Performance of contrast-enhanced mammography for detecting multifocal and multicentric breast cancer and evaluating tumour size, and implications for surgical management: Early experience in a tertiary centre.

INTRODUCTION: To compare diagnostic accuracy of contrast-enhanced mammography (CEM) with standard 2D digital mammography (equivalent to low-energy image; LEM) for detection of multifocal and multicentric breast cancer and evaluation of tumour size and disease extent for preoperative planning. METHODS: Biopsy proven breast cancer patients who underwent CEM preoperatively between January 2021 and January 2023 were included in this study. CEM and LEM images were independently reviewed by at least two blinded readers. Lesion location, number, size (maximal diameter) and extension across the midline and/or nipple invasion were recorded. Tumour number and size estimated on imaging were compared with final operative histology, which served as the gold standard. RESULTS: Forty-nine patients (48 females and 1 male) and 50 cases (one patient had bilateral breast lesions) were included in the analysis. Median patient age was 60 (IQR 51, 69). CEM had significantly higher lesion detection rate compared with LEM, with sensitivities of 78% for LEM and 92% for CEM for the index tumour and 15% for LEM and 100% for CEM for multicentric and multifocal cancer. We found no statistically significant difference in median tumour size measurements on CEM and final surgical specimen (P value = 0.97); however, a significant difference was identified in the tumour size measured on LEM and surgical specimen (P value < 0.001). CONCLUSION: CEM is superior to standard 2D digital mammography for detection of multifocal and multicentric breast cancer and is a reliable and more accurate method for estimating tumour size.

Introducing a secondary segmentation to construct a radiomics model for pulmonary tuberculosis cavities.

PURPOSE: Accurate segmentation (separating diseased portions of the lung from normal appearing lung) is a challenge in radiomic studies of non-neoplastic diseases, such as pulmonary tuberculosis (PTB). In this study, we developed a segmentation method, applicable to chest X-rays (CXR), that can eliminate the need for precise disease delineation, and that is effective for constructing radiomic models for automatic PTB cavity classification. METHODS: This retrospective study used a dataset of 266 posteroanterior CXR of patients diagnosed with laboratory confirmed PTB. The lungs were segmented using a U-net-based in-house automatic segmentation model. A secondary segmentation was developed using a sliding window, superimposed on the primary lung segmentation. Pyradiomics was used for feature extraction from every window which increased the dimensionality of the data, but this allowed us to accurately capture the spread of the features across the lung. Two separate measures (standard-deviation and variance) were used to consolidate the features. Pearson's correlation analysis (with a 0.8 cut-off value) was then applied for dimensionality reduction followed by the construction of Random Forest radiomic models. RESULTS: Two almost identical radiomic signatures consisting of 10 texture features each (9 were the same plus 1 other feature) were identified using the two separate consolidation measures. Two well performing random forest models were constructed from these signatures. The standard-deviation model (AUC = 0.9444 (95% CI, 0.8762; 0.9814)) performed marginally better than the variance model (AUC = 0.9288 (95% CI, 0.9046; 0.9843)). CONCLUSION: The introduction of the secondary sliding window segmentation on CXR could eliminate the need for disease delineation in pulmonary radiomic studies, and it could improve the accuracy of CXR reporting currently regaining prominence as a high-volume screening tool as the developed radiomic models correctly classify cavities from normal CXR.