Data-Driven Breast Care
By Rebecca Montz, EdD, MBA, CNMT, PET, RT(N)(CT), NMTCB RS
Radiology Today
Vol. 26 No. 2 P. 18
Transforming the Future of Breast Cancer Detection and Care
AI is revolutionizing breast cancer detection, offering new hope for early identification that could save millions of lives. In 2022, the World Health Organization reported approximately 2.3 million women were diagnosed with breast cancer, and 670,000 lost their lives to the disease. Every 14 seconds, a woman is diagnosed with breast cancer globally, according to the Breast Cancer Research Foundation. Early detection is critical to improving survival rates. When diagnosed at a localized stage, where treatment is more effective, the five-year relative survival rate exceeds 99%, as reported by the American Cancer Society. However, this rate drops to 87% if diagnosed at a regional stage and just 32% for distant-stage cancer, underscoring the importance of timely detection. Unfortunately, screening mammograms miss about 20% of breast cancers, according to the National Cancer Institute, and inaccurate results can lead to delayed treatments and more advanced-stage diagnoses.
AI is transforming breast imaging by assisting radiologists in detecting subtle and aggressive cancers that may otherwise be missed. By automating routine tasks and providing data-driven insights, AI flags areas of concern, allowing radiologists to focus on more complex cases. AI enhances diagnostic efficiency by automating tasks, speeding up processes, and reducing the risk of missed diagnoses. It improves outcomes by providing consistent, accurate results across diverse populations, addressing health inequalities. By using diverse datasets, AI can minimize biases, contributing to equitable access to early detection and quality care for all patients.
Enhanced Access and Equity
Diverse AI datasets are crucial in advancing breast cancer diagnostics, enabling algorithms to adapt to the different ways the disease manifests across various populations. These datasets enhance accuracy and personalize detection tools, ensuring the models are applicable to broader demographics. By incorporating patient data that reflects genetic, environmental, and cultural factors, AI systems can identify disparities and offer more equitable care. This strategy helps minimize bias, ensuring that women from all racial, ethnic, and socioeconomic backgrounds receive accurate and fair treatment, ultimately reducing health care inequalities.
Companies such as iCAD and Hologic are driving AI innovation in breast imaging, significantly improving diagnostic accuracy and patient care. These companies utilize diverse global datasets to train AI algorithms, ensuring their tools can detect breast cancer across various populations. Their focus on comprehensive datasets reflects global variations in breast cancer presentation, helping create inclusive diagnostic models. This approach plays a key role in reducing health care disparities worldwide by making breast cancer detection more accurate and equitable.
Siemens Healthineers does not yet offer AI-driven solutions for breast imaging, but the company has recently introduced an innovative combination of exclusive technologies designed to enhance breast imaging capabilities. One of these groundbreaking technologies is the proprietary next-generation 2D Shear Wave elastography, which has been specifically developed to improve the detection of stiff lesions in the breast. This advanced technology adapts to various tissue characteristics, offering improved sensitivity and accuracy in identifying potentially malignant lesions. Additionally, Siemens Healthineers has unveiled the HLX highfrequency transducer, which provides superior penetration and resolution, allowing for clearer and more detailed imaging. Together, these advancements empower clinicians with the diagnostic tools necessary to detect lesions more effectively and reduce the likelihood of false negatives. As a result, Siemens Healthineers is continuing to make significant strides in improving breast imaging technologies, enabling health care professionals to provide better, more accurate care for their patients. iCAD is committed to bridging gaps in health care equity through AI in breast cancer screening. Dana Brown, president, CEO, and chair of the board of iCAD, says the company’s technology addresses disparities in outcomes, particularly for Black women, and provides life-saving insights to underserved regions. By ensuring all patients, regardless of background, have better access to early detection and improved outcomes, iCAD is advancing more equitable health care solutions. The company’s global database, comprising over 7 million mammographic images and 8,000 biopsy-proven cancers, allows iCAD’s algorithms to be accurate and effective across different populations and clinical scenarios. Nikolaos Gkanatsios, vice president of research at iCAD, emphasizes the significance of this diverse dataset in enabling the detection of even the most subtle and aggressive cancers.
Similarly, Hologic employs large, highquality patient and imaging data sets to train and test algorithms. Paola Wisner, vice president of global research and development for breast and skeletal health at Hologic, explains that the company uses its broad install base to create diverse datasets representing various racial and ethnic groups, contributing to the development of inclusive models. These datasets, combined with biopsy results and patient outcomes, establish a “ground truth” for the AI algorithms. After extensive internal validation, the algorithms undergo clinical testing and FDA review for safety and efficacy. Hologic continues to refine its algorithms based on user feedback and technological advancements, submitting updates for FDA clearance.
Hologic also conducted a study, titled “Performance of a Digital Breast Tomosynthesis AI Detection Algorithm in Common US Racial/Ethnic Groups,” which found that the measured performance of its Genius AI Detection algorithm was similar across Asian, Black, Hispanic, and white women. The study helps increase clinical confidence in this AI solution’s ability to deliver results not materially impacted by racial bias.
Although Siemens Healthineers has not yet introduced AI for breast imaging, the company is making progress in AI development across its broader portfolio, underpinned by a global strategy that leverages diverse datasets from more than 200 partners spanning five continents. These comprehensive datasets— encompassing clinical images, lab data, genomic information, and patient histories— serve as the foundation for Siemens Healthineers’ robust repository, providing a controlled environment for training AI algorithms. With a “data lake” containing over two billion data points, Siemens Healthineers continually updates and refines its algorithms, ensuring daily improvements in accuracy. Jennifer Gregov, RDCS, RVT, CPM, ACPMPO, senior product manager at Siemens Healthineers, highlights that this expansive global network is critical in developing AI models that are not only high-quality but also inclusive and adaptable across various populations. This commitment to continuous improvement enables Siemens Healthineers to create AI solutions that are not only precise but also equitable and relevant across diverse patient backgrounds.
The use of diverse AI datasets is a transformative force in breast cancer diagnostics, allowing algorithms to adapt to the unique ways the disease manifests across various populations. By ensuring these datasets reflect genetic, environmental, and cultural differences, AI can offer more accurate, personalized, and equitable care, reducing health care disparities and improving outcomes for women from all backgrounds. The leading companies in breast cancer detection are not just advancing diagnostic accuracy, they’re actively dismantling biases to ensure fairness in care. By integrating diverse datasets, they’re shaping a future where early detection and high-quality treatment are available to everyone, regardless of race, ethnicity, or socioeconomic status. Their commitment is setting the foundation for a more equitable health care system, where every individual has the opportunity for timely diagnosis and improved outcomes.
Advancing Breast Imaging
Breast imaging technology has evolved significantly, from the introduction of mammography to the development of 3D mammography. Despite these advancements, the complexity of breast tissue and subtle abnormalities make early detection challenging. AI, particularly machine learning and deep learning algorithms, is transforming breast cancer detection. By analyzing vast datasets of mammograms, MRIs, and ultrasounds, AI can identify patterns and abnormalities that may be missed by human eyes. These AI systems support radiologists by offering second opinions, improving diagnostic accuracy, and reducing false positives. AI’s ability to detect even subtle abnormalities leads to more comprehensive evaluations of breast tissue, reducing the need for unnecessary exams and invasive procedures, as well as improving diagnostic efficiency, patient comfort, and outcomes. Advanced AI solutions are pushing the boundaries of early detection, allowing for more precise and efficient breast cancer screening.
ProFound Detection Version 4.0 by iCAD
iCAD’s ProFound Detection Version 4.0 is an advanced AI solution designed to enhance breast cancer detection in digital breast tomosynthesis. Leveraging deep learning algorithms, this software helps radiologists identify subtle abnormalities that might otherwise go unnoticed, highlighting areas of concern such as masses, lesions, or calcifications and offering an AI-driven second opinion. ProFound Detection 4.0 reduces false positives and unnecessary biopsies by accurately distinguishing between benign and malignant abnormalities, allowing radiologists to focus on more complex cases.
Chirag Parghi, MD, MBA, CMO at Solis Mammography, notes that the latest version excels at detecting subtle invasive cancers, especially in dense breast tissue, a traditionally difficult area for imaging. Integrated seamlessly into clinical workflows, ProFound Detection 4.0 delivers real-time results, boosting efficiency without disrupting patient care. Building on over 20 years of innovation, the software improves overall cancer detection by 22%, with notable gains in identifying cancers in dense tissue (50%), invasive lobular cancers (60%), and smaller cancers (38%).
Vasu Avadhanula, chief product officer of iCAD, highlights the critical role of early detection, noting that stage 1 diagnoses have a survival rate exceeding 99%. ProFound Detection 4.0 increases cancer detection rates by up to 23% compared with traditional methods, helping alleviate both the emotional and financial strain on patients and their families. iCAD is committed to seamlessly blending cutting-edge AI technology with clinical expertise by consistently refining their solutions to address the evolving needs of both health care providers and patients.
Hologic’s Genius AI Detection
Hologic’s Genius AI Detection products enhance breast cancer detection by analyzing each “slice” of a tomosynthesis image set and marking potentially cancerous lesions. These marks can be overlaid on synthesized 2D images or 3DQuorum SmartSlices, enabling radiologists to easily identify and navigate to the corresponding slice for further analysis. Genius AI Detection 2.0 provides automated lesion pairing across different mammography views, streamlining the review process. It also allows radiologists to prioritize their worklist by sorting cases based on cancer suspicion or complexity.
The new Genius AI Detection PRO solution—which expands on Hologic’s flagship Genius AI Detection 2.0 solution—integrates a red-yellow-green color-coding system to further aid radiologists in assessing cases, reducing reading time by 24%. The platform also helps solve the issue of interoperability by offering an all-in-one user interface through an application that can be downloaded on any reading workstation, allowing most customers to get Genius AI Detection PRO features such as lesion detection, density, image quality, and reporting features with little integration work needed. Additionally, AI-based assessments of image quality and breast density can improve workflow efficiency and diagnostic accuracy.
Siemens Healthineers ACUSON Sequoia
The ACUSON Sequoia ultrasound system from Siemens Healthineers, while not yet equipped with AI for breast imaging, offers a variety of advanced imaging technologies that enhance breast cancer detection, particularly in patients with dense breast tissue. The system’s highresolution imaging capability ensures that even subtle abnormalities are captured with clarity, which is crucial for early detection of breast cancer. Additionally, the ACUSON Sequoia integrates elastography technology, which measures tissue stiffness—a key indicator in distinguishing between benign and malignant masses. This ability helps reduce the need for unnecessary biopsies, streamlining the diagnostic process and minimizing patient discomfort.
The system also supports 3D imaging and high-frequency ultrasound, offering detailed, multidimensional views of the breast tissue, which can improve visualization of both superficial and deeper structures. These features are essential for clinicians to assess and differentiate lesions more effectively, providing greater confidence in their diagnoses. Designed with a focus on delivering comprehensive and reliable solutions, the ACUSON Sequoia ultrasound system plays a pivotal role in breast cancer screening and diagnosis, ultimately improving patient outcomes by ensuring more accurate and timely interventions.
A Healthier Future
The rise of AI in breast cancer detection is more than just a technological advancement— it’s a promise for a healthier, more equitable future. Industry leaders such as iCAD, Hologic, and Siemens Healthineers are driving the revolution in breast cancer detection, equipping radiologists with advanced tools that enhance the accuracy of detecting even the most subtle signs of cancer and simplifying their workflows.
iCAD and Hologic are incorporating diverse global datasets into breast imaging, with their AI systems tackling health care disparities head-on. By utilizing this wealth of data, they are ensuring that every woman—no matter her background or location—has an equal opportunity for early detection and timely treatment. As AI continues to evolve, it’s clear that its impact will go far beyond just improving diagnostic efficiency; it will help save lives, reduce emotional and financial burdens, and ultimately shape a world where breast cancer is detected earlier, treated more effectively, and faced with hope rather than fear.
— Rebecca Montz, EdD, MBA, CNMT, PET, RT(N)(CT), NMTCB RS, has worked at the Mayo Clinic Jacksonville and University of Texas MD Anderson Cancer Center in Houston as a nuclear medicine and PET technologist.
