Deep Learning-Based Computer-Aided Diagnosis Improves Breast Lesion Classification on Ultrasound
A recent study published in the American Journal of Roentgenology (AJR) found that deep learning-based computer-aided diagnosis (CAD) significantly improved radiologists’ diagnostic performance for breast lesion classification on ultrasound, particularly in reducing the frequency of benign breast biopsies. The study was conducted by a team of researchers from Peking University Third Hospital in Beijing, China.
The study involved patients scheduled to undergo biopsy or surgical resection of a breast lesion classified as BI-RADS category 3-5 on prior breast ultrasound at eight Chinese secondary or rural hospitals from November 2021 to September 2022. An investigational breast ultrasound was performed and interpreted by a radiologist with no expertise in the modality. Hybrid body-breast imagers, radiologists lacking breast subspecialty training or in whom breast ultrasound accounted for less than 10% of their ultrasounds performed annually, then assigned a BI-RADS category. CAD results were used to upgrade reader-assigned BI-RADS category 3 lesions to category 4A, as well as for downgrading BI-RADS 4A lesions to 3. Histologic results of biopsy or resection served as the researchers’ reference standard.
The application of CAD to interpretations by radiologists without breast ultrasound expertise resulted in the upgrade of 6.0% of BI-RADS category 3 assessments to category 4A, of which 16.7% were malignant, and the downgrade of 79.1% of category 4A assessments to category 3, of which 4.6% were malignant. This demonstrates the potential of CAD to improve the accuracy of breast lesion classification on ultrasound and reduce the number of unnecessary biopsies.
According to the authors, institutions lacking breast imaging expertise may also suffer from capacity issues to perform image-guided breast biopsies and pathologic evaluation of biopsy specimens. Therefore, decreasing benign biopsies with CAD could have a significant impact on these institutions.
The study’s results support the use of CAD in settings with incomplete access to breast imaging expertise. Compared with the literature supporting CAD at tertiary and/or urban centers, this prospective multicenter study of radiologists without breast ultrasound expertise provides evidence for the efficacy of CAD in improving diagnostic performance for breast lesion classification on ultrasound.
The American Roentgen Ray Society (ARRS), North America’s first radiological society, remains dedicated to the advancement of medicine through the profession of medical imaging and its allied sciences. The ARRS aims to improve health through a community committed to advancing knowledge and skills with the world’s longest continuously published radiology journal, the American Journal of Roentgenology, and various educational materials.
- Breast ultrasound diagnosis
- Computer-aided diagnosis
- Multicenter study
- Breast cancer detection
- Ultrasound expertise improvement
News Source : Mirage News
Source Link :Multicenter Study Shows Computer-Aided Diagnosis Enhances Breast Ultrasound Expertise/