LiverMultiScan

LiverMultiScan (LMSv4) is a post-processing software application that utilizes DICOM 3.0 compliant MR images to perform quantitative analysis of liver tissue, providing metrics such as cT1, T2*, and PDFF through automated segmentation and region of interest tools. Its purpose is to offer clinicians non-invasive, reproducible imaging biomarkers to aid in the evaluation and characterization of liver health as part of a comprehensive diagnostic process.

LiverMultiScan (LMSv4) is a standalone, post-processing software device designed for the non-invasive evaluation of liver tissue using Magnetic Resonance (MR) images. It assists trained operators at Perspectum's image analysis center in analyzing DICOM 3.0 compliant MR datasets. The software provides tools for automatic liver segmentation and region of interest (ROI) placement on previously acquired images to quantify key physiological parameters. LMSv4 generates a summary report containing these quantitative metrics and images, which is then sent to an interpreting clinician. This report serves as an adjunctive diagnostic tool, intended to complement existing clinical evaluation methods such as blood tests, ultrasound, and biopsy, rather than replace them. The software aims to help characterize liver tissue and identify patients who might benefit from further investigation, with the final diagnosis and treatment decisions remaining the responsibility of the clinician. LMSv4 quantifies three main metrics: iron-corrected T1 (cT1), T2*, and Proton Density Fat Fraction (PDFF). For cT1, it supports full liver segmentation or ROI-based methods, reporting median and interquartile range (IQR). T2* is calculated using either the LMS MOST method or a three-point DIXON method via ROI analysis. PDFF is determined using the LMS IDEAL method, with options for full liver segmentation or ROI placement. Performance testing using phantoms demonstrated accuracy for T2* and PDFF within physiological ranges, though cT1 (MOLLI-based T1) showed a consistent underestimation compared to ground truth, a known characteristic of MOLLI techniques. Clinical testing on volunteer data confirmed high repeatability and reproducibility for cT1, T2*, and PDFF measurements across different scanners and field strengths, with minimal variability introduced by different operators using both ROI and segmentation methods.

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. What are the specific advantages and limitations of the LMS MOST method for T2* calculation compared to the traditional three-point DIXON method, particularly in diverse patient populations or varying degrees of iron overload?; The document mentions cT1 underestimation with MOLLI techniques. How does LMSv4 account for or communicate this known bias in its reports to ensure clinicians interpret cT1 values correctly, especially when tracking longitudinal changes?; How does the performance (accuracy, precision, reproducibility) of LMSv4's automated liver segmentation compare to manual segmentation by experienced radiologists, and what is the typical processing time?; What is the range of MRI scanner manufacturers and field strengths validated for use with LMSv4, and are there any specific acquisition protocols required to ensure optimal performance of the quantitative metrics?; Given that LMSv4 is used by trained Perspectum operators, what is the training and quality control process for these operators to ensure consistent and accurate analysis across different cases and over time?