Melanoscope AI
A research project in medical informatics and computer vision aimed at supporting physicians in the analysis of dermoscopic images of skin neoplasms.
Melanoscope AI is being developed as a clinical decision support system for tasks related to early melanoma detection and differential diagnosis of skin neoplasms using machine learning methods, dermoscopy, and clinical expertise.
Project video
Why it matters
Focus on physician support
The project treats AI algorithms as a tool for supplementary analysis of dermoscopic images rather than as a substitute for clinical decision-making by a specialist.
Based on research activity
The development is accompanied by publications, dataset analysis, comparison of machine learning models, and dissertation work in medical informatics.
Connected with practical scenarios
The project is developed with preventive examinations, patient routing, and the use of mobile dermoscopy in clinical settings in mind.
Who it is for
Clinics and medical centers
For assessing the potential role of AI-based support in primary examinations, preventive programs, and digital patient routing.
Physicians and specialists
For using computational analysis as an additional source of information when working with dermoscopic images.
Research and engineering teams
For joint research in medical informatics, computer vision, biomedical data analysis, and digital diagnostic tools.
How the project works
Data formation and description
Dermoscopic images of skin neoplasms are used together with metadata and clinically significant features.
Models and computational analysis
Both multiclass and binary classification approaches are studied, including neural network architectures and visual transformers.
Interpretation in clinical context
The analytical output is treated as an element of clinical decision support and requires interpretation by a specialist.
Why it is trustworthy
The site presents articles and dissertation materials related to dermoscopy, image classification, and medical informatics.
The development has been demonstrated and used in the context of Melanoma Days, professional seminars, and early diagnosis activities.
The project is presented at conferences, exhibitions, and industry events where digital medicine and medical AI are discussed.
Project at a glance
The core task of the project is related to image analysis of skin neoplasms and support for early diagnostic workflows.
The research component of the project is reflected in scientific articles, reviews, dataset-related work, and dissertation materials.
The project is being developed with preventive examinations, first-line consultation scenarios, and mobile dermoscopy in mind.
The development is regularly presented at conferences, exhibitions, and discussions devoted to digital medical technologies.
Latest news
Third Scientific Council of the Center for AI in Medicine
On July 29, the Almazov National Medical Research Centre hosted the third meeting of the Scientific Council of the Centre for Artificial Intelligence in Medicine, bringing together …
PyCon Russia 2026
Staff members from the Ivannikov Institute for System Programming of the Russian Academy of Sciences (ISP RAS) participated in the exhibition at the PyCon Russia 2026 conference, …
Melanoma Day, June 29
On June 29, 2026, a Melanoma Day event was held as part of a nationwide initiative for the early detection of skin neoplasms.
The diagnostic process utilized the Melanoscope AI …
ISP RAS presented the Melanoscope AI system during a business mission to South Africa.
Staff members from ISP RAS participated in a business mission organized by the Russian Export Center in Johannesburg, South Africa. During the event, they showcased the Institute’s …
An innovative approach to the analysis of dermoscopic images was presented at the Ivannikov Readings
On May 28–30, 2026, the international conference “Ivannikov Readings” was held - a significant event in scientific life, organized in memory of Academician V. P. Ivannikov.
During …
Publications
Let’s discuss collaboration
If you are interested in research collaboration, discussion of clinical use scenarios, pilot projects, or professional interaction related to Melanoscope AI, please get in touch.