Smartphone-based compression-induced sensing system uses the light diffusion pattern to characterize the early-stage breast tumor noninvasively. The system is built on a smartphone and cloud platform to capture, transfer, and interface with the user. The compressed tissue's deforming pattern creates distinctive tactile images due to the size and hardness of the tumor. From the compression-induced images, we estimate the size of the tumor using projection analysis and the tumor's malignancy using the tissue deformation index ratio. Deformation index ratio is based on the changes of a healthy region over the tumorous region. By using the projection analysis, the human patient tumor size estimation resulted in 52.3% of the average error. For a small number (seven) of the feasibility test, the tumor's malignancy was classified based on the deformation index ratio with 67.0% of sensitivity and 100% specificity.