Diffuse Optical Tomography (DOT) is a recent novel and non-invasive technology for medical imagery using near infrared light. The principle is based on the assessments of spatial variation in terms of light absorption and scattering coefficients that can be used to detect foreign bodies and tumors inside the human tissues. However, the current DOT technology is limited to a two-dimensional image reconstruction. In this paper, we propose an innovative three-dimensional nonlinear reconstruction algorithm to improve spatial and temporal resolutions. This algorithm has been implemented on an integrated mobile DOT system which comprises a light sensors array front-end circuit, an analog to digital converter, a Bluetooth wireless communication module to send the results to the GUI interface on a remote displaying device. The functionalities of the proposed system were verified using an experimental static human breast phantom. The experimental results show that the proposed system can detect a tumor of the size of 0.445cm2 at a depth of 2cm from the surface of the skin.