Breast parenchymal patterns in processed versus raw digital mammograms: A large population study toward assessing differences in quantitative measures across image representations

Abstract

Purpose: With raw digital mammograms (DMs), which retain the relationship with x-ray attenuation of the breast tissue, not being routinely available, processed DMs are often the only viable means to acquire imaging measures. The authors investigate differences in quantitative measures of breast density and parenchymal texture, shown to have value in breast cancer risk assessment, between the two DM representations. Methods: The authors report data from 8458 pairs of bilateral raw ("FOR PROCESSING") and processed ("FOR PRESENTATION") DMs acquired from 4278 women undergoing routine screening evaluation, collected with DM units from two different vendors. Breast dense tissue area and percent density (PD), as well as a range of quantitative descriptors of breast parenchymal texture (statistical, co-occurrence, run-length, and structural descriptors), were measured using previously validated, fully automated software. Feature measurements were compared using matched-pairs Wilcoxon signed-ranks test, correlation (r), and linear-mixed-effects (LME) models, where potential interactions with woman-and system-specific factors were also assessed. The authors also compared texture feature correlations with the established risk factors of the Gail lifetime risk score (r G ) and breast PD (r PD ), and evaluated the within woman intraclass feature correlation (ICC), a measure of bilateral breast-tissue symmetry, in raw versus processed images. Results: All density measures and most of the texture features were strongly (r ≥ 0.6) or moderately (0.4 ≤ r < 0.6) correlated between raw and processed images. However, measurements were significantly different between the two imaging formats (Wilcoxon signed-ranks test, p w < 0.05). The association between measurements varied across features and vendors, and was substantially modified by woman-and system-specific image acquisition factors, such as age, BMI, and mAs/kVp, respectively. The strongest correlation, combined with minimal LME-model interactions, was observed for structural texture features. Overall, texture measures from either image representation were weakly associated with Gail lifetime risk (−0.2 ≤ r G ≤ 0.2), weakly to moderately associated with breast PD (−0.6 ≤ r PD ≤ 0.6), and had overall strong bilateral symmetry (ICC ≥ 0.6). Conclusions: Differences in measures from processed versus raw DM depend highly on the feature, the DM vendor, and image acquisition settings, where structural features appear to be more robust across the different DM settings. The reported findings may serve as a reference in the design of future large-scale studies on mammographic features and breast cancer risk assessment involving multiple DM representations. C

Incorporating Breast Anatomy in Computational Phenotyping of Mammographic Parenchymal Patterns for Breast Cancer Risk Estimation

“…In addition, our analysis was confined to a fixed feature set and, although elastic-net regression was used to alleviate model over-fitting, our reported model performance may be over-estimated due to the relatively small sample size as a single-institution evaluation. Considering the reported substantial differences in textural measurements across image acquisition settings 27 , different FFDM representations, and vendors 28 , in our future studies we will plan to more thoroughly test the robustness of our method by incorporating multiple FFDM vendors from larger populations. In addition, larger studies will allow us to more rigorously evaluate the added discriminatory capacity of such imaging biomarkers when considering additional demographic and clinical risk factors ( e.g ., age at menarche, parity, family history of breast cancer), potentially also expanding our feature set into higher phenotypic representations, including deep learning.…”

Breast density: why all the fuss?

“…Though multiparametric TA appears highly predictive of risk, comparative studies on one large dataset are lacking and more research is needed into optimised methodology, including location and size of ROIs chosen. It is also important to know whether the image format (raw or processed) and vendor unit affects the predictive abilities of TA, as this would have a significant impact upon the design of prospective multicentre and multivendor studies (181,182). Nonetheless the results are encouraging, particularly the fact that TA appears to confer information on risk separate from that provided by MD (175,180).…”

Comparisons of the Computed Tomographic Scan and Panoramic Radiography Before Mandibular Third Molar Extraction Surgery

“…Statistical analysis was performed using Instat (GrapPad Inc., USA). Two diagnostic modalities to predict the condition of IAN, the close relationship of tooth and canal, and the close relationship to the buccolingual position were compared by non-parametric Wilcoxon matched-pairs signed-ranks test [ 13 ] followed by Tukey’s post hoc test considering the critical value of the studied range [ q ] >4.136 ( http://elvers.us/stats/tables/qprobability.html ; number of means: 3; degree of freedom: 1,000; comparing with “hypothetical gold standard”) at a significant level [ 14 ]. The association of procedural factors and demographic factors with IAN sensory impairment was compared by non-parametric Mann-Whitney U test [ 4 ] followed by Tukey’s post hoc test considering q >4.136 at a significant level (number of means: 3; the degree of freedom: 956, compared with “hypothetical gold standard”) [ 14 ].…”