Our mission in the HITI lab is to apply machine learning to big problems in healthcare, including developing diagnostic and predictive imaging models, validation of existing models, and a focus on the fairness and explainability of AI.
Our projects encompass a wide range of topics from computer vision to natural language processing to radiology workflow. We strive to base each project on a real clinical need and end with the implementation of our products into a clinical workflow.
A large, multiinstutional collaborative paper “Evaluation of combined artificial intelligence and radiologist assessment to interpret screening mammograms” was published in JAMA. This work evaluates several deep learning models’ performance on detecting cancer in...
We are pleaseed to announch a new collaboration with Stanford University and University of California, San Francisco to automatically infer risk score for hepatocellular carcinoma (HCC) from MRI images.
Our paper “Fusion of medical imaging and electronic health records using deep learning: a systematic review and implementation guidelines“ provides an overview of different types of fusion models that can merge clinical and imaging data to imrpove diagnostic and...
Our paper ““Current clinical applications of #AI in radiology and their best supporting evidence “ was just published in the Journal of the American College of Radiology
The HITI lab is happy to announce a collaboration with U.K. based Kheiron Medical to test and improve performance of their breast cancer detection model – Mia. Our diverse patient population will allow us to ensure that these types of models serve all populations...
The HITI lab is proud to be one of the 7 pilot sites for the American College of Radiology AI Lab to work on federated learning. This work will improve robustness of machine learning models by allowing training across multiple institutions to improve generalizability
Type 1 Diabetes Management With Technology: Patterns of Utilization and Effects on Glucose Control Using Real-World Evidence Journal Article
In: Clinical Diabetes, 2021.
Fully Integrated Analog Machine Learning Classifier Using Custom Activation Function for Low Resolution Image Classification Journal Article
In: IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 68, no. 3, pp. 1023–1033, 2021.
Opportunistic Assessment of Ischemic Heart Disease Risk Using Abdominopelvic Computed Tomography and Medical Record Data: a Multimodal Explainable Artificial Intelligence Approach Journal Article
In: medRxiv, 2021.
PAC-Bayes Bounds on Variational Tempered Posteriors for Markov Models Journal Article
In: arXiv preprint arXiv:2101.05197, 2021.