The invention relates to the technical field of
image processing, in particular to an ultrasonic image
ovarian tumor intelligent classification method, device and equipment and a medium. According to the method, the
multiple tumor features are merged according to the similarity among the tumor features of the multiple ultrasonic images to obtain the redundancy-removed features, then the O-RADS
score of the tumor is obtained based on the redundancy-removed features, and the O-RADS
score comprises five scores, so that the
lesion degree of the tumor can be more refined, and the
evaluation result of the
lesion degree of the tumor is improved. In addition, according to the method, the tumor features of each ultrasonic image are extracted separately, and then the final O-RADS
score is obtained according to the tumor features of the multiple ultrasonic images. No matter how many ultrasonic images of a tumor are collected, the feature of each ultrasonic image can be independently extracted, and no matter how many tumor features are, an O-RADS score can be obtained, so that the collected ultrasonic images are not limited by evaluating the
lesion degree of the tumor based on the O-RADS score, and the accuracy of evaluating the lesion degree of the tumor based on the O-RADS score is improved. Therefore, more ultrasonic images are allowed to be used for evaluating the tumor lesion degree, and the tumor lesion degree evaluation precision is further improved.