To address the difficulty in accurately quantifying the number of broken wires in bridgecable magnetic flux leakage (MFL) testing, a quantitative identification approach for broken wires in bridge cables based on Cable-conrolutional neural network anisotropic dilated three dinensoral (Cable-AD3D-CNN) is proposed according to the distribution characteristics of MFL signals along the axial, circumferential, and radial directions of bridge cables. This method uses a bridge-cable MFL testing platform to fabricate specimens containing different numbers of broken wires and different broken-wire locations under the same broken-wire number. The axial component of the leakage magnetic field is acquired at different axial positions, circumferential positions, and radial lift-off distances to establish the dataset. Considering the directional distribution characteristics and data-scale differences of the acquired MFL signals, anisotropic convolution kernels are adopted to preserve directional resolution while enhancing the extraction of spatial features associated with broken wires. In addition, progressively increasing and pairwise coprime dilation rates are introduced along the axial direction to enlarge the receptive field for the MFL signals. Based on these designs, the Cable-AD3D-CNN model is established. The model is then trained and deployed to identify the MFL signals of bridge-cable specimens containing broken wires, thereby determining the number of broken wires. Experimental results demonstrate that the proposed method exhibits good convergence and achieves an accuracy of 96.94% for broken-wire-number identification. Compared with the method based on the baseline three-dimensional convolutional neural network (baseline 3D-CNN), the accuracy is increased by 13.61%. In addition, the proposed approach effectively reduces confusion among categories with similar broken-wire numbers and improves the discriminability of samples with different broken-wire numbers, providing a useful reference for accurate quantitative evaluation and maintenance decision-making for bridge-cable damage.