Combined Digital Colposcopy Analysis to Improve Cervical Precancer and Cancer Detection (MDC_Algo)

Precancerous lesions of the cervix occur frequently and are treatable. This justifies a population-based screening program. Following an abnormal Pap smear, patients are referred for a colposcopic exam to confirm the presence/stage of disease and select appropriate treatments. Unfortunately, these approaches do not detect all lesions or can sometimes give 'false positive' results (resulting in overtreatment). We are testing a device called a 'multispectral digital colposcope' to determine whether it is more effective at detecting precancerous cervical lesions than existing tools. Success in our study will make diagnoses more accurate and reduce the costs associated with unnecessary treatments.

The multispectral digital colposcope (MDC) is a specialized lightsource and imaging device mounted on a custom-made colposcope that acquires a sequence of fluorescence images. The MDC also consists of a point probe that is used to take detailed light measurements (call spectroscopy) from smaller regions of the cervix. The goal of the study is to see whether the MDC can perform better than standard methods in detecting cancerous or pre-cancerous abnormalities of the cervix.

Study subjects will be HPV swabbed. After vinegar is applied on the cervix as per usual colposcopy practice, the MDC will then be used to take a sequence of images of the cervix. Next 2-4 readings will be taken from the cervix using the point probe.

Study subjects may also be randomly asked to have additional pictures taken with another imaging device. a) The Fast Fluorescence Camera (FFC) is a modified digital camera that will capture images similar to that of the MDC. It will take a pair of images under two different lighting conditions - white light and violet light, both of which are used by the MDC. b) The Confocal Probe (CP) is a hand-held microscope used to collect images from the same location as the point probe. The cervix is topically stained with a fluorescent dye (Acriflavine) which is taken up by cell nuclei. The CP is then able to capture, in real time, images of the cells and their fluorescent nuclei.

This data will be used to train, test and valid an algorithm specific to the MDC. The ability of the algorithm to detect cervical intraepithelial neoplasia will be compared to standard clinical care (colposcopy).

The MDC images and pathology results will be compared to the spectroscopy data from the MDC probe. We will determine if spectroscopy has equal sensitivity to colposcopy for detecting high grade lesions.