- EHR Data Boosts Machine Learning Algorithms for Chronic Disease
- Machine Learning Algorithm Brings Predictive Analytics to Cell Study
The two conditions share several symptoms, including fevers and rashes. Early diagnosis is necessary to improve outcomes for KD and MIS-C, the press release notes.
Researchers from the University of California-San Diego have created an artificial intelligence (AI)-based machine-learning algorithm to differentiate between MIS-C and KD in an effort to enable early detection of the diseases.
“For 40 years, the Kawasaki research community has tried to create a diagnostic test for KD and failed,” said co-senior study author Jane C. Burns, MD, a pediatrician at Rady Children’s Hospital-San Diego and director of the Kawasaki Disease Research Center at UC San Diego School of Medicine, in a press release. “But now, in just the space of 18 months, we have created a physician support tool that differentiates MIS-C from KD in children using simple test results and five features of the physical exam that any health care provider, clinic or hospital can do, with accuracy exceeding 90 percent."
Researchers gathered data from 1,517 patients at Rady Children’s, all of whom previously received a diagnosis for either MIS-C, KD, or other febrile illnesses. They then used a deep-learning algorithm known as KIDMATCH to compare patient age, the five signs of clinical KD, and 17 other laboratory measurements.
“Deep-learning algorithms have shown high performance capabilities in industrial applications, such as voice recognition or machine translation, and are good at identifying multiplicative risk factors,” said Shamim Nemati, PhD, associate professor of medicine at UC San Diego School of Medicine, in the press release.
They found that the algorithm was effective in identifying the difference between the two conditions during the internal validation phase.
Researchers externally validated the algorithm using patient cohorts from the Boston Children’s Hospital, Children’s National Hospital in Washington, DC, and the CHARMS Study Group Consortium. They found that the algorithm achieved an accuracy rate of 90 percent or higher.
“The hope here, of course, is that KIDMATCH can help front-line clinicians distinguish between MIS-C, Kawasaki disease, and other febrile illnesses so that they can provide earlier, appropriate treatment and prevent severe complications,” continued Nemati.
The use of AI and machine learning to enhance the diagnosis processes, and in healthcare overall, is becoming more common.
The results of a study published in September described the capabilities of an AI-and nanotechnology-based blood testing system in accurately detecting tuberculosis among pediatric patients.
In August, researchers from the University of Florida announced plans to develop an AI algorithm to track COVID-19 variants. The researchers will train the algorithm using publicly available global data on the genetic sequences of the SARS-CoV-2 virus.
By 