# Radiomics analysis for the early diagnosis of common sexually transmitted infections and skin lesions

## TL;DR

An AI tool using radiomics and machine learning can help identify STIs from skin lesion images, with accuracy improving when infection location is considered.

## Contribution

A novel AI method combining radiomics and machine learning for early STI diagnosis from skin lesion images, with performance improvements when anatomical site data is included.

## Key findings

- Including infection site information improved model performance by 22.3% for anal infections and 3.8% for skin infections.
- Lesion texture and statistical radiomics features were most predictive for STIs.
- The best model achieved an average AUC of 0.681 when infection site information was unspecified.

## Abstract

Early identification of sexually transmitted infection (STI) symptoms can prevent subsequent complications and improve STI control. We analysed 597 images from STIAtlas and categorised the images into four typical STIs and two skin lesions by the anatomical sites of infections. We first applied nine image filters and 11 machine-learning image classifiers to the images. We then extracted radiomics features from the filtered images and trained them with 99 models that combined image filters and classifiers. Model performance was evaluated by area under curve (AUC) and permutation importance. When the information of infection sites was unspecified, a combined Gradient-Boosted Decision Trees (GBDT) classifier and Laplacian of Gaussian (LoG) filter model achieved the best overall performance with an average AUC of 0.681 (95% CI 0.628-0.734). This model predicted best for lichen sclerosus (AUC = 0.768, 0.740-0.796). The incorporation of infection site information led to a substantial improvement in the model’s performance, with 22.3% improvement for anal infections (AUC = 0.833, 0.687-0.979) and 3.8% for skin infections (AUC = 0.707, 0.608-0.806). Lesion texture and statistical radiomics features were the most predictive for STIs. Combining machine learning and radiomics techniques is an effective method to categorise skin lesions associated with STIs clinically.

We developed an artificial intelligence tool that can help identify sexually transmitted infections (STIs) from photographs of skin lesions. Using machine learning and a technique called radiomics—which extracts detailed information about texture and shape from medical images—we analysed 597 images from the STI Atlas database covering four common STIs and two skin conditions. Our approach combines computer algorithms with radiomics to automatically detect features in skin images that might indicate specific infections. We found that when we included information about where on the body the infection appeared (genitals, anus, or other skin areas), our tool’s accuracy improved significantly. The biggest improvement was for anal infections, where accuracy increased by over 22%. This technology could be particularly valuable in areas with limited access to healthcare specialists, allowing people to take photographs with their smartphones for preliminary assessment. While not intended to replace clinical diagnosis, our tool could help people decide whether they need urgent medical attention. This aligns with global health efforts to improve early detection and treatment of sexually transmitted infections, potentially reducing transmission and complications in communities worldwide.