SBIR Phase I: Compact, Low-cost, Automated 3D Ultrasound System for Regular and Accessible Breast Imaging

这个SBIR一期项目引入了一种早期监测和发现乳腺癌的新模式：量化自我检查。在美国，每年有30多万女性被诊断出乳腺癌，4万女性死于乳腺癌。如果早期发现，乳腺癌的存活率为99%，但现有筛查技术在成本、敏感性、可及性和便利性方面的限制导致三分之一的乳腺癌在早期阶段被遗漏。由于早期癌症的治疗费用比3期和4期癌症的治疗费用低一个数量级，因此开发更好的乳腺癌监测和筛查工具具有明显的经济和社会效益。为了应对这一挑战，该项目提出的技术利用了超声成像的优势和基于云计算的人工智能的新功能，提供了一种定期和可访问的自我监测工具，可以量化和跟踪乳房组织的可疑变化。该设备的便携性、低成本、2分钟扫描时间和乳房图像数据的自动分析将大大增加女性乳腺癌监测的可及性，从而减少美国医疗保健系统对该疾病的成本负担。该SBIR一期项目建议开发一种用于早期检测和监测乳腺癌的新工具：量化自我检查（QSE），该工具将低成本、紧凑型3D超声设备和定位配件与人工智能相结合，为女性及其医生提供适当和可操作的数据。本项目提出的QSE技术独立于用户技能，在2分钟内捕获整个乳房的3D体积图像。系统架构允许简化和低成本的超声硬件无线连接到智能手机/平板电脑，并将捕获的数据传输到安全的云进行高级图像处理和存储。超声波扫描仪连接到定位附件上，可重复成像，通过基于机器学习的图像分析，实现对异常生长的纵向3D监测。拟议的第一阶段研发工作侧重于四个目标：(I)优化电子硬件和低水平成像软件，以提高空间分辨率和图像质量；（ii）建立一个QSE扫描仪，最大限度地提高视野和体积完整性；（iii）构建定位附件，用于QSE扫描仪的定位；（iv）通过验证乳腺假体上3D超声体积的对齐来证明QSE成像的纵向可重复性。

This SBIR Phase I project introduces a new paradigm for early monitoring and detection of breast cancer: the Quantified Self Exam. In the United States over 300,000 women are diagnosed and 40,000 women die from breast cancer every year. Breast cancer has a 99% survival rate if detected early, but limitations in cost, sensitivity, accessibility, and convenience of existing screening technologies result in one third of breasts cancers getting missed at early stages. Since treatment for early stage cancer is an order of magnitude less costly than treatment for stage 3 and 4 cancers, there is a clear economic and societal benefit for the development of better breast cancer monitoring and screening tools. To address this challenge, the technology proposed in this project leverages the proven benefits of ultrasound imaging and the newfound power of cloud-based artificial intelligence to provide a regular and accessible self-monitoring tool that can quantify and track suspicious changes in breast tissue. The device portability, low cost, 2 minute scan time, and automated analysis of breast image data will greatly increase the accessibility of breast cancer monitoring for women, which in turn stands to decrease the cost burden of this disease for the US healthcare system. This SBIR Phase I project proposes to develop a new tool for early detection and monitoring of breast cancer: the Quantified Self Exam (QSE) that combines a low-cost, compact 3D ultrasound device and positioning accessory with artificial intelligence to empower women and their physicians with appropriate and actionable data. The QSE technology proposed in this project operates independent of user skill and captures 3D volumetric images of the whole breast in 2 minutes. The system architecture allows for simplified and low-cost ultrasound hardware that connects wirelessly to a smart phone/tablet and transfers captured data to secure cloud for advanced image processing and storage. The ultrasound scanner attaches to a positioning accessory for repeatable imaging that enables longitudinal 3D monitoring of abnormal growth using machine learning-based image analysis. The proposed Phase I R&D efforts focus on four objectives: (i) optimize electrical hardware and low-level imaging software for spatial resolution and image quality; (ii) build a QSE scanner that maximizes field of view and volumetric integrity; (iii) build a positioning accessory for positioning of the QSE scanner; (iv) demonstrate the longitudinal repeatability of QSE imaging by validating the alignment of 3D ultrasound volumes on a breast phantom.