Training Deep Learning Models to Work on Multiple Devices by Cross-Domain Learning with No Additional Annotations

In this paper, recently published in the journal Ophthalmology, Yue Wu and his co-authors developed a unique method for training a model to identify the retinal layers on optical coherence tomography images, with the key feature that the model works successfully on optical coherence tomography (OCT) images from a different OCT device than the one that was used to obtain the training data. Although the field of deep learning image analysis is rapidly advancing, one major roadblock is the problem of domain shift, where models trained on a particular dataset may experience significant performance degradation when applied to slightly different datasets from different hospitals, imaging protocols, or device manufacturers.

Continue reading "Training Deep Learning Models to Work on Multiple Devices by Cross-Domain Learning with No Additional Annotations"

Lee Lab launches AI-READI, a unique data generation project

AI-READI is part of the Bridge to Artificial Intelligence (Bridge2AI) program, a new initiative by the National Institutes of Health to expand the use of artificial intelligence in biomedical and behavioral research

The National Institutes of Health (NIH) Common Fund has launched the Bridge2AI program to accelerate the widespread use of artificial intelligence (AI) by the biomedical and behavioral research communities. The Bridge2AI program will be assembling team members from diverse disciplines and backgrounds to generate tools, resources, and richly detailed data that are responsive to AI approaches. Through extensive collaboration across projects, Bridge2AI researchers will create guidance and standards for the development of ethically sourced, state-of-the-art, AI-ready data sets that have the potential to help solve some of the most pressing challenges in human health. 

NIH has issued four awards for data generation projects to generate new biomedical and behavioral data sets ready to be used for developing AI technologies, and three awards to create a Bridge Center to integrate activities and knowledge across data generation projects while disseminating products, best-practices, and training materials. 

Cecilia and Aaron Lee will be the co-principal investigators of one of the the data generation projects, AI Ready and Equitable Atlas for Diabetes Insights (AI-READI). The goal of AI-READI is to develop a multi-modal atlas of type 2 diabetes mellitus (T2DM) by collecting data from a diverse population, while simultaneously creating a roadmap for ethical and equitable research that focuses on the diversity of the research participants as well as the workforce involved at all stages of the research process (collection, curation, and analysis). The AI-READI dataset will collect a range of health data from more than 4000 participants with diverse racial/ethnic backgrounds who represent all stages of T2DM, allowing researchers to better investigate health outcomes associated with T2DM in previously understudied populations, many of which have been significantly impacted by this disease. The project is structured with cross-disciplinary modules that focus on different aspects of the project, including data collection, team building, ethical oversight, development of a skilled diverse workforce, and the creation of data collection tools and standards. 

Read the UW Newsroom article about this exciting initiative here: UW Medicine will lead study arm of national AI initiative

To learn more about the Bridge2AI program, visit the Musings from the Mezzanine blog from the National Library of Medicine, watch this video about the Bridge2AI program, and read the NIH press release.

Policy-Driven, Multimodal Deep Learning for Predicting Visual Fields from the Optic Disc and OCT Imaging

In a new paper published in the journal Ophthalmology, Lee Lab members Yuka Kihara and Aaron Lee report the development of a novel deep learning approach to address a key problem in glaucoma research and treatment - predicting visual function from structural changes detected on optical coherence tomography (OCT). Physicians rely on visual field testing to assess whether glaucoma patients will progress to more severe disease, but the test is notoriously challenging for patients and the results can be inconsistent. Advanced retinal imaging such as OCT can reveal key structural changes in the retina and optic nerve, but previous methods for linking structural changes to visual function outcomes have been challenging. This approach provides a fully automated method for improving the prediction of standard automated perimetry sensitivity directly from structural imaging data and also enables artificial intelligence–derived structure–function mapping.

Continue reading "Policy-Driven, Multimodal Deep Learning for Predicting Visual Fields from the Optic Disc and OCT Imaging"

Association Between Cataract Extraction and Development of Dementia

In a new paper published in JAMA Internal Medicine, Dr. Cecilia Lee and her colleagues report exciting findings related to the association between cataract surgery and dementia risk. Using data from Adult Changes in Thought, an ongoing longitudinal study following 5000+ older adults for the development of Alzheimer disease and other dementias, they compared outcomes of participants with cataract who had surgery to those who did not. Participants who underwent cataract surgery had nearly 30% lower risk of developing dementia from any cause compared with those who did not, even after controlling for many health-related confounders and potential sources of biases.

Continue reading "Association Between Cataract Extraction and Development of Dementia"

Lee Lab Featured in Gates Notes

Several years ago, Bill Gates partnered with the Alzheimer’s Drug Discovery Foundation to develop a philanthropic fund called the Diagnostics Accelerator. Our own Dr. Cecilia Lee is one of the award recipients of this funding effort to discover and develop new diagnostic tests for Alzheimer's disease. Dr. Lee's lab is exploring retinal imaging techniques to identify early signs of Alzheimer’s, including by using artificial intelligence to analyze the imaging and potentially find irregularities that are invisible to the human eye.  Bill Gates featured her research on developing retinal biomarkers for Alzheimer's disease in a recent Gates Notes blog post. Watch the video of Bill Gates describing this intiative and Dr. Lee discussing her research here.