Multi-Vendor Multi-Site Novel Accelerated MRI Relaxometry

Overview

Currently the diagnosis of OA is based on radiographs and clinical findings, which is limited to detecting late-stage disease. There is a pressing, unmet clinical need for robust assessment of early changes in cartilage health. Towards this goal, extensive efforts are ongoing to develop quantitative MRI for cartilage matrix analysis. MR T1ρ and T2 relaxation times have shown to be promising imaging biomarkers for early cartilage degeneration and prediction of disease progression. However, many challenges remain to clinically applying these techniques, including lack of standardized acquisition and quantification methods, and long acquisition times. The study aims to develop novel, fast and reproducible MR T1ρ and T2 relaxation time imaging methods on MR systems from multiple vendors and establish a platform for standardization and cross validation of these measures as a tool for clinical trials using such techniques. Following method validation, patients at risk for osteoarthritis will be tested.

The significance of this study is twofold. First, one typical hurdle in applying advanced quantitative MRI to clinical protocols is the long acquisition time. The proposed novel acceleration techniques will facilitate clinical translations of T1ρ and T2 imaging by reducing the time and cost to integrate the sequences into standard clinical practice, increasing patient comfortableness and reducing potential motion artifacts. Second, it is an essential but sometimes overlooked step to investigate quantification variability across sites and MR systems in order to validate MR imaging biomarkers and to apply the measures in large scale multi-vendor multi-site trials. Specifically, there is a lack of systematic evaluation of inter-vendor inter-site variability of T1ρ imaging even though it has been widely applied in neural imaging, liver imaging, cardiac imaging, oncology imaging, and musculoskeletal imaging. Furthermore, no commercial T1ρ phantoms are available with validated reference values. The proposed study is addressing these significant gaps. By implementing and cross-validating T1ρ and T2 imaging on MR systems from different vendors, the outcomes from the proposed study (dedicated calibration phantoms, fast and standardized acquisition and analysis protocols) will provide the field with essential tools for future multi-vendor multi-site trials that will use these quantitative imaging techniques. In this proposal, the innovation and development in patients at risk for osteoarthritis will be tested. In this study, novel accelerated T1ρ and T2 imaging methods will be developed to systematically evaluate inter-vendor inter-site variation of these measures using dedicated T1ρ and T2 calibration phantoms (to be developed in this study) and traveling subjects, and demonstrate the feasibility in patients after ACL injury and reconstruction who are at risk of post-traumatic OA. The investigation has three Specific Aims: 1. \- Develop novel acceleration techniques for fast 3D T1ρ and T2 imaging. Despite the increasing availability of accelerated morphologic imaging using compressed sensing (CS) techniques, applications of CS to quantitative MRI are still very limited with the challenges of maintaining quantitative accuracy. A novel convex, model-based CS technique will be developed to take full advantage of the known model for T1ρ and T2 decay. 2. \- Develop a calibration phantom suitable for standardization of T1ρ and T2 measurements and implement acceleration techniques on MR systems of three major vendors (Siemens, GE and Philips). The phantom development and reference measurements will be in collaboration with the National Institute of Standards and Technology (NIST). The accelerated T1ρ and T2 sequences will be implemented on six MR systems at four sites from three vendors. Inter-site and inter-vendor variation of T1ρ and T2 will be quantified in phantoms and traveling human subjects. 3. \- Demonstrate the ability of the newly developed acceleration techniques to quantify cartilage degeneration longitudinally in a multi-vendor setting. Subjects with acute anterior cruciate ligament (ACL) tears, an established 'early OA' model, along with age-, gender-, and BMI-matched controls will be scanned on three MR systems using standard and accelerated T1ρ and T2 protocols at baseline and one-year. A novel atlas-based voxel-based relaxometry (VBR) analysis will be applied. Cross-sectional and longitudinal measures will be compared between standard and accelerated T1ρ and T2 protocols. Four groups of subjects will be recruited at CCF for the study. Group I: 30 volunteers (for Aim 1) The 30 volunteers from Group I will be recruited at CCF for evaluating differences between standard T1rho and T2 imaging vs accelerated T1rho and T2 imaging techniques that will be developed from this study. Group II: 8 traveling volunteers (for Aim 2) The 8 traveling volunteers in Group II will be recruited at CCF and be scanned at CCF, University of California San Francisco, University of Kentucky and Albert Einstein College of Medicine. Group III: 30 patients with acute ACL tears (for Aim 3) Group IV: 12 controls who are matched with Group III patients with age, sex and BMI (for Aim 3) Subjects in group III and IV will be recruited at CCF and scanned at baseline and 1-year at all of the three MR systems at CCF (Siemens, GE, Philips). Subjects of Group I will be recruited in year 1 and 2; Group II will be recruited in year 2 and 3, and Group III and IV will be recruited in year 3-5. Group II and IV share the same inclusion and exclusion criteria, therefore subjects can participate the study and serve as subjects within both groups. At two other sites involved in the study, University of California San Francisco (UCSF) and Albert Einstein College of Medicine at New York, 20 local volunteers will be recruited for sequence development in year 1 and 2.