WVU builds a bridge to better health using AI

Quality healthcare transcends the medical profession, as evidenced by a new project led by West Virginia University that includes not only health experts but engineers, a physicist, a lawyer, and a business data analyst. A multidisciplinary team at WVU will embark on a project that will leverage artificial intelligence and digital health – which includes data from mobile devices and wearables – to address rising healthcare costs, the expansion of the nation’s elderly population and health disparities.  CREDIT Aira Burkhart/WVU

“Bridges in Digital Health,” which recently received $3 million from the National Science Foundation, hopes to address the combination of rising healthcare costs, the expansion of the nation’s elderly population, and health disparities, particularly in rural communities, through advances in digital health and artificial intelligence, and training the next generation of professionals to develop and deploy such advances.  

Digital health is a rapidly growing field that involves clinical and biomedical data including prescriptions, medical images, ultrasound videos, electronic health records, and data from mobile devices and wearables, such as Fitbit, said Donald Adjeroh, lead investigator of the project and professor and associate chair in the Lane Department of Computer Science and Electrical Engineering.

“Two of our pathway themes in the project are focused on the use of data science and A.I. on two key areas in healthcare: namely, cardiovascular health (analysis of cardiac images, especially, echocardiograms), and genomics (analysis and functional annotation of long non-coding ribonucleic acids – a type of RNA - and their role in disease prediction and prognosis),” Adjeroh said. 

“Apart from traditional electronic health records, our health data will come from different sources and devices, including wearable devices such as hand-held mobile cardiac ultrasound devices, or pocket EKG monitors, low-cost mobile activity monitors, Fitbits, smartwatches, social media, etc. Such low-cost wearable devices and data sources are important in collecting health-related data from individuals in rural areas, and outside the hospital setting, important for preventive care.” 

Adjeroh noted that various recent reports, including results from WVU labs, document the success stories of A.I. techniques on health problems including breast cancer detection, diagnosing eye diseases, reading cardiac ultrasound images, early prediction of acute kidney failure, predicting adverse drug events, and visualization of neuronal structures in the brain.

“These methods have shown performance that is close to human performance, and at times outperform human professionals on some of these tasks,” he said. 

The NSF funding will help establish a new graduate education and traineeship model to prepare students to work in collaborative teams to develop and apply data science and A.I. techniques in addressing digital health issues. The project anticipates training 24 funded and 40 unfunded masters and doctoral students from different disciplines including engineering, computer science, medicine, health sciences, physical sciences, and economics.

Gay Stewart, a physicist who directs the WVU Center for Excellence in STEM Education, is one of the project’s co-investigators. 

“My focus is on improving access to STEM careers for West Virginians,” Stewart said. “Much of my focus has been on building the pipeline earlier, but traditional graduate programs do not provide the ability to work across disciplinary silos deeply enough to make the advances we need. ‘Bridges’ will address these challenges, by preparing trainees to work effectively in transdisciplinary teams that develop leading technology-driven solutions to challenging problems in DH, especially in rural communities.”

Stewart said the team will recruit participants from underserved groups – such as rural and first-generation students - in STEM.

“First-generation students tend to graduate college in STEM at lower rates than their peers and are less likely to pursue graduate studies,” she said. “Yet, we need their voices in this important work. I envision a much stronger motivation to pursue advanced studies when students can see the potential for significant impact on their families and communities.”

Dr. Michael Ruppert, another co-investigator on the project, explained the role of the research from a biomedical standpoint. 

“One of the stumbling blocks for biomedical researchers is that very diverse skill sets are required to develop new knowledge by analyzing large datasets such as clinical data,” said Ruppert, Jo, and Ben Statler chair of Breast Cancer Research at the WVU Cancer Institute and professor of biochemistry in the School of Medicine. “For example, you have to be good at biomedicine, which often involves moving molecules around the lab, and you also have to be able to move very large digital datasets around as well. The goal is to cross-train to generate students with all the necessary skill sets.”

Other members of the research team are Gianfranco Doretto, computer science, and electrical engineering; Dr. Partho Sengupta, of Rutgers University; Michael Humicrobiology, immunology and cell biologyValarie BlakelawBrad Pricemanagement information systemsNasser NasrabadiXin LiDon McLaughlin, and Brian Powell, all of computer science and electrical engineering; Michael Schaller, biochemistry; and Cathy MortonHealth Sciences and Technology Academy

NYU project trains Kenyan experts to bring social determinants to bear on modeling health outcomes

A data-science training program for equipping leaders to support the improvement of health outcomes in Kenya, led by  a team from NYU, Brown University, and Moi University in Kenya, was chosen as one of 19 initiatives funded by The National Institutes of Health (NIH) under its new Harnessing Data Science for Health Discovery and Innovation in Africa (DS-I Africa) program.

The $1.7 million award, part of the NIH’s mission to advance data science, catalyze innovation and spur health discoveries across Africa, establishes a consortium consisting of a data science platform and coordinating center, seven research hubs, seven data science research training programs, and four projects focused on studying the ethical, legal and social implications of data science research. Rumi Chunara, associate professor of computer science and engineering and biostatistics at the NYU Tandon School of Engineering and NYU School of Global Public Health (NYU GPH)

The main principal investigator for the NYU-Moi Data Science for Social Determinants Training Program (DSSD) is Rumi Chunara, associate professor of computer science and engineering and biostatistics at the NYU Tandon School of Engineering and NYU School of Global Public Health (NYU GPH). The DSSD training program represents a significant opportunity to leverage NYU's strengths in data science, machine learning and artificial intelligence in a collaborative fashion with global partners to improve data science capacity, specifically for health. 

The goal of the project is to develop future leaders in data science who are equipped to gather and analyze data to better leverage deep and rich surveys, as well as internet and other digitized data sources that can help the collaborators capture information on the social determinants of health. The project, includes researchers at NYU Courant, NYU GPH, NYU Wagner, the Center for Urban Science and Progress (CUSP), the NYU Center for Data Science, and the NYU Grossman School of Medicine. It constitutes an extension into a real-world training program of Chunara’s previous work on incorporating social determinants into predictive modeling for individual health outcomes.

“To develop best practices in treatment and analytics for health outcomes, social determinants must be part of the data mix because they provide context on broader forces impinging on the health both of individuals and for communities. I want to thank the NIH for their acknowledgment of this." said Chunara. "Besides advancing local efforts in Kenya in data science and health, we also envision our program will augment global knowledge on data science practices.” 

DSSD’s design will rapidly expand the local base of expertise via curriculum development, resulting in two Ph.D. (4-year training) and a total of six postdoctoral (2-year) and faculty (12-14 month) trainees, who will study at NYU. Additionally, eight masters and two Ph.D. trainees will commence or complete training (2-year and 4-year training, respectively) through newly developed data science tracks at Moi University.

Connecting with data science industries and organizations with a presence in Kenya, including IBM, Deep Learning Indaba, DataKind, AI.Kenya and Aga Khan University Nairobi and Karachi, will create intellectual meeting spaces for a variety of talented trainees from both data science and health backgrounds, to propel and sustainably advance the field’s capacity in Kenyan institutions as well as the DS-I consortium.

UC San Diego Health initiative translates clinical data into novel personalized therapies for breast cancer patients

In an important step that could help answer research questions about breast cancer and develop more personalized solutions for patients, philanthropists Richard and Carol Dean Hertzberg have committed $2.1 million to develop and maintain the Dean-Hertzberg Breast Cancer Database System (BCDS) at Moores Cancer Center at UC San Diego Health. The gift will support the work of Anne Wallace, MD, director of the Comprehensive Breast Health Center at UC San Diego Health, and her collaborators at Moores Cancer Center.

The interactive database will further UC San Diego Health’s efforts to advance the understanding of breast disease and develop new treatments. The BCDS will combine biological, biographical, and demographic data in novel ways that will allow researchers to study breast cancers with similar clinical features, as well as rare subtypes.  

“I am excited about the BCDS’s potential to bring research collaborators together with practicing providers to use advanced technologies, data, and knowledge to find better ways to improve each patient’s experience, based on their specific breast cancer,” Wallace said. “I am grateful to Carol and Dick for helping us launch this project.”

The Hertzbergs’ generosity has enabled Wallace and colleagues to begin collaborating with the laboratory of Thomas J. Kipps, MD, Ph.D., deputy director of research operations for UC San Diego Moores Cancer Center. Wallace and Kipps will use the system as a flagship for data analysis and accessibility. 

Previously, the Hertzbergs contributed two gifts of $100,000 and $200,000 to help create the BCDS. Their latest gift ($1.8 million) brings the BCDS initiative fully to life and includes the addition of a clinic data manager to support work.

“When we asked Dr. Wallace how we could help, she had a wish list of projects that could not be funded by traditional grant sources,” said Carol Hertzberg. “She described this project to us and we knew it was something we wanted to support. We are excited to see the impact that this collaboration will make for research and care.”