This blog is a contribution from our client University Medical Centre Mannheim, a leading university hospital in Europe. Learn how their team is using DataRobot to accelerate clinical research with AI.
As physicians and researchers, we are constantly striving to improve the quality of life. This requires a holistic, data-driven approach that helps us understand the full impact of specific treatments. How does a specific treatment affect a patient’s symptoms? What makes symptoms better or worse?
At the University Medical Center Mannheim, we are digging into the data to answer these questions. But it is difficult for our small team to balance research with the critical care of patients. We don’t learn data science in medical school, but it is an increasingly essential piece of the healthcare puzzle.
Fortunately, user-friendly AI platforms like DataRobot are helping to bridge the gap between our medical expertise and the data science needed to provide more thoughtful care to patients.
Benefits of AI in Healthcare: Reliable Results
What excites me most about AI in healthcare is its potential to uncover new explanations for diseases or breakthrough therapeutic effects that would otherwise be missed using traditional statistical methods. Our goal is to discover new influences on disease progression, predict disease progression, and better manage patient adherence to treatment.
DataRobot offers exciting new ways to gain insights from your data and empower your team without the need for a data scientist.
As clinicians, we can compare and validate models to find the one with the highest accuracy. In our clinical collaboration, Healthy Skin and Joints, we used AI to evaluate data from smartphone apps, including images of anonymized patient data and other clinical data sets.
Compliance is also important. From privacy measures surrounding patient information to GDPR regulations that protect and secure sensitive data. When we publish results, reproducibility is paramount. That’s why documentation and explainability behind models are so important. DataRobot seamlessly automates these typically labor-intensive processes.
With DataRobot, we trust our research results because we know they have been thoroughly trained, retrained, validated, and revalidated. We have a wealth of statistics that demonstrate the level of accuracy required to publish our results. That helps me sleep better at night and allows our center to make a greater contribution to the medical research community.
Uncovering the links in disease progression
Another example of AI being used in medicine is our collaboration with dermatologists and rheumatologists, where we applied AI to several cases.
In a recently published study, we used DataRobot to analyze clinical trial data from patients with chronic eczema or psoriasis. The analysis focused on itch, pain, quality of life, and tracking symptoms using a smartphone monitoring app. We looked to discover novel effects on disease progression, predict disease exacerbations, or promote patient adherence to treatment.
Analysis showed that about 30% of patients had improved quality of life after 6 months, while another 30% had a deteriorating or persistently poor quality of life. These and other insights inform treatment decisions. This data is transformative because it allows us to better understand patients and to know which patients benefit from certain treatments. It also informs when and how to change a patient’s course of treatment, if necessary.
Now we are helping other clinics in the medical center to find insights from their data. With the Internal Medicine department, we looked at blood lipids as a way to predict heart disease or heart attack. In just a few weeks/months, we were able to create a fairly accurate model and we expect to publish the results in the near future.
Accelerating medical research using AI
Without DataRobot, none of these discoveries would have been made. Instead, we were able to accelerate our research from hours to seconds, while continuing to focus on seeing patients and improving their quality of life.
AI helps us in our daily tasks, and most importantly, helps patients.
When I first partnered with DataRobot, I told people that this new technology was going to change the face of the planet and that they needed to learn about it. I still say that today. AI offers tremendous benefits to healthcare professionals and I am very excited to see the impact of what the University Medical Center Mannheim is doing.
About the author
Victor Olsavszky is a clinician and postdoctoral researcher at the Dermatology Clinic of the University Medical Center and Medical Faculty in Mannheim, Germany. His clinical specialties are dermatological surgery and plaque psoriasis, and his research interests include vascular biology and AutoML analysis of medical data sets using DataRobot.
Meet Victor Olsavszky, MD
