Three-dimensional visualization and quantitative analysis of cervical cell nuclei with confocal laser scanning microscopy

Abstract

Objective: To develop a method for the acquisition and processing of 3-dimensional images based on confocal laser scanning microscopy for the purpose of 3-dimensional visualization and quantitative analysis of cell nuclei. Study Design: A contour-based surface rendering method was used, and volume rendering was implemented according to the basic volume rendering pipeline. To extract quantitative features, a 3-dimensional labeling method based on slice information was used. After applying the labeling algorithm, the measurements for 3-dimensional quantitative analysis of nuclei were extracted: nuclear volume, surface area and spherical shape factor. We compared the 3-dimensional features of normal and abnormal cervical cell nuclei. Results: Comparison of the size of 3-dimensional cervical cell nuclei between normal and abnormal revealed a statistically significant difference. The proposed method could overcome the limitation inherent in 2-dimensional analysis and could become a way of improving the accuracy and reproducibility of quantification of cell nuclei. Conclusion: Three-dimensional visualization and quantification of cell nuclei provide valuable medical information that can lead to a more objective diagnosis.