Breast cancer is the most diagnosed cancer in the United States, with over 200,000 new cases each year. 

Mammograms, which are X-rays of breast tissue, are routinely used to screen patients for breast cancer. 

After the radiologist checks the X-ray, they may order more testing if needed. If the radiologist sees an unusual spot on the X-ray, surgery may be needed to remove the area. This is called a biopsy. 

The lab will create a microscopic slide from this tissue for the pathologist to review. Pathologists may look at 200-300 slides under the microscope each day. While incorrect diagnoses by pathologists are rare, 1-2% of cases may be given the wrong diagnosis [1].

The use of artificial intelligence (AI) to help reduce the work of pathologists is being explored by several labs. AI Systems are machines that replace the need for humans and can help doctors with their daily tasks. 

The most common AI system in the pathology laboratory is a digital slide scanner. Doctors use digital slide scanners to share tissue images with others worldwide without sending valuable slides through the mail [2]. 

Preparing the digital slide scanner requires loading a set of slides onto the machine, and creating a detailed picture of the specimen. These machines can be “trained” using a large set of existing slides that several pathologists have already evaluated. 

Machine software teaches the machine how to find key features in these slide sets that are normal or diseased. Based on this first set of data, the machine will then be able to recognize the same features in new tissues. As more biopsy data is added to the system, the machine becomes better at reading slides [3]. 

Newer pathologists may benefit the most from this technology as they are more likely to miss important details in a specimen. 

About 40% of women in the United States have dense breast tissue, which can be difficult to diagnose since these normal tissue areas can mimic cancer on an X-ray. These patients are often asked to return to the doctor’s office for more screening or testing. 

AI systems may be able to separate those patients into groups that need more testing versus those that do not [4]. 

While AI systems are promising, their use is limited in the United States until more data becomes available. 

Sources:

1. Dileep G, Gianchandani Gyani SG. Artificial Intelligence in Breast Cancer Screening and Diagnosis. Cureus. 2022;14(10):e30318. Published 2022 Oct 15. doi:10.7759/cureus.30318 
2. Madabhushi A, Doyle S, Lee G, et al. Integrated diagnostics: a conceptual framework with examples. Clin Chem Lab Med. 2010;48(7):989-998. doi:10.1515/CCLM.2010.193
3. Pallua JD, Brunner A, Zelger B, Schirmer M, Haybaeck J. The future of pathology is digital. Pathol Res Pract. 2020;216(9):153040. doi:10.1016/j.prp.2020.153040
4. How GE Healthcare and Intel and AI are Optimizing Cancer Screening of Dense Breast. GE Healthcare. Accessed January/ 01/2023. https://www.gehealthcare.com/

Thank you for reading Patient Education Essentials, the Write Shift RN blog.

Disclaimer: This article was written as a guest post for Write Shift RN LLC's blog. The information in it may not be wholly fact-checked or edited, allowing the reader to see the writer's work and skills firsthand. This information is not intended as medical advice. It is for informational and educational purposes only. Always talk to your doctor or other qualified healthcare providers about any questions or concerns you may have regarding medical conditions.