This article is part of a sponsored content series produced in partnership between Texas Monthly and Texas State University highlighting big ideas that move Texas forward.

The patient couldn’t go near a gas station. In fact, the mere smell of gas took him back to Iraq, where he fought in an endless succession of violent skirmishes. One day, an IED exploded right by his Humvee. The blast singed his hair, and the strong oil odor was in his nose, on his clothes, everywhere. The veteran has struggled with PTSD ever since and is one of the 40 million adults in the United States who suffer from some kind of anxiety disorder. Since he’s returned to the States, his car has always run on empty. By the time this patient met Alessandro De Nadai, Ph.D., the last thing he wanted to do was fill it up.

“That’s a completely natural reaction,” says De Nadai, an Assistant Professor of Psychology at Texas State University. “People naturally avoid the situations that make them anxious, but that often makes the anxiety worse.”

How do we automatically turn this info into something doctors can use in helping patients, De Nadai wondered. Better yet, how can we use it to get treatment right the first time?

Since he was a twenty-year-old undergraduate at the University of Georgia, De Nadai has been fascinated by two seemingly unrelated fields: clinical psychology and computer science. Yet as he continued his studies, earning two master’s and a doctorate, De Nadai realized how those two fields can complement one another and help patients get the treatments they need. This veteran was a prime example.

It was 2017, and De Nadai was wrapping up his doctorate in Clinical Psychology with an internship at the VA Medical Center in Jackson, Mississippi. While he watched the veteran detail symptoms, explain family history, and list a litany of other facts about his past and present, De Nadai had an idea.

“How do we automatically turn this info into something doctors can use in helping patients,” De Nadai wondered. “Better yet, how can we use it to get treatment right the first time?”

De Nadai is part of a talented cohort of Texas State professors leading the emerging field of computational medicine. With a slew of data-driven projects, professors of health administration, computer science, psychology and engineering are working together to make the world a healthier place.

“It’s an amazing, cross-disciplinary effort,” De Nadai says. “We have all of these professors coming together to use big data, and that’s pretty exciting.”

One cross-disciplinary project, led by professor Larry Fulton, employs machine learning to analyze MRI scans and predict Alzheimer’s, thereby allowing doctors to intervene and possibly even prevent the illness before it arises. Machine learning is only possible because of data, he says.

“This is the future of healthcare,” Fulton says. “Data is helping us get so much better at diagnostics and helping both doctors and patients. At the end of the day, all of our work is patient-focused and patient-centered.”

Figure 1: Percent of non-federal acute care hospitals that use their EHR data for at least one of the ten specified measures of hospital processes to inform clinical practice, 2015 – 2017.

Meanwhile, De Nadai is using data to help patients struggling with anxiety, OCD, PTSD, and other afflictions get on the right treatment path as early as possible. A combination of psychology and computer science, computational medicine has emerged as one of the most consequential fields of study.

“It’s really exciting, because there’s no single standard for what we’re doing right now,” he says. “We’re creating the standard that will help thousands and thousands of people.”

Professor De Nadai in his Texas State University office, comparing brain scans of patients suffering with obsessive compulsive disorder (OCD). Jeff Wilson

When De Nadai started his academic career, no one had ever heard of “computational medicine.”

“It didn’t have a name until relatively recently,” he says. “I started off in computer science, but I was interested in what computation can do to improve human welfare. I realized psychology and computer science are more related than we think.”

Specifically, De Nadai realized that computation can have a wide-ranging impact in mental health.

“Predicting what people are going to do is very difficult, and for this reason mental health data are some of the trickiest to work with,” he says. If the data was clean, he thought, then treatments would improve, and people could be happier and healthier.

“That’s been my biggest goal since I was twenty,” the professor says. “We may not need new medicines to improve lives. Instead, let’s use the tools that are already at our disposal to get people on the right path.”

Figure 2: How Much Data Is Created Every Day in 2020?

His passion for improving health inspired De Nadai to change his focus from computer science to psychology. After graduating from the University of Georgia, he earned a master’s in psychology from Stephen F. Austin State University, then a master’s in clinical psychology from the University of South Florida, where he would also earn his doctorate. Along the way, he has led several projects that merge his penchant for data with his passion for improving health, including an initiative involving multiple sclerosis (MS).

“When someone receives an MS diagnosis, they don’t know what’s going to happen next,” De Nadai says. “It could progress in one year, or fifteen years, and you have to plan your life very differently if it’s one or fifteen.”

De Nadai’s ongoing MS project compiles data on diagnoses and symptoms into a dashboard, allowing patients and doctors to predict what happens next based on their age, symptoms and other factors. Ultimately, doctors will be able to use this dashboard to practice “intervention,” identifying an illness before it arises and helping patients start the right course of treatment as early as possible.

“I don’t want people to live with ambiguity,” De Nadai says. “When a mother or father or grandparent gets a diagnosis, I want them to know what the rest of their life will look like, so they can live that life with their family as best they can.”

Figure 3: Percent of non-federal acute care hospitals that use their EHR data to perform each process that informs clinical practice, 2015-2017.

Eric Storch, Ph.D., has seen firsthand how data can improve lives. The experienced medical professional is a licensed psychologist and a professor at Baylor College of Medicine with a particular interest in anxiety disorders.

“In psychiatry, there really aren’t blood tests or biomarkers that clearly delineate who has a problem and what those outcomes are,” Storch says. “What Alex is doing is at the forefront of trying to tap into that and create those markers, and the result is a much more personalized approach to treatment.”

Storch has collaborated with De Nadai on several projects, including some that have used data to help treat people with anxiety and obsessive-compulsive disorder. The psychologist believes data is forging the next frontier of medicine, and with De Nadai at the helm, there’s no telling what breakthroughs are possible.

“If you think of early work by Sigmund Freud, it’s like that,” Storch says. “This is setting the foundation for medicine for years to come.”

De Nadai is currently hard at work on a variety of projects, including one that leverages data to predict depression and substance abuse, both of which may become more prevalent in a post-pandemic world. Smartphones can determine moving patterns—when you wake up, how often you move around or exercise—and, in turn, those patterns can indicate your susceptibility to depression and substance abuse. Like many of De Nadai’s projects, the professor is striving to prevent ailments before they complicate people’s lives.

“That’s the next big thing in medicine,” Storch says. “Predicting who is going to have what problem, what that problem can look like, and how we can treat it.”

And there’s no better place for De Nadai’s work than Texas State University. Texas is a nationwide leader in data science and computational medicine, and Texas State recently partnered with the tech company AMD to launch a project aimed at curbing the spread of COVID-19.

“Data science is great at merging lots of small signals and giving us something really meaningful,” De Nadai says. “If people can track changes in temperature and changes in heart rate for two to three weeks, then we know who has to isolate and who is safe.”

Of course, this is De Nadai’s first crack at a pandemic. Nevertheless, Stoch believes De Nadai can accomplish just about anything.

There’s no single standard for what we’re doing right now. We’re creating the standard that will help thousands and thousands of people.

“There are plenty of statistical whizzes who, by their trade, make the world a better place,” Storch says. “But not everyone is as personable as Alex. And that’s a huge asset. Because he cares, he sees the human behind the numbers, and he’s always thinking of how to improve that person’s life.”

De Nadai is humble about his own accomplishments, but he agrees with Storch: The person behind the data should always be top of mind.

“I want to see people get the right treatment for the first time,” he says. “Many suffer in silence, and I never want that to happen.”

Until he met De Nadai, that veteran-patient was suffering in silence. Often, he used substances to cope with his anxiety. By paying attention to the veteran’s daily habits and symptoms, De Nadai realized this man needed to expose himself to what gave him anxiety. So, the veteran started small, visiting a gas station every now and then, sometimes just to walk around. Eventually, he overcame his fear.

“Sometimes it’s just about practice, and taking those small steps,” De Nadai says. “That’s kind of what data is, too: Small parts of a big picture.”


Fig. 1 ONC/American Hospital Association (AHA), AHA Annual Survey Information Technology Supplement: 2015-2017. Note: The sample consists of 3,599 non-federal acute care hospitals.

Fig. 2

Fig. 3 ONC/American Hospital Association (AHA), AHA Annual Survey Information Technology Supplement: 2015-2017. Note: *Significantly higher than the previous year (p<0.05). The sample consists of 3,599 non-federal acute care hospitals.