On your phone: a real intelligent application reveals skin diseases!

There are a few applications that claim to be able to monitor and identify skin cancer markers but may be the most recent is the most accurate scientific yet.

The skin vision or "vision of the skin" application uses an algorithm developed two years ago, using more than 4,000 images from more than 500 skin assessments - and the accuracy of this algorithm is 90 percent. The developers of this algorithm say they do not replace medical advice because they do not provide a complete diagnosis but can be used to identify skin lesions as high risk if the algorithm shows that the images correlate with a specific skin disease.

The "vision of the heart" was developed to help users follow the moles appearing on their skin and to know the size of the risks these moles carry with them. Once the images are lifted, they are analyzed by a mathematical algorithm and then advanced to the gravity of the matter. If this algorithm identifies properties that can detect dangerous growth or unwanted growth, the application displays a warning message and guides you to the nearest doctor for your place to see your condition.

Users can also track the development of moles over time. The algorithm was developed by a team of certified dermatologists including Costas, Contantinos, Christina and some mathematics and image analysis experts. In order for the application to recognize the pattern of skin, it is based on fractal analysis, which treats the data of both skin lesions and surrounding tissues. Detection of the analysis algorithm for any abnormal growth patterns by calculating the fractional dimension of both the lesions and surrounding tissue.

So far more than 4,000 images have been reviewed and reviewed by dermatologists and more than 500 patients have already been evaluated. The algorithm is also involved in clinical research on the patient for comparison with the assessment of a dermatologist. It was found that by comparing live assessments with the results of the algorithm, it was found to be accurate by 74% and this percentage increased to 90% by comparing the results. The application cost is about five dollars and has an annual subscription of up to $ 28 per year and a monthly subscription of $ 5 per month. The application also suggests that the patient should visit the nearest doctor if the lesion is at high risk. Many have stressed that the application does not replace the doctor's advice.