Editorial Material
Biochemistry & Molecular Biology
Nehmat Houssami, Karla Kerlikowske
Summary: AI has the potential to be a new tool in the risk assessment and screening of breast cancer, but its impact on relevant clinical outcomes needs to be prospectively evaluated.
Article
Radiology, Nuclear Medicine & Medical Imaging
Yoel Shoshan, Ran Bakalo, Flora Gilboa-Solomon, Vadim Ratner, Ella Barkan, Michal Ozery-Flato, Mika Amit, Daniel Khapun, Emily B. Ambinder, Eniola T. Oluyemi, Babita Panigrahi, Philip A. DiCarlo, Michal Rosen-Zvi, Lisa A. Mullen
Summary: The study evaluated the use of artificial intelligence (AI) in improving the efficiency of digital breast tomosynthesis (DBT) screening. It found that AI can reduce the workload of radiologists without compromising sensitivity and recall rate compared to expert interpretations.
Article
Radiology, Nuclear Medicine & Medical Imaging
Clarisse Florence de Vries, Samantha Colosimo, Moragh Boyle, Gerald Lip, Lesley Anderson, Roger Staff, iCAIRD Radiology Collaboration
Summary: This study surveyed breast screening readers in the UK on their views of incorporating Artificial Intelligence (AI) technology into breast screening mammography. The results showed that readers supported AI as a partial replacement for human readers and preferred a graphical indication of the suspected tumor area as the AI representation option. National guidelines and independent prospective studies were considered crucial evidence prior to implementing AI.
INSIGHTS INTO IMAGING
(2022)
Article
Medical Informatics
Sian Taylor-Phillips, Farah Seedat, Goda Kijauskaite, John Marshall, Steve Halligan, Chris Hyde, Rosalind Given-Wilson, Louise Wilkinson, Alastair K. Denniston, Ben Glocker, Peter Garrett, Anne Mackie, Robert J. Steele
Summary: Artificial intelligence has the potential to accurately classify mammograms and replace or supplement radiologists in breast cancer screening. Evaluations of AI systems by the UK National Screening Committee focus on maximizing benefits and minimizing harms to women. Additional information on the spectrum of disease detected and interval cancers is crucial for understanding the benefits and harms of screening.
LANCET DIGITAL HEALTH
(2022)
Article
Medical Informatics
Christian Leibig, Moritz Brehmer, Stefan Bunk, Danaiyn Byng, Katja Pinkert, Lale Umutlut
Summary: This study evaluated the performance of a decision-referral approach integrating artificial intelligence (AI) into the breast-cancer screening pathway. The results showed that this approach outperformed standalone AI systems and radiologist decisions in terms of sensitivity and specificity, indicating its potential to improve screening accuracy.
LANCET DIGITAL HEALTH
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Stefanie Weigel, Walter Heindel, Hans-Werner Hense, Thomas Decker, Joachim Gerss, Laura Kerschke, TOSYMA Screening Trial Study Grp
Summary: The TOSYMA study showed that digital breast tomosynthesis plus synthesized mammography (DBT plus SM) has a higher invasive cancer detection rate compared to digital mammography (DM) in dense breasts.
Review
Radiology, Nuclear Medicine & Medical Imaging
Tone Hovda, Kaitlyn Tsuruda, Solveig Roth Hoff, Kristine Kleivi Sahlberg, Solveig Hofvind
Summary: The study reviewed mammograms for 1225 women with screen-detected breast cancer. Results showed that visible findings on prior screening mammograms were not necessarily indicative of screening failure. Some cases were classified as missed, aiming to improve the quality of breast cancer screening.
EUROPEAN RADIOLOGY
(2021)
Review
Oncology
Tong Li, Nehmat Houssami, Naomi Noguchi, Aileen Zeng, M. Luke Marinovich
Summary: Digital breast tomosynthesis (DBT) has differential incremental cancer detection and recall rates based on breast density. While the incremental cancer detection rate is higher in high-density screens, a substantial number of additional cancers can also be detected in low-density screens.
BRITISH JOURNAL OF CANCER
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Eleanor Cornford, Shan Cheung, Mike Press, Olive Kearins, Sian Taylor-Phillips
Summary: The study found that the performance of breast screening professionals is related to the number of mammograms examined per year. With increasing volume, the recall rate decreases and the positive predictive value increases, while the cancer detection rate remains stable. There is also some impact on performance based on the different occupational groups of readers.
EUROPEAN RADIOLOGY
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Gisella Gennaro, Letizia Povolo, Sara Del Genio, Lina Ciampani, Chiara Fasoli, Paolo Carlevaris, Maria Petrioli, Tiziana Masiero, Federico Maggetto, Francesca Caumo
Summary: This study aims to improve the individual performance of breast radiographers by using automated software to assess the correctness of breast positioning and compression. The results showed that active use of the software tool significantly increased the quality of breast positioning and compression, indicating the potential to improve screening outcomes.
EUROPEAN RADIOLOGY
(2023)
Article
Biochemistry & Molecular Biology
Adam Yala, Peter G. Mikhael, Constance Lehman, Gigin G. Lin, Fredrik Strand, Yung-Liang Wan, Kevin Hughes, Siddharth Satuluru, Thomas Kim, Imon Banerjee, Judy Gichoya, Hari Trivedi, Regina Barzilay
Summary: A reinforcement learning model, Tempo, was introduced to predict risk-based follow-up recommendations in breast cancer screening. The model was trained and validated using large datasets from multiple hospitals. The results showed that Tempo combined with an image-based AI risk model outperformed current clinical practice in terms of simulated early detection. The study demonstrated the potential of AI-based risk models and agile AI-designed screening policies in improving screening programs.
Article
Radiology, Nuclear Medicine & Medical Imaging
Melissa A. Durand, Sarah M. Friedewald, Donna M. Plecha, Debra S. Copit, Lora D. Barke, Stephen L. Rose, Mary K. Hayes, Linda N. Greer, Firas M. Dabbous, Emily F. Conant
Summary: This study investigated the impact of digital breast tomosynthesis (DBT) on breast cancer screening, and found that DBT can improve sensitivity and specificity for detecting breast cancer, identifying more invasive cancers.
Article
Radiology, Nuclear Medicine & Medical Imaging
Christoph I. Lee, Linn Abraham, Diana L. Miglioretti, Tracy Onega, Karla Kerlikowske, Janie M. Lee, Brian L. Sprague, Anna N. A. Tosteson, Garth H. Rauscher, Erin J. A. Bowles, Roberta M. diFlorio-Alexander, Louise M. Henderson
Summary: This study aimed to establish performance benchmarks for digital breast tomosynthesis (DBT) screening and evaluate performance trends over time in U.S. community practice. The results showed that most radiologists achieved acceptable performance ranges for screening performance metrics with DBT.
Article
Multidisciplinary Sciences
Divya Bhandari, Akira Shibanuma, Junko Kiriya, Suzita Hirachan, Ken Ing Cherng Ong, Masamine Jimba
Summary: Breast cancer burden is increasing in low-income countries due to increasing incidence and delayed presentation. Many women in these countries do not participate in breast cancer screening despite its effectiveness. This study in Kathmandu Valley, Nepal, found poor screening behavior among women, with higher rates of breast self-examination compared to clinical breast examination and mammography. Multidimensional culturally sensitive interventions are needed to enhance screening intentions, focusing on improving attitude, family support, and addressing fatalistic beliefs towards cancer. Proper availability of screening methods and encouraging early screening before symptoms appear are also essential.
Article
Radiology, Nuclear Medicine & Medical Imaging
Kristin Johnson, Kristina Lang, Debra M. Ikeda, Anna Akesson, Ingvar Andersson, Sophia Zackrisson
Summary: Interval cancer rates were lower in DBT screening compared to DM screening, potentially indicating screening benefits. Interval cancers in the trial generally exhibited nonfavorable characteristics.