Segmentation-based BI-RADS ensemble classification of breast tumours in ultrasound images

• Published in: International journal of medical informatics (Shannon, Ireland) Ireland), 2024-09, Vol.189, p.105522, Article 105522
• Main Authors: Bobowicz, Maciej, Badocha, Mikołaj, Gwozdziewicz, Katarzyna, Rygusik, Marlena, Kalinowska, Paulina, Szurowska, Edyta, Dziubich, Tomasz
• Format: Article
• Language: English
• Subjects: BI-RADS; Breast cancer; Ensemble approach; Image classification; Segmentation; Ultrasound
• ISSN: 1386-5056; 1872-8243; 1872-8243
• DOI: 10.1016/j.ijmedinf.2024.105522

[Display omitted] •Novel method for segmenting breast ultrasound scans by identifying the boundaries of lesions.•An ensemble of segmentation and classification methods for lesion classification.•Evaluating breast ultrasound segmentators within a consistent framework.•Automated ultrasound BI-RADS classification based on a new UCC BUS dataset. The development of computer-aided diagnosis systems in breast cancer imaging is exponential. Since 2016, 81 papers have described the automated segmentation of breast lesions in ultrasound images using artificial intelligence. However, only two papers have dealt with complex BI-RADS classifications. This study addresses the automatic classification of breast lesions into binary classes (benign vs. malignant) and multiple BI-RADS classes based on a single ultrasonographic image. Achieving this task should reduce the subjectivity of an individual operator’s assessment. Automatic image segmentation methods (PraNet, CaraNet and FCBFormer) adapted to the specific segmentation task were investigated using the U-Net model as a reference. A new classification method was developed using an ensemble of selected segmentation approaches. All experiments were performed on publicly available BUS B, OASBUD, BUSI and private datasets. FCBFormer achieved the best outcomes for the segmentation task with intersection over union metric values of 0.81, 0.80 and 0.73 and Dice values of 0.89, 0.87 and 0.82, respectively, for the BUS B, BUSI and OASBUD datasets. Through a series of experiments, we determined that adding an extra 30-pixel margin to the segmentation mask counteracts the potential errors introduced by the segmentation algorithm. An assembly of the full image classifier, bounding box classifier and masked image classifier was the most accurate for binary classification and had the best accuracy (ACC; 0.908), F1 (0.846) and area under the receiver operating characteristics curve (AUROC; 0.871) in the BUS B and ACC (0.982), F1 (0.984) and AUROC (0.998) in the UCC BUS datasets, outperforming each classifier used separately. It was also the most effective for BI-RADS classification, with ACC of 0.953, F1 of 0.920 and AUROC of 0.986 in UCC BUS. Hard voting was the most effective method for dichotomous classification. For the multi-class BI-RADS classification, the soft voting approach was employed. The proposed new classification approach with an ensemble of segmentation and classification approaches proved more accurate than most pub