PIKSL

PIKSL is dedicated to harnessing artificial intelligence and image analysis techniques to make imaging a reality in precision medicine. This work aims to address a wide range of challenges in imaging informatics, medical image analysis, and signal processing, creating robust and scalable resources for clinical imaging and translational research. This involves close collaboration with bioinformatics, clinical, and engineering groups across Johns Hopkins.

PIKSL was created to bring multiple imaging modalities and artificial intelligence methods under one unified research goal: “Precision Imaging”. A subset of precision medicine, “precision imaging” involves reimagining medical image analysis at the individual level. By utilizing the vast corpus of imaging data available throughout the health system and beyond, we aim to develop robust, scalable methods of quantitative analysis, bringing big data to the standard of care. Our current work includes generative AI models for medical imaging, foundation models for image embedding, and processing methods to streamline analysis workflows in highly variable clinical data.

PIKSL is led by Dr. Blake Dewey, an Assistant Professor of Neurology at the Johns Hopkins School of Medicine. His work is at the intersection of medical imaging, signal and image processing, artificial intelligence, software engineering, and clinical research. Dr. Dewey leads PIKSL’s committment to open and team-oriented science, working to distribute code and model weights for projects alongside academic publication.