After many years treating vulvovaginal disorders, gynaecologist Ahinoam Lev-Sagie observed that vaginitis was frequently misdiagnosed, causing patients like hers needless suffering and more serious consequences. Vaginitis includes vaginal disorders caused by infection or inflammation. Lev-Sagie asked two men close to her to find a solution: husband Meny, a software engineer, and brother Nimrod Lev, a mechanical engineer. “She said: ‘You two have so much experience with medical devices. Do something about vaginitis diagnosis’,” recalls Lev, chief executive and co-founder of Gyntools, the Israeli start-up that coordinated EU project VagX. The family invented a device for gynaecologists and GPs, called Gyni. A feasibility test funded through the EU project has shown the tests to be 95 % accurate. Thanks to the market development carried out with its EU partners, including AiM and Stratex, the devices could be sold in Europe by the end of this year and in the United States (US) and other regions by the third quarter of next year, forecasts Gyntools.
“Gyni will allow caregivers to provide a correct diagnosis of the vaginitis condition at the first clinic visit within five minutes,” says Lev. Currently about one in three women suffers from vaginitis at least once in their lifetime, 40-50 % of first clinic visits are misdiagnosed and about 30 % of women with vaginal complaints remain undiagnosed even after extensive testing, reports Medscape. That often means women don’t receive timely treatment for complaints, risking reduced fertility, sexually transmitted infections and premature deliveries or low birth weight babies in pregnant women. The Gyni devices work using automated microscopy, pH measurements and the consideration of patient symptoms in an automated diagnostic process carried out through deep learning. Gyni provides six different diagnoses, including mixed conditions, just by taking a single vaginal swab.
Bettering the expert
The test results were so positive that gynaecologists and specialists in Israel have joined Gyntools’ seed investment round. The study also sparked interest from European specialists when presented in November at the III ISIDOG Congress from 31 October to 2 November. The study collected 5 000 microscopy images from 300 patients, which were classified by a specialist. A total of 85 % of them were used to train Gyni’s software model; the rest were used for testing. In one case, the developers noticed the Gyni device’s diagnosis did not match the experts’ one. However, when the expert reviewed the images taken by the Gyni device, she agreed its mixed infection diagnosis was better than her single condition diagnosis. “Modern deep learning image analysis results at times even surpass human performance,” adds Lev. Gyntools is now applying for ISO certification, completing its clinical trial in Israel and hoping to gain a CE mark during the third quarter of this year. It plans a larger trial in the EU, United Kingdom and the US, with a view to FDA clearance. “Women will no longer suffer for a long time from misdiagnosed vaginitis and incorrect treatments, will suffer fewer treatment side effects and will get a faster improvement in their quality of life,” concludes Lev.
VagX, vaginitis, vaginal infections, gynaecologists, deep learning, Gyni, Gyntools