Machine Learning Algorithm Detects Cervical Cancer

A novel colposcope uses an automated visual evaluation (AVE) algorithm to detect cervical cancer from even a single image.

The MobileODT (Tel Aviv, Israel) Enhanced Visual Assessment 3 (EVA3) system is a compact colposcope designed for durability and portability. Features include an ultra-bright powered light source with cross-polarization (to reduce glare); a complementary metal-oxide semiconductor (CMOS) sensor with 13 megapixel resolution; and a powerful 4X optical/16X digital zoom magnification lens that provides a working distance of 225-400 mm. The rechargeable, long-lasting battery provides up to 10 hours of continuous use.


Secure software allows for real-time visualization of the cervix, with enhancement filters that can be applied directly to captured images. Secure online data management allows users to document cases, add annotations, and export the information to an electronic medical record (EMR), simplifying the medical workflow. In addition, a cloud-based information system (EVA Cloud) provides secure access to real time data so as to monitor provider utilization, identify cases reviewed, collect anonymized patient statistics, and enhance quality control and quality improvement opportunities.

The EVA Colposcope is currently used in 29 countries, using smartphone technology and augmented intelligence cervical cancer detection to improve cancer identification. The augmented AVE algorithm can identify problematic lesions with greater reliability than traditional Pap cytology testing, and with a higher level of accuracy than expert human colposcopists, as validated by the U.S. National Cancer Institute (NCI, Rockville, MD, USA) and the National Library of Medicine (Bethesda, MD, USA).

“For this new technology to move out of the laboratory and into healthcare practice, a practical application was needed. The EVA System is the only colposcope on the market ready to deliver AVE at the point-of-care,” said Ariel Beery, CEO of MobileODT. “We are excited by the new AVE algorithm and the promise it holds in fighting cervical cancer. Our team is proud to make available an AVE enabled colposcope to reach more women and save more lives.”

Cervical cancer is the fourth most common cancer in women, with more than 500,000 new cases occurring annually worldwide. The two most common detection methods include the Pap smear, which can be performed by a non-specialist, and colposcopy, which requires visualization of the cervix using a speculum, a colposcope, and a trained professional to administer the test. Colposcopes and people who know how to use them are difficult to find in many low-income regions.

Related Links:
MobileODT