Technological advancements in medical imaging/probing systems (e.g., MRI, CT, Ultrasound, PET, Diffuse optical imaging, Electrophysiology, etc.) resulted in the exponential data growth both in quantity and complexity (thousands to billions of data points per patient). Despite such comprehensive information yields great promise for driving more precise medicine (e.g., diagnostic biomarkers, treatment assessment, intervention-guidance, etc.), such promise is currently compromised by the lack of sufficient computational analysis methods that can fully utilize such massive complex data. Consequently, current methods rely primarily on some form of dimensionality reduction or a simplified aggregation of this rich multi-dimensional, potentially multi-million point, data into simple statistical descriptives (e.g., total sum, maximum, average, etc.) that results in substantial data underutilization of such powerful source of medical decision-making information. To address this unmet need, we develop fully scalable mathematical methods that derive unique quantitative data signatures that utilize the entire data dimensionality of any point-distributed imaging data field. Our techniques are based, in part, on developing and utilizing novel computational stochastic-based data quantification for encoding local and global high-dimensional data field associations throughout the organ or space of interest (e.g., heart, lungs, brain, etc.). Example data fields include vector fields (e.g., color Doppler velocity, 4D flow three-directional velocity vectors), tensor field (e.g., DTI) or scalar field (e.g., CT, PET, MRI, T1 mapping, T2 mapping, fMRI, Angiography, etc.). Importantly, we are interested in developing techniques that are scalable to any dimension (e.g. 1D, 2D, 3D, 4D, 5D,… ND; N is any positive integer number > 0). We aim to utilize these advanced computational techniques to derive unique patient-specific physiological (e.g., cardiovascular and neurovascular blood flow, functional activity) and structural (e.g., extracellular volumes, tumor composition, infarction structure, thrombus formation, structural deformations, etc.) signatures while exploiting the entire data dimensionality. Hence, permitting to identify precisely personalized biomarkers of anomalies in various diseases throughout the body.
Related Publications
- M.S.M. Elbaz, J. Baraboo, A. DiCarlo, D. Lee, D. Kim, R Arora, M Markl, P Greenland, and R Passman “Novel left atrial 4D hemodynamics signature index for comprehensive assessment of 4D flow alterations in atrial fibrillation and degree of flow restoration 6-month post-ablation”. Society for Cardiovascular Magnetic Resonance (SCMR) 24th Scientific Meeting, February 18-20, 2021 (Accepted)
- Elbaz MSM, Scott MB, Barker AJ, McCarthy P, Malaisrie C, Collins JD, Bonow RO, Carr J and Markl M. 4D Flow Hemodynamic Signatures: A novel technique for assessing hemodynamics in aortic valve disease – evaluation in 418 patients and controls. Society for Cardiovascular Magnetic Resonance (SCMR) 23rd Annual Meeting , Orlando, Florida,12-15 February, 2020
- Elbaz MSM, Scott MB, Barker AJ, McCarthy P, Malaisrie C, Collins JD, Bonow RO, Carr J and Markl M. Stochastic Flow Co-expression Signatures: A novel concept for volumetric 4D flow assessment with application to aortic valve disease. International Society of Magnetic Resonance in Medicine (ISMRM) 27th Annual Meeting , Montreal, Canada, 11-16 May 2019 (Three Awards: Summa Cum Laude Merit Award, Research Presentation Award from ISMRM Cardiac MR study group, Travel Award).
- Elbaz MSM, Scott MB, Barker AJ, McCarthy P, Malaisrie C, Collins JD, Bonow RO, Carr J and Markl M. 4D Hemodynamic Signatures: A novel concept for the evaluation of abnormal flow dynamics in aortic valve disease. Society for Magnetic Resonance Angiography (SMRA) 31st Annual Meeting , Nantes, France, 28-31 2019
Patents
Patent-pending: application# 62/892,234 : Mohammed S.M. Elbaz, Michael Markl “Co-Expression Signatures Method for Quantification of Physiological and Structural Data”.
Funding
–AHA 20TPA35490311 (PI: Elbaz) (1/1/2021 – 12/31/2023)
American Heart Association (AHA) Transformational Project Award
“4D Hemodynamic Signatures of Atrial Fibrillation, Ablation Efficacy and Risk of Stroke”
–Sylvia Wolff Research Fund (PI: Elbaz) (01/01/2020-12/31/2021)
Northwestern Memorial Hospital
project: “Left Atrial 4D Hemodynamic Signatures of Atrial Fibrillation”