In scalp skin examinations, it is difficult to find a previously treated region on a patient's scalp through images captured by a camera attached to a diagnostic device because the zoom lens on camera has a small field of view. Thus, doctors manually record the region on a chart or manually mark the region. However, this process is slow and inconveniences the patient. Thus, we propose a new system for tracking the diagnostic device for the scalp skin of patients. Our research is novel in four ways. First, our proposed system consists of two cameras to capture the face and the diagnostic device. Second, the user can easily set the position of camera to capture the diagnostic device by manually moving a frame to which the camera is attached. Third, the position of patient's nostrils and corners of the eyes are detected to align the position of his/her head more accurately with the recorded position from previous sessions. Fourth, the position of the diagnostic device is continuously tracked during the examination through images that help detect the position of the color marker attached to the device. Experimental results show that our system has a higher performance than conventional method.