Smarter cervical screening: How AI is transforming the Pap smear

Artificial intelligence (AI) is helping to usher in a new era for cervical cancer screening. At MUSC Health, providers are using AI-powered technology – the Genius Digital Diagnostics System – to make reading Pap smears faster, more accurate and less labor-intensive.

MUSC is one of the first academic health systems in the U.S. to use the Food and Drug Administration (FDA)-approved system – it is now using the new technology from Hologic to examine all Pap smears analyzed at its Charleston laboratory.

“The Hologic Genius system marks a transformative step forward,” said MUSC Cytopathology Manager Patricia M. Houser. “I’m incredibly proud of our entire team and so excited we can offer this state-of-the-art technology to our patients. It truly reflects MUSC’s continued commitment to innovation and enables us to provide the highest level of quality testing.”

A digital assist

Like most medical procedures, cervical cancer screening is a team effort. It starts with a clinician, usually a gynecologist or family medicine doctor, collecting a sample from the patient’s cervix, known as a Pap test or Pap smear, and sending it to the cytology laboratory for examination. A certified cytologist then carefully examines the cells to look for abnormal ones. If any abnormal cells are detected, the sample is sent to a pathologist for final review to determine if follow-up is needed.

As the first-line screeners, cytologists have a time-intensive job. Traditionally, they manually examine every cell in a Pap smear, which can contain anywhere from 5,000 to over 100,000 cells per slide. Unsurprisingly, that process takes considerable time, and, to ensure quality, federal law limits how many Pap smears cytologists can review in a day.

 

The new system helps overcome those hurdles by drastically speeding up and automating the screening process. It scans and digitizes the slides and then uses AI to identify representative cells from the thousands collected and displays them in 36 panels on a monitor, highlighting the cells most likely to represent precancerous lesions or cervical cancer cells.

Thus, instead of needing to examine thousands of cells manually, cytologists can focus on a more limited set of potentially problematic ones. That makes them less likely to overlook abnormalities in a sea of normal cells. The result is a faster and more focused screening process that improves diagnostic accuracy. Importantly, pathologists can still make the final call using their expertise – but with far less time spent on routine review.

“The system doesn’t replace people – it helps them,” explained Jack Yang, M.D., an MUSC pathologist and champion for the new system. “By bringing suspicious cells to the front, it makes the process more sensitive and efficient – ultimately getting patients faster results.”

Doubling efficiency and fostering collaboration

The impact of the new system has been immediate. Cytologists can now review up to 200 slides in the same amount of time it previously took to read 100, effectively doubling their efficiency. That improvement is critical as many labs grapple with staffing shortages. Quicker turnaround also translates into faster results for patients who can then initiate care with the gynecologic oncologists at MUSC Hollings Cancer Center sooner, which can be critical in the fast-moving environment of cancer cells.

Equally important, the digital platform enables remote access and easier collaboration. Because the images are securely stored and transmitted via the internet, clinical staff no longer need to ship slides between hospitals; instead, the images can be shared instantly. That means multiple experts can review the same case in real time, a major advantage when dealing with second opinions or complex diagnoses.

“This technology makes it possible to centralize expertise,” Yang said. “A specialist in one location can instantly consult with another across the system.”

And, as the cytology system becomes more widely used, it could open doors for patients living in rural or underserved areas to benefit from the same high-quality cancer screening as those treated in major medical centers.

“This is just the start. With the rapid advances in technology, AI will play a very significant role in the field of cytopathology, making diagnoses easier, faster and more consistent. And, in the end, patients benefit most.”

Jack Yang, M.D.

Like any new tool, the system requires training, and the FDA mandates that cytologists receive formal instruction before using the technology. At MUSC, all staff using the system have been formally trained, either by trainers sent by the company or by going to a training site.

“Everyone who uses the system has to be retrained, even though cytologists already know the basics of Pap smears,” Yang explained. “That added some extra effort at the beginning, but the payoff is worth it. Over time, I think the process will get even easier.”

While some staff initially had concerns about missing abnormalities by relying on computer-selected cells, confidence is growing as more cytologists and pathologists use the technology. Yang noted that, while the system cannot replace expert diagnosis, it can raise the baseline level of accuracy and consistency, especially for less-experienced staff.

“If you don’t trust the flagged cells, you can always go back and review the physical slide,” he noted. “You don’t lose anything – you just gain efficiency.”

While Pap smears are leading the pack, AI has the potential to reshape many areas of medicine, from radiology to pathology.

“This is just the start,” Yang emphasized. “With the rapid advances in technology, AI will play a very significant role in the field of cytopathology, making diagnoses easier, faster and more consistent. And, in the end, patients benefit most.”

Related Links

More MUSC Hollings Cancer Center news

Subscribe to our newsletter

Hollings Horizons magazine