Visualizing healthcare data

In today's healthcare systems, vast amounts of data are collected and tracked around the clock. They come from many sources, such as electronic health records, hospital management systems, and more. In addition, there is data from diagnostic centers, laboratories, pharmaceutical companies, and other devices in hospitals. But access to all this data is worthless without proper data visualization to help medical staff interpret it faster, recognize trends, and make better decisions.

Kaloian (Kal) Petkov has made it his mission to leverage the latest developments in visualization and computer graphics to support applications in healthcare. As a senior key expert in imaging and visualization at the Technology and Innovation Center at Siemens Healthineers in Princeton, United States, he investigates new techniques to help healthcare staff and improve patient outcomes through better visualization.

Kal’s passion for computer graphics started in middle school in Bulgaria when he took computer science classes. He knew very quickly that he wanted to pursue this passion. So he studied computer science and mathematics at the Lakes Forest College in Illinois, United States. Later he specialized in medical and immersive visualization at the State University of New York at Stony Brook.

His passion quickly became a calling when he gained his first insights into medical visualization in healthcare, especially rendering technology. “This is a good way to convey information, to synthesize information, and help people understand. And this was something that I liked, and I wanted to pursue a career in it,” Kal remembers.

In the area of surgical guidance, for example, they use computer graphics for minimally invasive surgery, such as laparoscopic procedures. “We generate overlays using computer graphics techniques that synthesize patient data, pre-operative data, and planning data, and display it on top of the video feed in such a way that we can help the surgeons do a better job,” Kal explains. “And by using the correct depth visualization techniques, we allow them to see 3D structures in ways they could not before.”

Another field of application is education and training. Here they also use techniques from the movie industry to create photorealistic renderings of patient cases. The idea is to train future doctors and clinicians using actual patient- and disease-specific data. Another field that is a focus of Kal and his colleagues: They want to use 3D image visualization to support dialogue with patients by allowing physicians to communicate information about the disease in a more illustrative way.

For these visions to become a reality, the visualization techniques must take many things into account. Whether it is the quality of the data or the environment in which they are used, the challenges are manifold.

One way to address these challenges is to use artificial intelligence techniques to help generate the visualizations. For photorealistic renderings, for example, Kal and his colleagues up-sample and denoise data to speed up the rendering process. “In the future, we want to do much more than that,” Kal says. The idea is to automate the visualization in order to provide the relevant images with minimal user interaction when the clinician opens a specific case.

His work combines different aspects of research and practical software development, which is exactly what Kal likes about it: “We investigate new techniques of how to improve patient outcomes through better visualization, but we also implement and deliver the software that clinicians use to accomplish this.” In the end, the combination of practical software development, research, and art is what makes his job fulfilling – and that he can have an impact on patients.