ClariSIGMAM

ClariSIGMAM is an automated image processing software designed to enhance breast cancer screening by objectively quantifying breast density from 2D digital mammograms. It utilizes algorithms to calculate the ratio of fibroglandular tissue to total breast area, providing a binary BI-RADS density classification (fatty vs. dense) as adjunctive information for clinicians.

ClariSIGMAM is a standalone software application designed to automatically analyze "for presentation" 2D digital mammograms to aid in assessing breast tissue composition. It functions by calculating the percent breast density, defined as the ratio of fibroglandular tissue area to total breast area, for each breast. The software then categorizes breast density into two groups: BI-RADS A+B (fatty) and BI-RADS C+D (dense), aligning with the American College of Radiology’s BI-RADS density classification. This information, including specific area measurements (cm²) for breast and fibroglandular tissue and percent density, is outputted as a DICOM structured report or secondary capture for display on mammography workstations or PACS. ClariSIGMAM is intended for use with compatible Hologic and GE full-field digital mammography systems and provides adjunctive information, not serving as a direct diagnostic aid. The performance of ClariSIGMAM was validated through non-clinical testing, comparing its automated breast density estimates against a gold standard established by expert radiologists using interactive thresholding software (Cumulus). Validation also assessed reproducibility concerning age-related density decrease, consistency over time with paired mammograms, and similarity in density estimates between left and right breasts. In a key comparison with a consensus visual assessment by four expert readers on a dataset of 837 mammograms, ClariSIGMAM demonstrated substantial agreement for binary breast density classification (fatty vs. dense), achieving a Kappa statistic of 0.734. The software showed a sensitivity of 86.6% in identifying consensus-defined fatty breasts and a sensitivity of 87.3% in identifying consensus-defined dense breasts. The software meets established international standards and guidance for medical device software and interoperability.

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 ClariSIGMAM's algorithm differentiate between fibroglandular tissue and other breast tissues in 2D mammograms, and what are its limitations with overlapping tissues?; Given the Kappa value of 0.734, what were the primary sources of disagreement between ClariSIGMAM's automated assessment and the expert reader consensus for binary density classification?; The software outputs a binary classification (BI-RADS A+B vs. C+D). How does this simplification compare to the clinical utility of the full four-category BI-RADS assessment, particularly for borderline cases between B and C?; What specific image quality parameters or acquisition artifacts from the compatible Hologic and GE systems most significantly impact the accuracy of ClariSIGMAM's density calculations?; How were the 'substantial data sets' for validation testing characterized in terms of patient demographics (age, ethnicity), BI-RADS density distribution, and image acquisition variability, and how might these factors influence real-world performance?