Artificial Intelligence (AI) is one of the hottest topics in healthcare currently. The subject is permeating most industries, and breast imaging is no different. Many may wonder how the use of AI will affect their breast program’s current policies and procedures.
AI and Skin Markers: Systematic Skin Marking is Essential
AI is quickly being adopted to aid radiologists and act as a second set of eyes while reading a mammogram. Implementing AI within your breast center can be very challenging if you use incorrectly shaped skin markers for areas of clinical concern and/or landmarks. Using the incorrect marker may mislead the AI engine. For example, if a pellet or BB marker is placed on a palpable mass in the retroareolar area, the algorithm may assume that the density adjacent to the BB is a nipple out of profile rather than a true lesion that requires follow-up.
Tomosynthesis excels at visualizing architectural distortion related to a prior surgery. Often, the architectural distortion has not been visualized on previous 2D mammograms, going back five, sometimes ten years.1 If a surgical scar is not marked with a scar marker, the AI engine may flag the area as a potential cancer and interrupt the radiologist’s workflow as they investigate the patient’s history and/or years-old images. Ultimately, the patient may be recalled.
The ACR’s Data Science Institute has collected use cases to support the implementation of artificial intelligence within breast centers. They state, "Our use cases help radiologists and allied professionals by ensuring that AI tools provide needed information, can be efficiently implemented into daily workflow, and have the potential to improve the quality and efficiency of patient care." The mammography skin marker use case that AI developers look toward clearly identifies which shape should be used for nipples, scars, and moles. Implementing this type of skin marking system can avoid confusion and potential callbacks.
ACR Practice Parameters: Current Recommendations on Skin Marking
According to a mammography survey2 conducted in 2023-2024, the prevalence of 3D imaging in breast centers across the U.S. has increased to 92%. With that increase, the ACR has amended its practice parameters for the use of skin markers to include suggested use of radiographically distinct markers for skin lesions, palpable masses, and surgical scars. Using distinct and specific shapes to mark these applications decreases unnecessary callbacks and increases communication between the technologist and radiologist. Skin marking in mammography has remained essential, with the majority of the country performing 3D mammograms.
Renowned radiologist Dr. Michael Linver explains why the use of radiographically distinct skin markers is important in the age of tomosynthesis: "If there are two areas of clinical concern in the same breast, and identical skin markers were applied to both areas, then localization of each separate region on the mammogram could be compromised, and the wrong marker applied to the wrong lesion. This could lead to incorrect characterization of the areas of concern on the mammogram, potentially allowing cancer or cancers to be missed."
Furthermore, using a defined and consistent approach to skin marking has been shown to reduce the radiologist’s reading time by 1 minute and 34 seconds per patient3. While a minute and a half may not seem like a lot, when potentially reading 50 -100 mammograms a day, 75-150 minutes could be spent elsewhere, like interacting with your patients. According to the National Institute of Health, the more a patient interacts with their provider, the higher their patient satisfaction.
An Easy, Cost-Effective Solution
Since the Data Science Institute has provided a skin marking use case for the development of AI datasets, the best practice is to follow the ACR’s practice parameters and use different and correctly shaped skin markers for mammography. For those facilities that have not yet implemented AI – making an easy change by implementing skin markers today will be an investment in your future.
To learn more about TomoSPOT® skin markers for 3D mammography or to receive samples, contact your Account Manager at 800.233.5539 or info@beekley.com.
1Gruen, David, 2016. Benign Intraparenchymal Scarring in the DBT Era
2Based on a 2023-2024 anonymous market survey of mammography facilities in the United States.
3Merkulov, Alex, 2014. What Happens when a Consistent Versus Sporadic Use of a Five Shape Breast Skin Marking System is Used in Digital Breast Tomosynthesis (DBT)?
Related articles:
Colleen O'Flaherty
Product Manager - Mammography and Breast Biopsy
