Breast lesions associated with mammographic architectural distortion: a study of 588 core needle biopsies

Combination of Deep Learning Grad-CAM and Radiomics for Automatic Localization and Diagnosis of Architectural Distortion on DBT

RATIONALE AND OBJECTIVES

Detection and diagnosis of architectural distortion (AD) on digital breast tomosynthesis (DBT) is challenging. This study applied artificial intelligence (AI) using deep learning (DL) algorithms to detect AD, followed by radiomics for classification.

MATERIALS AND METHODS

500 cases with AD on DBT reports were identified; the earlier 292 cases for training, and the later 208 cases for testing. The DL Gradient-weighted Class Activation Mapping (Grad-CAM) was applied to automatically localize abnormalities and generate a region of interest (ROI), which was put into the radiomics model to estimate the malignancy probability for constructing ROC curves. Radiologists delineated ROI manually for comparison. Cases were categorized into pure AD and AD associated with other features, including mass, regional high-density, and calcifications. The ROC curves were compared using the DeLong test.

RESULTS

The overall malignancy rate was 57% (285/500). Of them, 267 cases were classified as pure AD, and the malignancy rate (106/267 = 39.7%) was significantly lower compared to AD cases associated with other features (179/233 = 76.8%, p < 0.01). In the testing set, the diagnostic AUC was 0.82 when using the manual ROI and 0.84 when using the DL-generated ROI. In the more challenging pure AD cases, DL-generated ROI yielded an AUC of 0.77, significantly lower than 0.86 for AD associated with other features.

CONCLUSION

DL could detect AD on DBT, and the diagnostic performance was comparable to manual ROI. The strategy worked for pure AD, but the performance was worse than that for AD with other features.

Diagnostic accuracy of the breast MRI Kaiser score in suspected architectural distortions and its comparison with mammography

Suspicious architectural distortion is an elusive finding in breast cancer diagnosis. This study aimed to evaluate the diagnostic accuracy of the Kaiser score for suspicious architectural distortions observed on mammography and compare it with the BI-RADS score of the lesion. Mammograms performed between January 2013 and March 2023 were retrospectively analyzed for the presence of suspicious architectural distortion. Forty-one patients, who had at least 1 year of radiological follow-up or pathology results, and underwent breast MRI, were included in the study. Mammography findings and the BI-RADS category of the lesion were assessed. MRI findings were evaluated and Kaiser scoring was performed according to the tree flowchart. Ninety-one percent of the enhanced lesions had a Kaiser score of 5 and above. In the diagnosis of malignancy, the Kaiser score yielded an accuracy of 75.61% (AUC 0.833). A statistically significant correlation was observed indicating that a malignant diagnosis was more prevalent in patients with a Kaiser score of 5 and above (p < 0.05). Additionally statistically significant relationship was also observed between the BI-RADS category of architectural distortions on mammography and the Kaiser score (p = 0.007). The combined utilization of mammography findings and the evidence-based Kaiser score in suspected architectural distortions provides more accurate results in the differential diagnosis of breast cancer.

High-Risk Lesion Management

High-risk lesions or lesions of uncertain malignant potential are frequent findings on image-guided needle biopsy of the breast and comprise a number of distinct entities. These lesions are known for having risk of underlying malignancy and are usually associated with an increased lifetime risk for breast cancer. Surgical excision was traditionally recommended for all high-risk lesions but recent studies have demonstrated that vacuum-assisted excision or surveillance may be adequate for some lesions. While management of high-risk lesion varies among institutions, this chapter describes the management recommendations based on recent literature of the most frequent types of lesions.

Diagnosis of architectural distortion on digital breast tomosynthesis using radiomics and deep learning

Purpose

To implement two Artificial Intelligence (AI) methods, radiomics and deep learning, to build diagnostic models for patients presenting with architectural distortion on Digital Breast Tomosynthesis (DBT) images.

Materials and Methods

A total of 298 patients were identified from a retrospective review, and all of them had confirmed pathological diagnoses, 175 malignant and 123 benign. The BI-RADS scores of DBT were obtained from the radiology reports, classified into 2, 3, 4A, 4B, 4C, and 5. The architectural distortion areas on craniocaudal (CC) and mediolateral oblique (MLO) views were manually outlined as the region of interest (ROI) for the radiomics analysis. Features were extracted using PyRadiomics, and then the support vector machine (SVM) was applied to select important features and build the classification model. Deep learning was performed using the ResNet50 algorithm, with the binary output of malignancy and benignity. The Gradient-weighted Class Activation Mapping (Grad-CAM) method was utilized to localize the suspicious areas. The predicted malignancy probability was used to construct the ROC curves, compared by the DeLong test. The binary diagnosis was made using the threshold of ≥ 0.5 as malignant.

Results

The majority of malignant lesions had BI-RADS scores of 4B, 4C, and 5 (148/175 = 84.6%). In the benign group, a substantial number of patients also had high BI-RADS ≥ 4B (56/123 = 45.5%), and the majority had BI-RADS ≥ 4A (102/123 = 82.9%). The radiomics model built using the combined CC+MLO features yielded an area under curve (AUC) of 0.82, the sensitivity of 0.78, specificity of 0.68, and accuracy of 0.74. If only features from CC were used, the AUC was 0.77, and if only features from MLO were used, the AUC was 0.72. The deep-learning model yielded an AUC of 0.61, significantly lower than all radiomics models (p<0.01), which was presumably due to the use of the entire image as input. The Grad-CAM could localize the architectural distortion areas.

Conclusion

The radiomics model can achieve a satisfactory diagnostic accuracy, and the high specificity in the benign group can be used to avoid unnecessary biopsies. Deep learning can be used to localize the architectural distortion areas, which may provide an automatic method for ROI delineation to facilitate the development of a fully-automatic computer-aided diagnosis system using combined AI strategies.