This study will examine the use of timed image acquisition, wavelength filtered light and ace to whitening kinetics to improve the accuracy of colposcopy for the routine clinical diagnosis of cervical cancer. A colposcopy workstation will be built with enhanced software and wavelength filtered light capability to collect necessary patient data. A cohort of 25 patients with high grade Cervical Intraepitheliel Neoplasia (C/N II/III will be imaged with the modified I workstation to quantify how these image enhancements improve the predictive value of colposcopy. The time-dependent and spectral data will be used to improve the accuracy of colposcopy in the discrimination of high grade lesions from other conditions including low grade lesions (C/NI) and normal tissue. A time-sequenced series of images will be collected after the uniform and timed application of 5 percent acetic acid onto the cervix. Images will be processed to find predictive markers such as (1) lifetime and intensity of acetowhitening and resulting temporal decay curves, and (2) spectral changes in reflectance due to acetowhitening. Techniques will be improved for the processing and analysis of temporal curves to enhance their predictive value.
Completion of this Phase I program will lay the groundwork for significantly improved digital colposcopic instrumentation. The timed spectral images of cervical acetowhitening will improve the quality of clinical treatment while reducing costs. As with any new technology, market commercialization will depend on the willingness of physicians to accept the new technology. The device developed in this application, however, is similar to existing colposcopes, and the barrier to acceptance should be less as users in the OB/Gyn community are familiar with this equipment. The device will allow for improved accuracy in the mapping and grading of cervical cancer.
