The tinkerer-turned-tech leader

By: Christina Hernandez Sherwood

When Mitchell Schnall, MD, PhD, was a young faculty member in Penn Radiology in the 1990s, he realized that the standard medical equipment used in the department wasn’t advanced enough for his purposes. Schnall, a radiologist specializing in magnetic resonance imaging (MRI) of the breast, said one problem was the radiofrequency coils, the small pieces of hardware that receive the signals making MRI possible. “The coils available from the commercial vendors at that time were limited,” he said. “So we [in the department] did our own coil development.” 

As a result, many of the radiofrequency coils used for routine clinical service at Penn Medicine in the 1990s were custom-built on site. For a time, Penn Medicine was the only health system with such advanced magnetic resonance imaging for the breast.  

While it was immensely gratifying to make such an impact, Schnall said, the one-of-a-kind innovation presented a practical problem: upkeep. “I’d get calls all times of day if a coil stopped working,” he said, “and I’d have to run over with my soldering iron to fix it.” (Eventually the coils developed at Penn were emulated by vendors and became standard of care.) 

Those early days of breast MRI are akin to today’s generative artificial intelligence (AI) landscape, said Schnall, who is now Penn Medicine’s inaugural senior vice president for data and technology solutions. Health technology vendors are developing generative AI tools that could help clinicians with some of their biggest burdens, like responding to online patient queries and documenting patient interactions. But, Schnall said, most of these products aren’t ready for broad use in the clinical setting. 

“We’re in the very early days,” he said. “So much of medical practice is context specific. Vendors don’t quite understand how local medical practice operates within a specific medical environment.” 

Still, medicine is on an unsustainable trajectory, Schnall said, with utilization of medical services accelerating and a workforce that’s resource constrained. “We need to find a different way,” he said. “Technology can help us develop creative approaches to how we engage patients and how we make the system more effective and efficient.” Schnall said he sees his new role, leading the Offices of the Chief Information Officer, Chief Medical Information Officer, and Chief Analytics Officer, helping to usher Penn Medicine into this new technology-enabled world. 

There will be growing pains, he said, like there were when radiology first transitioned from film to digital images. Schnall remembered, at first, trying to view the new digital images in the same way he used to view films, next to each other on a large sheet. “The aha moment is when we realized that we don’t want to look at the images next to each other [on a series of clunky computer monitors side-by-side]. Instead, we wanted to scroll through them in a single frame, and we can’t do that on film,’” he said. “This simple insight of reimagining how we view images not only made reading images on computer feasible, it was even better than film.” 

Always ready to flex for a goal  

As a child, Schnall was a below-average student up until about the fourth grade, struggling with learning difficulties that today would be more readily recognized as dyslexia. He adapted by inventing his own accommodations like memorizing text passages ahead of his turn to read aloud in class. Eventually, as the emphasis shifted from learning reading skills to critical reasoning, he said, “I found I picked up concepts quickly and didn’t need to read every last word of texts.” Looking back, he said, he has consistently played to his strengths in creativity and analysis to succeed academically, even as he still finds reading even long emails a challenge. 

Schnall came to the University of Pennsylvania as a 17-year-old undergraduate with a penchant for gymnastics and an interest in math and physical sciences. While most of his family members were doctors, Schnall struggled to define his own career goals. His acceptance into the University Scholars program helped Schnall plot out his own unique, and accelerated, trajectory. By the end of his third year at the University, Schnall had completed his undergraduate degree in physics. As a fourth-year student, he started the MD/PhD program while continuing as a member of the men’s varsity gymnastics team. 

After earning his PhD in biophysics and completing a residency in radiology, Schnall stayed as a member of the Penn Radiology faculty in 1990, often flexing his schedule to start work as early as 4 a.m. to make time for his family. In fact, Schnall first met University of Pennsylvania Health System CEO Kevin B. Mahoney on a youth softball field before their paths crossed professionally. Mahoney, who ran the local Little League and needed a coach for the 9-year-olds’ tournament team, asked Schnall if he would do it. 

“He said, ‘Sure.’ He didn’t say, ‘How many hours is it going to take?’ He didn’t say, ‘What are the rules?’” Mahoney said. “You ask Mitch Schnall to do something and the first answer is ‘yes,’ then he figures it out.” 

Schnall became chair of Penn Radiology in 2012. During his two terms as chair, Schnall was credited with doubling the size of the department and its research funding, as well as developing training programs for radiology clinician-scientists. Despite his increased responsibilities, Schnall has continued to devote at least one full day a month to performing MRI-guided breast biopsies, said Mark Rosen, MD, PhD, chief of the division of Abdominal Imaging. “At this stage of his career, to still be rolling up his sleeves and doing that is unique and speaks to his personality,” he said. 

Becoming chair was Schnall’s first time managing faculty members, and he went all-in on leadership development. With the help of Cindy Morgan, Penn Medicine’s vice president for learning and organizational development, Schnall organized quarterly leadership conferences for some 50 Penn Radiology leaders, including faculty from all radiological disciplines and clinicians from all Penn hospitals, nurses, technologists, businesspeople, and others. 

“It was very prescient of him,” Rosen said, “and suggested that he recognized, maybe for himself as well as others, that those of us who learned how to be radiologists, researchers, or educators may or may not know how to be leaders.” 

An integrated system and technological transformation 

Schnall was one of the first department leaders to capitalize on the power of “system-ness,” Morgan said, by building strong relationships with radiology departments across the Penn Medicine system rather than focusing only on the downtown campuses.  

Schnall also presided over a period of technological growth in Penn Radiology. Radiology faculty were among the developers of Penn AInSights, a real-time, AI-guided imaging system and clinical support tool that helps overburdened radiologists by creating more-precise, three-dimensional views of internal organs. “Mitch has always been an early pioneer in a department that has always been an early pioneer,” Mahoney said. 

Penn Radiology, which interacts regularly with just about every other medical discipline, is a natural leader within the health system, Schnall said. “We want to be in the early phase of working with prototypes and early products and helping to develop them with vendors,” he said. “I would take on stretch assignments and tasks as chair that spanned the health system. I learned a ton about how the health system worked.” 

An ambitious vision  

Schnall’s new health system role puts him in a different position than he was in as Radiology chair—leading a team not of faculty colleagues, but of information technology and data professionals. His plans range from efficiency efforts, like automating access by improving how patients are channeled through the Penn Medicine phone system, to system overhauls, such as reshaping the telemedicine service, Connected Health. 

Even more exciting are the plans that could have direct health care implications, such as leveraging artificial intelligence models to predict patient trajectories, Schnall said. “We’re going to be able to scour through records, find patients, and automatically give them insights,” he said. For instance, a patient at risk for developing diabetes could be reminded to get a blood glucose test. 

To facilitate this, Schnall is aiming to streamline Penn Medicine’s electronic health records (EHR) system, a data-rich but overwhelming platform that is part patient archive, part regulatory compliance system, and part transactional agent. “That imposes a certain level of structure on that EHR, which frankly makes it awkward to interact with it,” he said. “AI tools are going to allow us to put a layer between the EHR and us. We’re going to be leveraging AI technology and large language models and generative technology that are going to catalyze most of our interaction with the EHR.” 

But scaling at that level of innovation won’t be easy, Schnall said, in part because of the health system’s fragmented information technology landscape. Because of Penn Medicine’s growth through mergers over the years, the health system has two instances of its electronic health records system, along with multiple instances of other technology platforms. 

“To try to layer AI-based automation on a fragmented environment is going to be really difficult,” Schnall said. “So, in parallel, we’re talking about how to make investments to harmonize our systems, so they become the right foundation for this innovation. That’s a very serious lift. It keeps me up at night. But it’s absolutely necessary for us to scale across the system and see the benefit.” 

Still, Schnall said he is excited and motivated to be in a position that could make such significant change and impact. “I approach this not from a position of stress and fear, but from a position of excitement and anticipation,” he said. “I’m the ultimate glass–half-full guy.” 

Explore more stories about technology, transformation, and innovation 

This article is part of a collection that will be published in the spring 2025 issue of Penn Medicine magazine. See related articles including: