The Quantitative Imaging Analysis Program provides comprehensive services for research, pre-clinical and clinical imaging studies, including protocol development, image acquisition consulting, quality control, image analysis, statistical analysis, visualization and result reporting. Briefly, our mission is to:
- Provide image-based quantitative analysis services for both clinical and research studies, using advanced computer-aided methods
- Promote intra- and inter-departmental multidisciplinary research collaborations
- Develop novel machine learning techniques for image analysis, to recognize and characterize normal and abnormal regions of the heart, brain and other organs
- Enhance and quantify the information derived from complex multidimensional cardiac images
One focus of research in the Quantitative Imaging Analysis Program is to measure the myocardial blood flow (MBF) of first-pass perfusion at magnetic resonance imaging. Unlike conventional qualitative and semiquantitative methods, our approach generates fully quantitative measurements by using Fermi function deconvolution to determine the absolute myocardial perfusion value. Meanwhile, a dual-resolution perfusion sequence is employed to provide accurate assessment of arterial input function, which is essential in absolute MBF quantification.
Another research focus is on brain image analysis. We provide services on brain structural and functional image processing, statistical analysis and visualization. Briefly, we are working on structural, functional and diffusion MR images on the following topics: tissue segmentation, spatial normalization, brain labeling, cortical surface analysis, voxel-based morphometry, resting- and task-state functional connectivity, diffusion-tract-based spatial statistics, fiber tractography, brain connectome, and computer-aided disease diagnosis and prediction.
Additionally, our research emphasis is on:
- Automated derivation of imaging measures from noninvasive cardiac images
- Clinical implementation of novel automated computer processing algorithms
- Machine learning integration of imaging biomarkers to predict risk
Working environment of cardiac image processing for scar signal quantification.
The Quantitative Imaging Analysis Program brings together a team of scientists, biomedical engineers, neurologists and psychiatrists, as well as developing collaborations with other specialty areas. The research also involves collaborations with other institutions, including UCLA.