Medical images provide intuitive visualization methods and provide reliable basis for disease diagnosis and treatment. In the process of medical image acquisition, transmission and recording, noise information will inevitably be introduced due to the influence of transmission medium errors, measurement errors and the quantization of digital information. Noisy images not only reduce the image quality and affect the visual effect, but also blur or obscure the important information of the image. In order to obtain clear and noise-free medical images, this paper proposes an adaptive median filter image denoising method. Firstly, the noise points and signal points in the image are judged, and only the noise points are processed. Then, in the process of removing noise signal, we fully consider the pixel distribution feature information around the noise point and replace the pixel value of the noise point. This method has a good ability to remove noise and it can effectively deal with impulse noise while protecting the details of the image.