The Computer and Information Science (CIS) Department at the University of Pennsylvania tackles problems in machine learning, artificial intelligence (AI), data privacy, sensors and the Internet of Things, cloud and distributed computing, computational social science, and human-computer interaction. CIS teams often collaborate with other groups within Penn Medicine to work on the critical health informatics problems of today and the future, and to design and implement impactful solutions.
Within the CIS department are several different groups doing applied health informatics work.
The PRECISE group conducts research on the Internet-of-things and medical cyber-physical systems. PRECISE aims to provide safe health care solutions that integrate medical devices, patients, clinicians, and personalized automation to improve health outcomes.
The PRECISE team has developed a medical device integration platform, VitalCore, that supports access to medical device data in real-time to address issues that can arise when vendor-specific applications on medical devices stop working. During the time the systems are down, medical professionals have to manually document patient data in the electronic health record (EHR), which creates delays and leaves it vulnerable to errors, manipulation, and omissions.
Internet-of-Medical-Things researchers in CIS seek to revolutionize health care through unprecedented access to real-time health data that will enable actionable clinical decision support and safe, autonomous medical systems.
A central theme of their research is the desire to design devices and alerting systems to support the ability of clinicians to make appropriate care decisions. For example, they have recently tackled the problem of strokes that occur in the hospital, which account for up to 17% of all strokes in the US each year and that are associated with significant mortality and increased cost.
This team created an automated, continuous, and non-invasive monitor to rapidly identify symptoms of stroke in hospitalized patients and enable more prompt use of proven interventions to save lives and decrease costs.
The World Well-Being Project is pioneering scientific techniques for measuring the determinants of psychological well-being and physical health from social media and text message language, combined with cellphone sensor data and surveys.
CIS professors are developing scalable machine learning methods for data mining and text mining, including deep learning methods for natural language processing. These methods support projects ranging from diagnosing depression using data from SMS and Facebook messages, to analyzing Twitter to create targeted messages addressing vaccine hesitancy.
This team works with a broad range of psychologists and medical researchers to shed new light on the psychosocial processes that affect health and happiness, and to explore the potential for their unobtrusive well-being measures to supplement—and in part replace—expensive survey methods.
The Penn Database Group has a long legacy of collaborating with other groups in health care.
Professors in the Departments of Bioengineering; Biostatistics, Epidemiology, and Informatics; and CIS develop tools and platforms for large-scale data sharing across neuroscience and many other domains, addressing key questions in data reuse, trust and provenance, and semantic integration. These informatics tools, exemplified by the open-source Pennsieve platform, have thousands of users in areas such as neuroscience, novel medical devices, cancer treatment, and more.
Penn’s CIS department has collaborated with other groups in the Center for Applied Health Informatics including the Center for Health Care Innovation (CHCI). In one case, CHCI took several data sharing platforms and algorithms related to epilepsy and helped CIS teams prototype proofs-of-concept for how the same platform could be integrated into a clinical diagnostics setting. This ultimately allowed CIS to create a startup, Blackfynn Inc., that uses data and technology to help develop treatments for neurodegenerative diseases.
A key challenge in health care today is how best to capture ever-increasing amounts of patient health and medical data — whether from wearable devices, social media, or from clinical notes — and then integrate it into a single, cohesive picture that informs personalized treatment. Penn CIS is excited to join the Center for Applied Health Informatics and to work with collaborators across Penn Medicine to develop new artificial intelligence, data management, and human-computer interaction tools to address this and other challenges going forward.