UNIVERSITY PARK, Pa. — Two Penn State College of Information Sciences and Technology (IST) students are part of an interdisciplinary team working to create a health care application that can assist doctors with diagnosing schizophrenia in remote regions.
Gia Nguyen, a senior studying data sciences, and Maria Czura, a junior pursuing a degree in human-centered design and development, are contributing their skills and knowledge to the mental health screening tool SchizophrenAI. Their project is one of 10 selected to receive funding to develop a minimum viable product (MVP) in the 2022 Nittany AI Challenge.
“It is amazing to be able to develop SchizophrenAI in such an encouraging environment and have the opportunity to reach an even greater audience at this level of the competition,” Czura said.
The team is made up of 10 students from IST, the College of Engineering, the Eberly College of Science and the Smeal College of Business. Loc Phan, a sophomore studying physics, came up with the concept for SchizophrenAI after a personal struggle with finding proper mental health screening for his uncle in Vietnam. Phan’s uncle was eventually diagnosed with schizophrenia, but the long and arduous process to receive that diagnosis delayed treatment and added undue stress to his family.
When the Nittany AI competition challenged undergraduate students to tackle global issues through AI and machine learning solutions, Phan recruited a team to build a solution that could improve the diagnostic process.
“The baseline for diagnosing schizophrenia that doctors use is the DSM-5, but there is also a lot of subjective content that goes into that diagnosis,” Nguyen said. “That can result in misdiagnosis or delay the time that it takes to diagnose the patients, which can impact their quality of life.”
SchizophrenAI uses a machine learning algorithm that will take the individual patient’s data from a variety of cognitive tests. The resulting output could indicate whether a patient is at risk of having schizophrenia.
Schizophrenia is often referred to as a diagnosis of exclusion, but medical advancements have established additional substantive ways to identify this mental illness. The SchizophrenAI team hopes to build on these advancements in their solution.
They are currently working to create a way to interpret a patient's vision pattern using a flashing dot on their computer screen that will track their eye movement. This biomarker will be converted into a number on a scale that indicates how likely it is that the afflicted patient has schizophrenia. This testing model could be applied to other diagnostic tools, like ink blot testing, that is used by mental health professionals.
“It’s been shown that people with indicators of schizophrenia have different eye movements than other people would, which can aid in pointing out some things that maybe other people can miss,” Czura said.
Nguyen’s role on the team is to focus on forming connections between her team and individuals who could use this app in order to determine what features to include, but her IST background in applied data science prepared her to tackle one of the biggest challenges SchizophrenAI faced: layers of privacy protection that surround medical records, making data collection an arduous and complex task.
“I really used what I learned from my major classes; it was a big help,” Nguyen said. “We had to resort to open-source data, instead of just getting data from the psychiatrists or clinics, so I helped the artificial intelligence (AI) team with finding data and making sure it was clean.”
Czura assists in the app’s development, helping to determine useful features and then designing them. This work has taught her a lot about the differences in needs between patients in the United States and those in foreign countries.
“I've found a lot of interesting little nuances between Vietnam and the U.S. … like transportation,” Czura said. “In more rural areas, it's really difficult to get to a hospital and transportation either takes a long time or is expensive. There are rural areas in the U.S., too, but I found that you're not going to spend days traveling to get to the hospital. It was surprising to find that in some countries it's really a roadblock to be able to just go to the hospital.”
Nguyen is hoping that the technology she is helping to develop will bridge the gap between doctors and patients in need, which is one of the main reasons that she got involved with the Nittany AI Challenge.
“I want to take the knowledge I've learned, like coding, and use that to solve real-life problems,” Nguyen said.
The SchizophrenAI team is entering the minimum viable product (MVP) phase of the Nittany AI Challenge. By Aug. 11, the team must produce a product that is feasible, impactful and sophisticated enough to tackle the problem that they have outlined. Stay tuned to more information on the next phase of the challenge at nittanyai.psu.edu.
Editor's note: This story is informational in nature and should not be considered an endorsement of any product or application.