New AI Smartphone App Diagnoses STIs from Uploaded Images

New AI Smartphone App Diagnoses STIs from Uploaded Images

Facebook
Twitter
LinkedIn

Researchers at the Melbourne Sexual Health Centre and Monash University have developed a new AI-powered smartphone app capable of diagnosing sexually transmitted infections (STIs) from uploaded photos of the infected area.

As reported by AlfredHealth, the app allows users to upload photos of abnormal lesions, which are then processed using AI to detect STIs and other common genital skin illnesses. This brings hope for effortless STI diagnosis in the future.

The app’s development framework was based on a database of over 5,000 collated images gathered with the help of patient volunteers and MSHC clinicians who gave prior consent for the images of their infected areas to be used for research.

In the initial testing, the app showed an impressive accuracy of about 60 to 70 percent, with MSHC epidemiologist Professor Lei Zhang confident that the accuracy will improve with the addition of more images and metadata.

An Effective Mobile Tool for Self-screening

According to Prof Zhang, the main goal behind the innovative self-testing smartphone application is to make it publicly accessible to the community, allowing those with STI symptoms to self-screen and get accurate diagnosis by following a few steps.

Considering that many people who experience unusual lesions around their genitals often feel too shy to talk to a doctor about it, it often leads to delayed diagnosis and treatment; the new AI-powered tool allows such individuals to test themselves as soon as the symptoms appear.

Early testing leads to early diagnosis, which can help in early clinical intervention. On the other hand, a negative result can save the user’s clinical visit. According to Zhang, the smartphone tool can also provide valuable suggestions to clinicians for a better diagnosis.

While the STI app is still under development, the researchers anticipate that the technology will soon be used in multiple areas, including in melanoma detection and early detection of anal and cervical cancers, as high-resolution images of affected areas become more available.