AI Enabled Technology

PCCI’s AIM-UP predictive models drive augmented intelligence for clinicians and personalized patient engagement. The core technologies powering these models include Isthmus™, our interoperable, cloud-based digital data environment; ISLET™, a web-based predictive model visualization tool that seamlessly integrates model results into existing EHR systems; and our automated AI model monitoring technology.

AI Tools In Action: Expanding the Foundational Tools Powering Our AI Solutions

PCCI has created a healthcare-focused, secure, and private digital platform called Isthmus™, where healthcare data can be safely stored and analyzed using cloud technology and industry-standard tools. The platform is deployed within a hospital’s existing electronic environment to protect personal health data from being exposed to the outside world. This protected environment provides confidentiality and security for sensitive patient information, while enabling advanced analytics and modeling capabilities. Isthmus serves as a research data hub via open-source programming, such as R and Python, along with other standardized data models (OMOP) and software tools (ATLAS) for data studies and outcomes sharing across organizations.

In 2024, we announced PCCI’s collaboration with University Health (UH) Transplant Institute to enhance its digital data environment through creation of a first-of-its kind Transplant Quality and Research DataMart integrating OHDSI, the OMOP Common Data Model, and the ATLAS software applications into the UH Isthmus ecosystem. This will enable University Health teams’ access to the open community data standards, open-source software, NMDoH data (through PCCI’s CVC), and ways to apply scientific best practices to generate reliable clinical evidence and enable deeper understanding of the context and complexities of the social barriers to health, access, and well-being of Transplant Institute patients.

Visualizing Critical Data using ISLET

Supplementing Isthmus, PCCI developed ISLET™, a web-based, rich, predictive model visualization tool that seamlessly integrates model results into existing systems, such as EHR or case-management systems, making it easy for clinicians to access and utilize the insights generated by the model. It requires no additional logins or workflow changes. We believe that models shouldn’t be “Black Boxes” and ISLET was developed to allow us to prioritize transparency by providing clear explanations of the important, actionable factors that influence a model’s predictions.

As an early use case, ISLET has been integrated into the PCCI Inpatient (IP) Sepsis predictive model at Parkland, predicting the risk of a patient becoming septic in real time. The integration of ISLET allows care teams to easily view and understand model output and actionable risk factors to ensure care teams don’t miss (or misinterpret) key data insights. This use of data reduces risk and supports more informed decision-making in patient care, helping to save lives.

ISLET’s use with Parkland’s IP Sepsis model has been recently expanded to 10 new Parkland hospital units, and its capabilities are also being incorporated into several emerging programs where visualization of patient-specific predictive model data can be leveraged to improve patient outcomes.