DIGITAL IMAGING COLPOSCOPY, IMAGE-ANALYSIS AND QUANTIFICATION OF THE COLPOSCOPIC IMAGE

Objective To investigate the use of a digital imaging system for colposcopy, its use for image analysis and quantification of the colposcopic features that may predict histological outcome as defined by large loop excision of transformation zone. Design Prospective programme study of all patients undergoing colposcopy for cytological abnormalities. Setting Dudley Road Hospital, Birmingham. Subjects Fifty consecutive patients having cytological and colposcopic abnormality treated with large loop excision of transformation zone formed the study group. Main outcome measures Pretreatment and colposcopic features correlated with histological diagnosis of excised transformation zone. Results Index cytology and current smoking status are the most important variables for prediction of histological diagnosis. Other important variables are focality of lesion, surface pattern, intercapillary distance and degree of acetowhiteness. Conclusion Digital imaging colposcopy allows image capture, processing and objective analysis. This methodology holds advantages for basic and clinical research, teaching, diagnostics and clinical audit. The system can act as a quality control tool for colposcopy units. This system should prove invaluable for further quantitative studies, for natural history studies and for those patients with deferred treatment of their cytological and colposcopic abnormalities. The statistical models described may be incorporated into the system and can aid the colposcopist in management of the woman with abnormal cervical cytology and colposcopic abnormality.