Multi-scale Hyperspectral Imaging Of Cervical Neoplasia

Cervical cancer is a prevalent disease in many developing countries and is the one of leading reason of female cancer death. In China, Colposcopy screening method is one of the most common approach for screening cervical intraepithelial neoplasia (CIN). However, its clinical efficacy heavily relies on the examiner’s experience. Spectroscopy is a potentially effective method for noninvasive diagnosis of cervical neoplasia which can improve the accuracy and availability of cervical cancer screening. In this paper, we introduce a diagnosis method involves hyperspectral imaging (HSI) technique to image changes in reflectance at many wavelengths across the entire cervical epithelium for noninvasive detection and quantitative analysis of cervical neoplasia. A hyperspectral camera provided by Brimrose was used to collect the reflectance images of the entire cervix under Xenon lamp illumination, followed by standard colposcopy examination and cervical tissue biopsy at both normal and abnormal sites, which were indicated on the colposcopy image by clinician. The collected reflectance data were calibrated. A wide-gap second derivative analysis was applied for the differentiation of cervical tissue. Further image segmentation based on the minimum distance classification method was carried out to classify tissue into different types according to the wide-gap second derivative characteristics at different stages of cervical intraepithelial neoplasia. The consistent between the histopathological results demonstrates that the hyperspectral imaging system can distinguish between the neoplastic and normal tissue. The hyperspectral camera was also coupled with a lab microscope to acquire the transmittance images of the pathological slides. Based on the spectral and spatial differences, the similar wide-gap second derivative analysis and image classification method were applied on the transmittance images, and the classification result was compared to that of clinical findings, which verifies the accuracy and efficacy of the method. The multi-scale hyperspectral imaging on cervical tissue provides a possibility for the application of a low-cost, noninvasive, portable, and quantitative imaging system to detect CIN.