When he begins to take an interest in health issues “forsaken” of women, the president of Ziwig, Yahya El Mir, quickly understands that endometriosis wins the prize. Within this biotech, which already employs around fifty people in Lyon, we are “amazed” by diagnostic error and its consequences on the lives of patients: “Waiting on average between eight and twelve years, depending on the country, before being able to detect the presence of a pathology that affects millions of women seems completely unreal in 2022! »
→ ANALYSIS. “This pain has been silent for too long”: endometriosis, how to take care of women
How to explain this interminable wait? Beyond the lack of training of health professionals, this delay is due to the diversity and complexity of the disease, the symptoms of which can be confused with other pathologies, but above all to detection tools – MRI or endovaginal ultrasound. – which are not suitable or sometimes considered too “invasive”.
Finding a good test
After first developing a questionnaire capable of « presay » a positive or negative result thanks to artificial intelligence, the Ziwig team, which combines technology engineers, scientists and health professionals, is embarking on a biological shift.
For several years now, researchers around the world have been hoping to develop a high-performance diagnostic test by focusing on a few microRNAs known to be involved in pathology. “We have therefore chosen to consider all of the 2,600 microRNAs of the human species, most of which are present in saliva, and to analyze them with precision using an unprecedented technique: NGS (new generation sequencing, Editor’s note) », explains Professor François Golfier, who has joined the project.
→ DEBATE. Can the endometriosis plan put an end to diagnostic wandering?
That is, for each woman, millions of data to sift through. “The human mind being incapable, we have exploited the new possibilities of artificial intelligence, with machine learning (automatic learning, Editor’s note) », continues the head of the gynecological-obstetric surgery department at the Lyon Sud hospital center.
Available next year.
The algorithms allow such a precise analysis that the team manages to isolate 109 micro-RNAs involved in the physiopathology of endometriosis in order, finally, to establish a ” signature ” precise of the disease. More concretely, a woman could simply take a sample of her saliva before sending it to an equipped laboratory, to get a result in less than ten days, “Is enough to revolutionize the care, and therefore the terrible evolution of the disease”enthuses Yahya El Mir.
→ READ. Endometriosis: the long battle of associations
On the initial sample of 200 women, the reliability displayed exceeds 98%, although an external validation study – currently being carried out on 1,000 patients in several French centers – is still necessary before considering marketing.
This step should also serve to improve the performance of the assay in detecting, for example, “of an early form or a risk of infertility”, says the company. In France, where discussions are underway with the High Authority for Health in order to obtain reimbursement by Social Security, the “endotest” could be available as early as next year.
We would love to give thanks to the writer of this article for this awesome content
Endometriosis: the “endotest”, a promising diagnostic tool
Visit our social media profiles and other pages related to themhttps://www.ai-magazine.com/related-pages/