Quantib Prostate

Quantib Prostate version 3.0 is an image post-processing software intended to aid medical professionals in interpreting prostate MRI studies by providing tools for visualization, semi-automatic segmentation, volume computation, and structured PI-RADS scoring. The system utilizes an improved and retrained algorithm for prostate and sub-region segmentation from multiparametric MRI data, enhancing the analysis and review of MR datasets.

Quantib Prostate version 3.0 is an image post-processing software designed to assist trained medical professionals in the analysis and interpretation of prostate MRI studies. It provides functionalities for processing, visualizing, and editing MRI images, enabling review of MR data sets with additional mathematical and statistical analysis. The software supports registered multiparametric-MRI (mpMRI) viewing, offers an option to combine images into a biparametric view for better visualization, performs semi-automatic segmentation of anatomical structures, calculates volumes, includes tools for manual editing, and facilitates PI-RADS scoring through a structured workflow. The output, including images overlaid onto source MRIs, is intended to aid clinical interpretation but not serve as the sole basis for diagnosis. The key modification in version 3.0 is an improved prostate and sub-region segmentation algorithm, which was updated, retrained on over 400 scans, and consolidated from two separate algorithms into one. Non-clinical performance testing involved comparing automatic segmentations against ground truth using Dice overlap and Mean Surface Distance. Clinical performance was assessed qualitatively, with radiologists scoring sub-region segmentations and ROI initial localizations. Human factor enhancements include the ability for users to switch between editing steps and an initialized ROI drawing for PI-RADS. The device also generates 3D segmentations of the prostate and ROIs as secondary captures and adds these images to PDF reports, ensuring it remains as safe and effective as its predicate.

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 specific deep learning architecture or machine learning techniques are employed in the improved combined prostate and sub-region segmentation algorithm, and how do they contribute to its enhanced performance over version 2.0?; Could you provide the quantitative Dice overlap and Mean Surface Distance metrics for the improved segmentation algorithm (version 3.0) and how they statistically compare to both the predicate device (version 2.0) and inter-observer variability referenced in bench testing?; How does the 'biparametric combination image' feature quantitatively impact lesion conspicuity or diagnostic efficiency compared to viewing individual mpMRI sequences separately within the Quantib Prostate software?; What were the specific Likert scale results or qualitative feedback themes from radiologists regarding the accuracy of sub-region segmentations and the clinical utility of the initialized ROI localizations in the PI-RADS workflow?; Beyond assisting interpretation, does the software offer any quantitative tracking of lesion characteristics or prostate volume over time if multiple studies are processed for the same patient, and how is this data presented?