E010 Reliability of automatic estimation of striatal volumes using synthetic MRI in Huntington’s disease
Year: 2024
Type: conference-abstract
Abstract: Background/Aims Synthetic MRI (SynMRI) offers rapid data acquisition (5 min scan), which could reduce MRI time while providing versatile tissue contrasts for Huntington’s disease (HD) assessments. Broad application requires validation to ensure precise depiction of brain structural changes (especially striatum), as these are used to monitor HD progression. Methods MRI scans were collected from 10 HD participants using synMRI (spatial-resolution 2.2 mm³) and conventional T1-weighted MRI (T1w-MRI) at two resolutions (1 mm³ and 2.2 mm³). Striatal volumes were evaluated using AssemblyNet. The reliability of volume estimations was assessed using root mean squared error (RMSE), using the conventional scan (1 mm³) as the reference. Correlations between striatal volumes and Unified HD Rating Scale (UHDRS) total motor scores and disease exposure (CAP score) were computed using Pearson’s correlation coefficient. Results A fourfold increase in RMSE for putamen volumes and a threefold increase for caudate volumes were revealed when comparing synMRI to T1w-MRI at 2.2 mm³. Furthermore, correlations between CAP score and volumes decreased by 3 and 9 percentage points for putamen and caudate, respectively, with synMRI. However, correlations between striatal volumes and total motor scores (TMS) were comparable for both T1w-MRI and synMRI figures 1–4. Conclusions While synMRI offers rapid MRI acquisition, we found increased error in volume estimation and reduced correlation with known disease markers. Given the pivotal role of striatal volume in HD studies, optimizations of synMRI (high-resolution, multi- contrast) are imperative before its adoption for disease monitoring and interventional assessment. The utility for other quantitative measures will be assessed in future studies.
Authors: Kilian Hett, Paula Trujillo, Tang Haiying, Daniel O Claassen
Institution: Vanderbilt University Medical Center
Language: English
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: gold
OpenAlex ID: W4402374662