Article Text
PDF
Abstract
Background The magnitude of both healthcare data and artificial intelligence (AI) model complexity limits their use on traditional computational infrastructure. AI Labs for Patient Safety (ALPS) is a novel infrastructure designed to apply cutting-edge AI tools to quality improvement (QI) projects. ALPS uses a secure high-performance computing environment capable of storing and analyzing large amounts of healthcare data based on a common data model. This self-service approach mitigates computation and administrative burden in streamlining QI projects, balancing the needs of researchers and individuals conducting QI.
Objectives Illustrate ALPS’s capability to streamline QI.
Methods Figure 1 shows ALPS’s process, where the QI community can develop AI models in a protected health information (PHI)-free environment. Based on a project idea, participants can select patient and healthcare delivery data from a self-service database (Health Data Sciences and Informatics’ Observational Medical Outcomes Partnership (OMOP)). The selected data are securely transferred to HiPerGator, a high-performance, HITRUST-certified, GPU-accelerated computing environment for preprocessing and analysis of data containing PHI. Projects are managed using version control and machine learning operations (MLOps) platforms to ensure sustainable use of analyses over time.
Results ALPS can host over 169 million records (table 1) in a secure high-performance computing environment, ready for analysis. The infrastructure can arrange data in both flat files and in a graph database, reflecting the connection between caregivers and services to patient factors and health care delivery.
Conclusions ALPS provides a secure, efficient, and powerful infrastructure for AI-driven QI initiatives in health care, enabling rapid translation of ideas into solutions for patient safety. ALPS has the potential to revolutionize the way QI projects are conducted by integrating secure high-performance computing, accessible databases, and balancing the needs of researchers and those conducting QI work. This innovative infrastructure can improve patient outcomes and transform the healthcare experience.
ALPS process
ALPS data infrastructure
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.