FIELD: medicine.
SUBSTANCE: invention refers to medicine, namely to a method for remote diagnosis of skin new growths by photography. Disclosed is a method automated remote diagnostics of skin new growths, in which an image of a skin area with an investigated neoplasm is obtained, the image is processed with quality control, at which the compliance of the image with the requirements of illumination and sharpness is determined, a neoplasm is detected on the image corresponding to the requirements, in which a segment including a suspected neoplasm is extracted from the image, the compliance of the selected segment with the segment size requirements is determined and the presence of all boundaries of the suspected neoplasm in the segment, an image with the selected segment is sent to the service unit to confirm the presence of the neoplasm in the selected segment, after receiving confirmation, the suspected neoplasm is diagnosed by means of a pre-trained classifying convolutional neural network with subsequent formation of a list of possible diagnoses with determination of their probability indicators or one diagnosis, presenting the diagnostic result to the user. Preferably, when determining compliance of an image with illumination requirements, uniform shadowless illumination is controlled, wherein control of uniformity of illumination of the image is carried out by determining the level of local contrast. Preferably, to determine compliance of the image with sharpness requirements, the maximum value from the Laplace filter applied to the image matrix is measured.
EFFECT: invention provides higher accuracy of determining probability of diagnosis or diagnoses corresponding to the investigated new growth of skin, in simplification of classification of new growth.
4 cl, 3 dwg
Title | Year | Author | Number |
---|---|---|---|
SYSTEM FOR DIAGNOSING MALIGNANT AND BENIGN SKIN NEW GROWTHS | 2022 |
|
RU2814539C1 |
METHOD FOR DETECTING ONCOLOGICAL DISEASES IN PELVIC ORGANS AND SYSTEM FOR IMPLEMENTING METHOD | 2023 |
|
RU2814790C1 |
METHOD OF PIGMENT SKIN GROWTHS RECOGNITION | 2018 |
|
RU2712919C1 |
METHOD AND SYSTEM FOR AUTOMATED DIAGNOSIS OF VASCULAR PATHOLOGIES BASED ON AN IMAGE | 2020 |
|
RU2741260C1 |
METHOD FOR AUDIOVISUAL RECOGNITION OF PERSONAL PROTECTION EQUIPMENT ON HUMAN FACE | 2022 |
|
RU2791415C1 |
METHOD FOR AUTOMATED MICROSCOPIC EXAMINATION OF SPECIMENS | 2007 |
|
RU2330265C1 |
METHOD FOR MICROSCOPIC EXAMINATION OF A SPECIMEN | 2006 |
|
RU2318201C1 |
METHOD OF PRELIMINARY DIAGNOSTICS OF LOWER LIMB VEIN DISEASES BASED ON ARTIFICIAL INTELLIGENCE ALGORITHMS | 2022 |
|
RU2804279C1 |
METHOD OF TEACHING AND MASTERING INSTRUCTION MATERIAL | 2008 |
|
RU2413298C2 |
METHOD FOR DIAGNOSIS OF LUNG CANCER BASED ON INTELLECTUAL ANALYSIS OF THE SHAPE, INTERNAL AND EXTERNAL STRUCTURES OF NEW GROWTHS | 2018 |
|
RU2694476C1 |
Authors
Dates
2024-04-17—Published
2022-12-15—Filed