The invention discloses an
automatic segmentation typing and thickness measuring method based on an endometrial
ultrasound image, which comprises the following steps: 1) collecting an endometrial
ultrasound scanning image set, manually selecting a standard
uterus longitudinal
cutting image, sketching an endometrial boundary, making a
typing label, and establishing a training and testing sample set; 2) carrying out image preprocessing such as filtering on each image; removing
noise influence and achieving an
image enhancement effect; 3) respectively constructing a
deep learning model for segmenting an endometrial region and
typing, and respectively inputting a
training set for training; 5) performing the same image preprocessing on the
test set image, inputting the
test set image into the trained
deep learning model, and performing segmentation to obtain a segmented endometrial region and a typing result; 6) calculating the thickness of the
endometrium by using a
digital image morphological
algorithm and a segmentation result of the
endometrium ultrasonic image; by means of the method, the endometrial area can be accurately and effectively extracted, and more accurate thickness measurement and endometrial typing can be achieved.