Melanoscope AI
Download catalogue of technologies
Melanoscope AI
Melanoscope AI - the system for early detecting melanoma and other malignant skin neoplasms based on analysis of dermatoscopic images of melanocytic skin neoplasms.
Melanoscope AI detects malignant skin neoplasms at early stages by obtaining a preliminary diagnosis by non-specialist at first-line medical institutions.
Relevance
In 2020 due to untimely diagnosis of the disease the mortality rate was more than 20 percent.
According to the statistics of cases of melanoma in 2020 and the forecast of melanoma incidence by researchers from IARC (International Agency for Research on Cancer) for 2040.
Key indicators
The advantages of our technology: high diagnostic accuracy, extensive data and local adaptation for maximum efficiency.
90%
accuracy of diagnosis, which guarantees high reliability and minimizes the risk of errors
> 35 000
clinical cases are included in deep learning has been used in an algorithm that could detect melanoma in images which roughly the same accuracy seasoned professional dermatologists
> 500
unique images taking into account the skin phenotype of the population of the country, which allows us to adapt algorithms to specific features and provide more accurate results for patients
How Melanoscope AI Works
Visual example of melanoma diagnosis using Melanoscope AI on a computer and mobile device.
Examples of detected melanomas
with confirmation of diagnosis Melanoscope AI
Benefits of technology
- High accuracy. Achieving accuracy comparable to professional dermatologists.
- Speed. Quick analysis and obtaining the result.
- Availability. Can be used on different devices, including smartphones and tablets.
- Efficiency. AI doesn’t get tired, which allows it to operate without fatigue-related errors, providing consistent and high-quality diagnostics in various conditions.
System structure
The system for early detection of melanoma and skin cancer based on the analysis of dermatoscopic images of melanocytic skin neoplasms has the following modules:
- Collection module dermatoscopic images of melanocytic lesions to form a data set.
- Training module artificial intelligence models.
- Marking system dermatoscopic images of melanocytic lesions.
- Classification module dermatoscopic images of melanocytic lesions based on modern neural network architectures.
For whom
The system is intended for use by junior medical personnel in private and public first-level medical institutions with a dermatoscope to form a preliminary diagnosis.
Application process
Back to the list of technologies of ISP RAS 