CLINICAL AI

Sexual health remains one of the most challenging global public health issues, with accessibility and early detection serving as critical barriers to effective disease management. The recognition of Professor Lei Zhang's groundbreaking research in the 2025 Clunies Ross Technology Innovation Award from the Australian Academy of Technological Sciences and Engineering underscores the transformative potential of artificial intelligence in addressing this persistent challenge. Zhang's development and commercialization of AI-assisted diagnostic tools for sexually transmitted infections (STIs) and cervical cancer screening represents a pivotal moment in how healthcare systems can leverage technology to democratize access to sexual health services.
The clinical need for Zhang's innovations cannot be overstated. With approximately 1 million STI diagnoses occurring globally every day, current diagnostic approaches remain constrained by limited access to specialized venereologists, dermatologists, and sexual health clinicians in many regions. Traditional STI screening relies heavily on clinical expertise and laboratory resources that are often unavailable or economically inaccessible to vulnerable populations. AI-powered image recognition systems address this gap by enabling preliminary diagnostic assessment through visual analysis of genital lesions, combined with patient-reported symptoms and clinical metadata. This multimodal approach has demonstrated significantly improved diagnostic accuracy compared to image analysis alone, with integrated models achieving area under curve (AUC) values of 0.893 compared to 0.859 for image-only analysis.
Zhang's research team developed deep learning models trained on thousands of clinical images collected at the Melbourne Sexual Health Centre, achieving remarkable performance across multiple STI presentations. The integration of clinical metadata—including patient demographics, symptom duration, and sexual history—with image analysis substantially enhanced diagnostic discrimination, improving model performance by up to 6.7% compared to baseline image-only approaches. This methodological advancement reflects a sophisticated understanding of how clinical context enriches algorithmic decision-making in dermatological diagnostics. Furthermore, existing AI platforms like HeHealth have demonstrated diagnostic accuracies of 86% for syphilis, 93% for genital herpes, and 96% for genital warts, establishing proof-of-concept for scalable screening solutions.
The cost-effectiveness implications of AI-assisted screening are equally compelling. Economic modeling demonstrates that AI-assisted liquid-based cytology for cervical cancer screening, when implemented at five-year intervals, could achieve comparable cost-effectiveness to human papillomavirus DNA testing while reducing healthcare provider burden. These findings suggest that AI technologies are not merely clinically effective but economically justifiable for health systems grappling with resource constraints. Importantly, Zhang's vision extends beyond technological innovation to encompass international implementation and capacity building. His research centers across Australia and China, combined with mentorship of early-career researchers from more than 10 developing countries, reflect an explicit commitment to translating these innovations into real-world impact across diverse healthcare contexts.
The recognition of this work comes at a critical juncture when healthcare systems worldwide are seeking solutions to reduce diagnostic disparities and improve sexual health equity. AI-assisted screening tools have particular value for populations experiencing stigma or geographic barriers to traditional sexual health services, enabling confidential preliminary assessment before clinical consultation. As digital health technologies continue advancing, Zhang's pioneering research provides a compelling model for how epidemiologically-informed, clinically-validated artificial intelligence can address fundamental access challenges in sexual health diagnostics while maintaining rigorous scientific standards and patient-centered implementation approaches.