New article on artificial intelligence in diagnostic imaging
Artificial intelligence (AI) has potential as a tool that can increase the efficiency and quality of diagnostic imaging after breast cancer screening.
Today, screening for breast cancer consists of a standardised high-quality X-ray of each breast, which is reviewed by two mammary radiologists, and in case of disagreement, a third review is performed. It is inherently resource-intensive, and therefore the potential to use AI in the examination of the images is being investigated in a new research project.
The article describes how artificial intelligence (AI) has the potential to make breast cancer diagnostics more efficient in screening. Modern technology, i.a. image analysis algorithms and new graphics processing units enable human-level performance to be achieved, and commercial AI systems are already in use worldwide.
AI thus constitutes a promising element in terms of relieving the burden on the screening programmes without compromising on quality, and in the long term AI may be able to replace the one radiologist in the examination of images. With its organised databases, Danish mammography screening has good breeding grounds for clinical validation of AI solutions prior to implementation.
The article was written by Mohammad Talal Elhakim, Ole Graumann, Lisbet Brønsro Larsen, Mads Nielsen and Benjamin Schnack Rasmussen. Read the article in Danish.