Evaluating the completeness of scanning in hand-held whole breast ultrasound exams

点击翻译按钮获取中文摘要

ABSTRACT: The project entails the development of a position-tracking system compatible with existing hand-held ultrasound systems that can track the complete breast volume coverage with a sufficient level of precision to avoid missing small breast cancers. Scanning the entire breast for screening breast cancer gained increased interest in recent years, as it was shown that supplemental screening in women with dense mammograms can significantly increase the detection rate of small cancers. While detecting sub-centimeter lesions is essential for the successful treatment of breast cancer, it was shown that around half of all lesions smaller than 9 mm were not detected by experienced operators12. This significant shortcoming can be partly attributed to an incomplete scanning of the breast caused by operator's imprecise visual estimation and also by large-scale deformation of the breast during the procedure. The resultant localization error can be in the order of a few centimeters and will increase the likelihood of missing regions of the scanned breast. Current hand-held ultrasound systems cannot track and document completeness of scanning or alert operators of regions with incomplete scanning. Metritrack Inc. developed a proprietary method of mapping hand-held breast ultrasound images that was incorporated in its FDA-cleared automated breast ultrasound system BVN G-1000. The proposed project intends to develop a method to assess the volumetric completeness of breast ultrasound scanning, based on the proprietary mapping method and the obtained clinical data. The project is segmented into two aims. For aim 1, we will build a software application designed to automatically detect the breast regions without optimal coverage and guide the operator to obtaining a complete exam. Aim 2 will evaluate the accuracy and precision of detecting breast regions without optimal coverage. For this purpose, a customized ultrasound breast phantom with multiple small internal targets simulating small tumors will be built and scanned under different conditions of deformation, by 2 operators to simulate clinical conditions of use. We expect that the outcome of this Phase 1 project will demonstrate the feasibility of building the completeness of scanning verification and guidance application into an automated breast ultrasound system to overcome the current limitations of breast ultrasound. The prototype will be developed into a commercially successful product to provide standardized, complete breast ultrasound exams of the whole breast with a higher cancer detection rate than possible today.