What kind of examination
does it provide?
Our diagnostic system allows for non-invasive testing. Doctors photograph the affected area with a digital camera,
placing no burden on the patient. The software installed on the computer then analyzes these images, assisting
doctors in diagnosing the disease. In the workflow from capturing images to diagnosis, our system plays a crucial
role in image extraction and analysis.
This software provides support for doctors in diagnosing dermoscopic images of nail pigmentation streaks. Despite the current trend of
employing AI technologies for image analysis, recent studies from the European Dermatology AI Diagnosis Research Group, utilizing over
50,000 cases, suggest AI's recognition ability is relatively low for diagnosing subungual melanoma.
Instead of relying on AI, our software quantifies the diversity of nail color tones based on RGB values in dermoscopic images, and identifies
subungual melanoma through an objective differential index (Di). As a result, our software's Area Under the Curve (AUC) when using the
final diagnosis noted in the medical record as the gold standard exceeded previously reported values using AI algorithms. Furthermore, our
research shows that the software's ability to distinguish between benign and malignant conditions surpasses that of qualified dermatologists.
Current definitive diagnosis for suspected nail pigmentation streaks involves total nail avulsion biopsy under local anesthesia. This invasive
test often results in the semi-permanent loss of the nail. Given the rarity of subungual melanoma, dermatologists find it challenging to gain
experience in assessing this disease. Our software offers an objective Di value, addressing these issues by compensating for differences in
medical experience among doctors among physicians, standardizing the level of care, reducing the occurrence of overlooked early-stage
melanomas, and avoiding excessive pathological tests.