Sensitive area detection method and device

The invention provides a sensitive area detection method and device. The sensitive area detection method comprises the steps that a medical image is acquired; the medical image is segmented so as to obtain an under-detection area image; the under-detection area image is preprocessed; the preprocessed under-detection area image is inputted to a trained residual network, and sensitive area candidate areas are extracted from the under-detection area image; and the under-detection area images of the extracted sensitive area candidate areas are inputted to a 3D convolutional neural network through deep learning, and the sensitive area is extracted from the sensitive area candidate areas. With application of the technical scheme, the sensitive area can be automatically detected from the medical image, and the technical scheme can be applied to digital medical clinical diagnosis and treatment so that the workload of the medical staff can be reduced, the identification process is more objective and the leak detection