Mammography is a method for the diagnosis and screening of the human breast using low-energy X-rays. Mammograms tend to detect breast cancer early, usually by detecting standard masses, or by detecting microcalcifications.Adescriptive analysis of mammogram diagnosis using Convolutional Neural Network (CNN)for spectral detection is presented in this study.Initially, the color components are separated as red, green and blue. Only green channel is used for analysis because green is sensitive for humans. Finally,CNN is used for mammogram color spectraldetection. The performance of proposed system is analyzed by CNN in terms of accuracy.