Mueller matrix imaging for skin cancer detection

Over one million Americans are afflicted with skin cancer each year. Even though skin cancer has a 95% cure rate, approximately 10,000 people die in the United States each year of this disease. The current ABCDE(F) detection method is not sensitive enough to detect skin cancer in its early stages and requires a biopsy for any suspicious lesions. A lot of unnecessary biopsies, which are painful and costly to the patient, are taken. Therefore, a noninvasive technique is needed that can accurately detect the presence of skin cancer. In this thesis, an optical approach will be presented that has potential to be a noninvasive skin cancer detection technique. Several morphological and biochemical changes occur as tissue becomes cancerous, and therefore the optical properties of the tissue can be used to detect skin cancer. A Mueller matrix imaging system has been developed by our group that measures the 16 or 36-element Mueller matrix, which completely describes the optical properties of the tissue sample. The system is automated and can collect the Mueller matrix in less than one minute. This system will be used to
image Sinclair swine, and data analysis techniques will be employed to determine if the system can distinguish between cancerous and noncancerous tissue. System software improvements will also be made, and a new calibration technique will be presented.