Over more than a century, advancements in medical imaging have resulted in several powerful imaging modalities (MRI, PET, CT, Echocardiography, Ultrasound, etc). However, each modality is usually suited for imaging certain types of structures but not as efficient for others (e.g. CT can be suited for lungs and for bone imaging, where MRI can be more suited for imaging the heart and its function). Over the past three decades, several methodological developments have been made towards medical image fusion i.e. combining multiple images from multiple modalities to deduce complementing imaging information to increase the clinical applicability and enable more precise diagnostics markers. However, imaging fusion is primarily based on the registration of images of the same structure (e.g. heart imaged by PET to be fused with that imaged by MRI). Therefore, registration-derived image fusion is fundamentally limited to complementing data of the same structure/organ and cannot be directly applied if the data comes from different organs/structures. Here we work on developing profoundly novel computational registration-free concepts and techniques, crossing the registration barrier, to enable integrating imaging data, at the voxel-level, from different modalities to study associations between different organs or structures. We will utilize these techniques to reveal novel insights into the pathophysiology of the complex coupling between multi-organs in various diseases (e.g. Lung-heart coupling in pulmonary hypertension and brain-heart connection in AF-induced stroke, and impact of tumors in various regions of the body on cardiovascular function).
Patents
- Patent-pending: application# 62/892,234: Mohammed S.M. Elbaz, Michael Markl “Co-Expression Signatures Method for Quantification of Physiological and Structural Data”.