GLOBAL DERMATOLOGY ESSENTIALS: Telemedicine and AI

Presentation on artificial intelligence in dermatology given by Professor Peter Soyer. AI in dermatology is composed of computer science, mathematics, and domain expertise (dermatology). In 2017, a paper published in “Nature” by Stanford showed that AI could diagnose melanoma with accuracy comparable to expert dermatologists. A few years later, a NatureMED paper from Google Health showed a deep learning system for the differential diagnosis of skin diseases with high accuracy based on 26 common skin conditions, patient metadata, and clinical images.

Dr. Victoria Williams conducted a pilot implementation and feasibility study for Visual DX, a mobile clinical decision support tool. Healthcare in developing countries, including Botswana, heavily relies on primary care physicians and healthcare workers, who face challenges in providing early diagnosis and managing complex medical conditions due to a shortage of specialists and lack of sufficient reference material available at the point of care. Patients often have to wait weeks or months to see a specialist, which may exacerbate their condition or result in missed appointments, inability to pay, or transportation issues. Dr. Williams found it to be a helpful tool in her practice and believes it could improve access to dermatology knowledge and skills for other providers, especially those working in remote areas across the country.

Dr. Prajwal Pudasaini discusses the use of teledermatology service in government-based hospital of low-income countries. This is his second time presenting and we are excited to gain new insights from his experiences in Nepal and the Southeast Asian region, highlighting the importance of teledermatology in these developing countries

Teledermatology is a good tool to provide dermatological expertise at distance, including new healthcare professionals or processes and new skills for dermatologists. In comparison to conventional dermatology practice, it highly stands on the quality of images or information provided for a probabilistic diagnosis and management plan. This lecture will illustrate organizational challenges and clinical pitfalls related to teledermatology practice.

Dr Gkini has a special interest in inflammatory skin diseases, hair diseases, and psychodermatology. In her presentation, will discuss the application of artificial intelligence in the management of psoriatic disease. The topic starts by defining Artificial intelligence which is a term that has gained popularity in recent times, but there is currently a lack of a standard definition. The umbrella term of artificial intelligence also includes machine learning and deep learning. Deep learning refers to neural networks that have multiplayers of neurons with adjustable weights, inspired by biology. Then the presenter moves to psoriasis.