Study Finds Robots can Detect Breast Cancer as well as Radiologists

Breast cancer is the most common cancer in women, and despite important improvements in therapy, it is still a major cause for cancer-related mortality, accounting for approximately 500,000 annual deaths worldwide. Breast cancer screening programs using mammography are effective in reducing breast cancer-related mortality. However, current screening programs are highly labor intensive due to the large number of women that have to be screened. Considering the increasing scarcity of breast screening radiologists in some countries, many researchers believe other screening methods may be worth investigating.

A new paper in the Journal of the National Cancer Institute suggests that artificial intelligence systems may be able to perform as accurately as radiologists in the evaluation of digital mammography in breast cancer screening.

Since the 1990s, computer-aided detection systems have been developed to detect and classify breast lesions in mammograms automatically. However, no studies to date have found that these systems directly improve screening performance or cost effectiveness. This has precluded their use as a method for screening mammography.

In this study researchers compare, at a case level, the cancer detection performance of a commercially available AI system to that of 101 radiologists who scored nine different cohorts of mammography examinations from four different manufacturers as part of studies previously performed for other purposes.

Each dataset consisted of mammography exams acquired with systems from four different vendors, multiple radiologists' assessments per exam, yielding a total of 2,652 exams (653 malignant) and interpretations by 101 radiologists (28,296 independent interpretations).

The performance of the artificial intelligence system was statistically not-inferior to that of the average of the 101 radiologists. The evaluated system achieved a cancer detection accuracy comparable to an average breast radiologist in this retrospective setting.

"Before we could decide what is the best way for AI systems to be introduced in the realm of breast cancer screening with mammography, we wanted to know how good can these systems really be," said Ioannis Sechopoulos, one of the paper's authors. "It was exciting to see that these systems have reached the level of matching the performance of not just radiologists, but of radiologists who spend at least a substantial portion of their time reading screening mammograms."

Alejandro Rodriguez-Ruiz, Kristina Lång, Albert Gubern-Merida, Mireille Broeders, Gisella Gennaro, Paola Clauser, Thomas H Helbich, Margarita Chevalier, Tao Tan, Thomas Mertelmeier, Matthew G Wallis, Ingvar Andersson, Sophia Zackrisson, Ritse M Mann, Ioannis Sechopoulos.
Stand-Alone Artificial Intelligence for Breast Cancer Detection in Mammography: Comparison With 101 Radiologists.
JNCI: Journal of the National Cancer Institute, djy222. doi: 10.1093/jnci/djy222.

Most Popular Now

Open Call HORIZON-JU-IHI-2022-03-05: Dig…

Mental health disorders represent an area of severe unmet public health need. This has been further negatively impacted by the COVID-19 pandemic, with a substantial increase in the number and...

The Future of Medicine is Data

At the 2023 Annual J.P. Morgan Healthcare Conference, Owkin Co-founder and CEO Thomas Clozel, MD will outline how data is the future of medicine - from the development of new...

Young People Want Trustworthy Mental Hea…

Young people may not be receiving the mental health support they need because of a lack of trust in the smartphone apps that deliver such services, a study suggests. Questions about...

Brain Area Necessary for Fluid Intellige…

A team led by UCL and UCLH researchers have mapped the parts of the brain that support our ability to solve problems without prior experience - otherwise known as fluid...

Study Surveys Landscape of Apps Built on…

A study led by Regenstrief Institute Research Scientist Titus K. Schleyer, DMD, PhD, is among the first to survey the current landscape of FHIR® apps, providing a snapshot of how...

New Computer Program 'Learns' to Identif…

Genetic mutations cause hundreds of unsolved and untreatable disorders. Among them, DNA mutations in a small percentage of cells, called mosaic mutations, are extremely difficult to detect because they exist...

Allscripts Announces Corporate Name Chan…

Allscripts Healthcare Solutions, Inc. announced that, effective January 1, 2023, it has changed its name to Veradigm Inc. (NASDAQ: MDRX). Allscripts had been transitioning its solutions to the Veradigm brand...

Applications Open for SpinLab Accelerato…

The start-up accelerator supports entrepreneurial and innovative teams that want to grow sustainably and successfully scale their business model. With a strong hands-on mentality and a lot of passion, the...

Bayer to Accelerate Drug Discovery with …

Bayer AG and Google Cloud today announced a collaboration to drive early drug discovery that will apply Google Cloud's Tensor Processing Units (TPUs), which are custom-developed accelerators designed to run...

220M€ Investment in Testing and Experime…

To make the EU the place where AI excellence thrives from the lab to the market, the European Union is setting up world-class Testing and Experimentation Facilities (TEFs) for AI. Together...

Artificial Nerve Cells - Almost Like Bio…

Researchers at Linköping University (LiU), Sweden, have created an artificial organic neuron that closely mimics the characteristics of biological nerve cells. This artificial neuron can stimulate natural nerves, making it...

AI Tool Developed to Predict Risk of Lun…

Lung cancer is the leading cause of cancer death in the United States and around the world. Low-dose chest computed tomography (LDCT) is recommended to screen people between 50 and...