Fig. 3From: Effective automatic detection of anterior cruciate ligament injury using convolutional neural network with two attention mechanism modulesStructure of our convolutional neural network model. The CNN model proposed in this paper is roughly divided into two parts. The first part is mainly supported by ATM1, which enhances feature extraction. The latter part is mainly supported by ATM2 to enhance the classification. The lower left is the structure of the ResNet, and the “ReLU” is the activation function of the ResNet. The method used for down-sampling was max pooling with a step size of 2Back to article page