets, reliable clinical phenotyped material, and appropriate legislation for products that do not follow traditional forms,” he says.
Thanks to machine learning, researchers in Helsinki discovered latent properties in baby’s motion signals that could not be identified by traditional heuristic design.
“At the same time, we need to remember that artificial intelligence in medical applications can only be as intelligent as we let it be,” Vanhatalo adds. “Real-world situations are much more complex than we hope, and the ambiguity of many clinical situations or diagnoses significantly limits our ability to build as accurate artificial intelligence solutions as we hope. For example, it is not possible to train and validate a classifier for countless medical diagnoses without clear boundaries.”
The Old House also believes that the medical community needs to identify sensible targets for artificial intelligence.
“It is much more fruitful to train clinical decision support systems (CDSS) than to train clinical decision systems,” he argues. “All men hope for the latter, and others for fear; but the responsibilities arising from the decisions, including legal ones, are so great that it is difficult for me to see that any company would dare to commercialize such solutions. In fact, I can already see how the legal risks of such responsibilities, whether indirect or illusory, create a bottleneck for the commercialization of many good artificial intelligence products.
At the forefront of oncology
One area of medicine where artificial intelligence has great potential to revolutionize treatment is oncology. Professor Karol Sikora, Chief Medical Officer (CMO) of Cancer, Rutherford Housebelieves that machine learning can benefit physicians by assisting in complex treatment decisions.
“Several commercial solutions are available to identify and map nearby organs at risk for cancer,” Sikora explains. “Precision oncology requires the analysis of large amounts of data in an unprecedented way, and we hope that artificial intelligence will benefit patients in the long run.”
Rutherford Health’s network of oncology centers leverages the latest innovations in cancer technology, such as artificial intelligence in the design of radiotherapy.
According to Sikora, machine learning can also be a huge benefit in the future, increasing patient choice. “Artificial intelligence could make patients understand the risk-benefit equation associated with any intervention,” he says.
Mystical artificial intelligence
But in order for healthcare organizations to take full advantage of the potential of artificial intelligence, the “noise” around it needs to be addressed, says Atif Chaughtai, head of the global healthcare and life sciences business. software company Red Hat.
“Properly used artificial intelligence has enormous potential to save lives and manage ever-increasing health care costs,” Chaughtai says. “In the future, artificial intelligence will continue to evolve and will be widely used as an assistive technology to perform tasks more accurately and efficiently with the people who make the final decisions.”
He adds that the success of artificial intelligence requires organizations to change at a manageable pace and work together to innovate intelligent business processes.
“Often, as data scientists or IT professionals, we don’t spend time understanding our customers’ business processes, which leads to poor change management, ”he says.
Old House, Sikora and Chaughtai will speak at the hearing Unlocking the future of artificial intelligence at HIMSS22 European Health Conference and Exhibition, June 14-16, 2022.