Researchers develop a model of artificial intelligence that can diagnose eczema

A Japanese multidisciplinary research team has developed an innovative model of artificial intelligence that can objectively evaluate the intensity of “eczema” by using photos that patients take through their smartphones. Atopine dermatitis (eczema) suffers from regular chronic symptoms, which need continuous monitoring, and to adjust the treatment. Despite the spread of health applications and social media platforms that enable patients to detect their condition, the self -assessment of symptoms such as itching or sleep disorders does not necessarily reflect the actual intensity of the disease. Artificial intelligence and diagnosis of disease and artificial intelligence provides an accurate and critical assessment that depends on the visual analysis of skin pests. The study was published in allergy. The new model of artificial intelligence works by analyzing eczema images that increase patients through their smartphones, using three advanced algorithms that work in integration. Initially, the first algorithm determines the part of the affected body accurately, whether it is the face, arms, legs or others, followed by the second algorithm that detects the distinctive skin lesions of eczema such as red spots and scales. After that, the third role of the algorithm, which is the most advanced, comes the severity of the condition, using a medical tis scale, as it measures three main indicators exactly, which is “the degree of redness, the size of swelling, the amount of scratches or ulcers.” A large database. The system relies on a large database of over 57 thousand real photos, collected by the ‘atopio platform’ of 28,000 patients since 2018, giving it a high ability to analyze and diagnose carefully. After training the model on 880 photos with evaluation of self -hinging, it showed a high diagnostic accuracy. In verification tests with 220 photos, researchers found that the evaluation of artificial intelligence is strongly compatible with the evaluation of dermatologists. “Many patients struggle to judge the severity of eczema themselves. This model allows objective monitoring in real time, with only a smartphone, which better improve the disease control.” The study also showed the poor connection between the evaluation of artificial intelligence and the self -evaluation of itching, which confirms the need for “digital signs), which increases the accuracy in diagnosis and treatment. The team is currently expanding the extent of the model to combine different types of skin, different ages, in addition to combining additional features of clinical evaluation systems.