Lab manager Martin Muinde scans pap smears at the Kinondo Clinic in Kenya.
Lab manager Martin Muinde scans pap smears at the Kinondo Clinic in Kenya.
Source: Oscar Holmström

AI method can detect precursors to cervical cancer

Using artificial intelligence and mobile digital microscopy, researchers hope to create screening tools that can detect precursors to cervical cancer in women in resource-limited settings. A study led by researchers at Karolinska Institutet and University of Helsinki now shows that AI screenings of pap smears carried out with portable scanners were comparable to analyses done by pathologists.

“Our method enables us to more effectively discover and treat precursors to cervical cancer, especially in low-income countries, where there is a serious lack of skilled pathologists and advanced laboratory equipment,” says corresponding author Johan Lundin, professor at the Department of Global Public Health, Karolinska Institutet.

In countries with national screening programmes designed to detect cell abnormalities and human papillomavirus (HPV) in cervical samples, the number of cases of cervical cancer has dropped dramatically. Despite this, the global case total is expected to increase in the coming decade, largely due to shortages of screening resources and HPV vaccines in low-income countries.

Need for innovative solutions

Innovative diagnostic solutions that take into account local conditions and constraints are needed if more women around the world are to be offered gynaecological screening.

For this study, the researchers trained an AI system to recognise cell abnormalities in the cervix, which when detected early can be successfully treated. Smears were taken from 740 women at a rural clinic in Kenya between September 2018 and September 2019. The samples were then digitalised using a portable scanner and uploaded via mobile networks to a cloud-based deep-learning system (DLS). Just under half of the smears were used to train the programme to recognise different precancerous lesions while the remainder were used to evaluate its accuracy.

The AI assessment was then compared with that made by two independent pathologists of the digital and physical samples. The study shows that the assessments were very similar. The DLS had a sensitivity of 96–100 percent as regards identifying patients with precancerous lesions. No patients with more serious high-grade lesions received a false-positive assessment. As regards identifying smears without lesions, the DLS made the same assessment as the pathologists in 78–85 percent of cases.

Could free up time and resources

The researchers believe that the method can be used to exclude a majority of smears, which would free up time for local experts to examine the ones that stick out. Before this can happen, however, more research is needed on larger and more diverse patient groups, including more smears and different types of lesions as well as biopsies with confirmed precursors to cervical cancer.

“With the portable online microscope, the DLS can act as a ‘virtual assistant’ when screening for cervical cancer,” Lundin explains. “The AI assistant can be accessed globally 24/7 and help local experts examine many more smears. This method will make it possible for countries with limited resources to provide their population with screening services much more efficiently and at a lower cost than is currently the case.”

The results are published in the journal JAMA Network Open.

Subscribe to our newsletter

Related articles

AI finds patterns of mutations in tumour images

AI finds patterns of mutations in tumour images

Researchers have developed an AI algorithm that uses computer vision to analyze tissue samples from cancer patients.

Lab-on-a-chip detects cancer less invasively

Lab-on-a-chip detects cancer less invasively

A new ultrasensitive diagnostic device could allow doctors to detect cancer quickly from a droplet of blood or plasma, leading to timelier interventions and better outcomes for patients.

AI, holographic microscopy beat scientists at analyzing immunotherapy​

AI, holographic microscopy beat scientists at analyzing immunotherapy​

AI is helping researchers decipher images from a new holographic microscopy technique needed to investigate a key process in cancer immunotherapy “live” as it takes place.

Lung cancer: AI could predict risk of recurrence

Lung cancer: AI could predict risk of recurrence

Computer scientists working with pathologists have trained an AI tool to determine which patients with lung cancer have a higher risk of their disease coming back after treatment.

AI model accurately classifies colorectal polyps

AI model accurately classifies colorectal polyps

An AI model for automated classification of colorectal polyps could benefit cancer screening programs by improving efficiency, reproducibility, and accuracy.

AI taps human wisdom for better cancer diagnosis

AI taps human wisdom for better cancer diagnosis

A new system combining artificial intelligence with human knowledge promises faster and more accurate cancer diagnosis.

Using AI to detect and grade prostate cancer

Using AI to detect and grade prostate cancer

Researchers have developed a method based on artificial intelligence for histopathological diagnosis and grading of prostate cancer.

AI tracks down acute myeloid leukemia

AI tracks down acute myeloid leukemia

Artificial intelligence can detect one of the most common forms of blood cancer—acute myeloid leukemia (AML)—with high reliability.

AI identifies features associated with cancer recurrence

AI identifies features associated with cancer recurrence

An AI has successfully found features in pathology images from human cancer patients, without annotation, that could be understood by human doctors.

Popular articles

Subscribe to Newsletter