Recently, a new study revealed that a computer algorithm was not only able to accurately analyze digital images drawn from cervical screenings but also detect precancerous changes that required more medical follow-up.
The new technique dubbed automated visual evaluation boasts the ability to transform point-of-care cervical screening.
“Our findings show that a deep learning algorithm can use images collected during routine cervical cancer screening to identify precancerous changes that, if left untreated, may develop into cancer,” Mark Schiffman, MD, MPH, of the National Cancer Institute’s Division of Cancer Epidemiology and Genetics, and senior author of the study said in a press release.
“In fact, the computer analysis of the images was better at identifying precancer than a human expert reviewer of Pap tests under the microscope (cytology).”
SEE MORE: UK’s National Health Service Trialing AI Software for Diagnosing Cancer
SEE MORE: AI Able to Differentiate Between Different Cancer Cells
The release also showed that the artificial intelligence-based approach is easy to perform.
In fact, health workers can utilize their cell phones or similar camera devices for cervical screening and treatment in one visit.
Since minimal training is required to carry out the technique, it is suitable for nations that have limited healthcare resources, whereby cervical cancer is the top cause of sickness and death among women.
To develop the algorithm, both Schiffman and his colleagues leveraged the information gathered from population-based longitudinal cohort research involving 9,406 women aged between 18 and 94 years in Guanacaste, Costa Rica.
For seven years, women underwent, numerous histopathologic confirmation of precancers and cervical screening techniques.
SEE MORE: Top 10 Ways Artificial Intelligence is Impacting Healthcare
A tumor registry helped in linking detected cervical cancers to a maximum of 18 years from particular research.
Over 600,000 stored, digital cervical images that were taken during screening were utilized for validation and training of the deep learning-driven algorithm.
The automated visual evaluation technique outperformed all standard screening techniques when it was used in detecting cervical cancer cases diagnosed in the course of the Costa Rica study.
The AI approach had better accuracy (AUC=0.91) compared to conventional cytology or human expert review, which attained AUC= 0.71 and AUC= 0.69 respectively.
Although he was not part of the study, the President of Cancer Treatment Centers of America Maurie Markman told Cancer Network that the release was a provocative report that uses modern technology in dealing with a crucial cancer screening matter in limited-resource settings.
“While additional work is required to optimize this deep learning algorism employing an automated evaluation of images of the cervix the potential exists that this strategy could substantially favorably impact cervix cancer-associated morbidity and mortality in these clinical settings,” he said.
