LiverMultiScan (LMSv3)

LiverMultiScan (LMSv3) is a software application that enhances non-invasive liver assessment by post-processing MR images to quantify key biomarkers like Proton Density Fat Fraction (PDFF), iron-corrected T1 (cT1), and T2*. It employs techniques such as automated liver segmentation and parametric mapping, including the LMS IDEAL and three-point Dixon methods for fat quantification, to provide clinicians with reproducible quantitative data intended to aid in liver disease diagnosis.

LiverMultiScan (LMSv3) is a standalone software application designed for non-invasive liver evaluation using magnetic resonance (MR) imaging data. Its primary purpose is to assist clinicians in diagnosis by generating, displaying, and reviewing 2D MR medical images and pixel maps of MR relaxation times. The software processes DICOM 3.0 compliant MR image datasets from compatible MR systems to visualize the liver's internal structure and derive quantitative physical parameters. Key functionalities include automated liver segmentation and region of interest (ROI) placements, enabling detailed assessment of selected liver regions for use as an aid to diagnosis. The methodology involves post-processing MR data to create parametric maps for Proton Density Fat Fraction (PDFF), T2*, T1, and iron-corrected T1 (cT1). LMSv3 offers flexibility in PDFF calculation, supporting both the LMS IDEAL method and the three-point Dixon methodology. This quantitative assessment provides measurements of liver triglyceride fat fraction, T2*, and cT1, which are compiled into a summary report, including histograms of segmented areas and PDFF intensities, for interpretation by specialists like radiologists and hepatologists. Performance testing on phantoms and in-vivo volunteer data demonstrated high repeatability and reproducibility for these measurements, with accuracy levels established for T1 (noting a consistent underestimation characteristic of MOLLI techniques), T2* and IDEAL PDFF (accurate over physiological ranges), and DIXON PDFF (relatively accurate, with minor deviations due to known fat bias).

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the performance of the LMS IDEAL PDFF methodology compare to the three-point Dixon methodology in various liver disease states, particularly concerning accuracy and sensitivity to subtle changes in fat fraction?; What is the clinical impact of the automated liver segmentation feature on workflow efficiency and inter-operator variability compared to manual or ROI-based methods in routine practice?; Given the documented underestimation of T1 values by MOLLI-based techniques, what specific calibration or correction factors, if any, are applied or recommended when interpreting LMSv3-derived cT1 values in longitudinal patient monitoring?; How does the software handle motion artifacts or other common MR image quality issues, and what are the thresholds for image acceptability for reliable quantitative analysis?; What are the specific differences in accuracy and clinical utility between PDFF values derived using MAGNITUDE ONLY-IDEAL (for GE/Philips scanners) versus COMPLEX-IDEAL (for Siemens scanners) as implemented in LMSv3?