LungQ v3.0.0

LungQ v3.0.0 is a software tool designed to provide quantitative analysis of pulmonary tissue from CT thoracic datasets, supporting physicians in diagnosis and follow-up. It employs automated 3D segmentation, volumetric and density analysis, and fissure evaluation to generate reproducible metrics from CT scans, with its performance validated through non-clinical testing including comparison to a predicate device and human expert assessments.

LungQ v3.0.0 is a software application designed to aid physicians by providing reproducible quantitative CT values for pulmonary tissue, thereby supporting the diagnosis and follow-up of pulmonary conditions using CT thoracic datasets. It functions as a command-line tool without a graphical user interface, processing DICOM images to perform automated three-dimensional segmentation of lungs, lobes, pulmonary sub-segments, and airways. The software offers capabilities for volumetric analysis, density evaluations (including 15th percentile density), fissure integrity assessment, and generates reports based on these analyses; user-initiated manual editing of segmentations is not supported. The underlying technology of LungQ v3.0.0 relies on computational algorithms to analyze CT scan data for quantitative measurements of lung structures. Its performance was evaluated through non-clinical testing, which included software verification, validation, and a head-to-head substantial equivalence study against its predicate, LungQ v1.1.0. This study compared key metrics such as lung and lobar volumes, density scores (LAA-950HU, LAA-910HU), 15th percentile lung density (PD15), and fissure completeness, using Bland-Altman plots and ROC analysis. The results indicated that LungQ v3.0.0 met predefined equivalence criteria, with differences from the predicate device falling within acceptable limits (e.g., volume differences ≤10%, fissure completeness Az value 0.97). Additionally, its (sub)segmental volume and density outputs showed good agreement with analyses corrected by human experts across various CT scanner models.

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 LungQ v3.0.0's algorithm for (sub)segmental segmentation compare in accuracy and robustness to other leading automated methods, particularly in cases with severe pathology or anatomical variations?; What is the specific impact of using a command-line interface without manual segmentation editing capabilities on clinical workflow efficiency and user adoption, especially for complex cases requiring adjustments?; Can the quantitative outputs of LungQ v3.0.0 (e.g., LAA-950HU, PD15) be reliably correlated with specific clinical outcomes or disease progression markers across diverse patient populations and CT acquisition protocols?; How sensitive is the fissure completeness evaluation to variations in CT slice thickness and reconstruction kernels, and what are the implications for its reliability in assessing collateral ventilation?; What are the performance characteristics of LungQ v3.0.0 when analyzing low-dose CT scans or scans with significant motion artifacts, and how does this affect the reproducibility of its quantitative measurements?