App that detects STIs from a “selfie” could warn when you need to see a doctor
Monash University and Alfred Health researchers have tapped into the power of artificial intelligence (AI) to develop an app that determines whether someone has a sexually transmitted infection by taking a “STI-selfie”.
By simply uploading a photo of the abnormal lesion area on their smart phone, the app will detect STIs and other common genital skin conditions. At the moment, the app has only been used in Melbourne Sexual Health Centre (MSHC, part of Alfred Health) with the presence of a clinician, which ensures the correct interpretation of the results. However, the app has demonstrated an accuracy between 60-90 per cent and researchers expect to further improve accuracy with more images collected.
The app has recently been trialled on the detection of Monkeypox by Dr Nyi Nyi Soe, a PhD candidate of Professor Lei Zhang, who is affiliated with both School of Translational Medicine, Monash University and MSHC, Alfred Health, and published in the Journal of Medical Internet Research.
Around 16 per cent of Australians have had an STI in their lifetime and, in 2020, there were about 57,500 new cases of notifiable STI among females and 67,400 among males. Alarmingly, between 2015 and 2019, STI notification rates increased by 17 per cent for chlamydia; nearly 79 per cent for gonorrhoea and 95 per cent for infectious syphilis.
The particular ‘STI-selfie’ app was based on more than 5000 STI lesion images with invaluable contributions from MSHC clinicians who collected the images and volunteers who provided consent for their lesion images to be used for research purposes.
According to Professor Zhang, the eventual goal of this web-based app is to alert a person early as to whether they should go and see their doctor when they have a skin abnormality. “Many people delay or avoid seeking testing and treatment, often due to a lack of knowledge about their own risk for developing HIV/STIs, limited availability of testing services, and the social stigma attached to them,” he said.
Professor Zhang, together with another PhD candidate Dr Phyu Mon Latt, have also led the development of a web-based tool that helps people predict their risk of developing any of the four major infections—HIV, syphilis, chlamydia and gonorrhoea. Called MySTIRisk, the tool (mystirisk.mshc.org.au) can assist individuals in assessing their risk of contracting STIs and encourage early testing and diagnosis. They used 2008–2022 data from more than 216,000 HIV, almost 228,000 syphilis, more than 262,000 gonorrhoea, and more than 320,000 chlamydia consultations at the MSHC to develop MySTIRisk.