Abstract
Introduction: In a previous study, we developed a chromatin homology profile (CHP) method using the mathematical concept of homology and an analysis application, “Cell Checker,” to differentiate histological types of lung cancer in respiratory cytology. However, issues such as the analyzed field of view and data volume arose because the CHP method is based on images taken with a 100× objective lens. In this study, to overcome these challenges, we investigated the accuracy of the CHP method for differentiating histologic types of lung cancer using images obtained with a 40× objective. Methods: In total, 35 cases of lung cancer and benign tissue were selected, and 6 cancer cells and ciliated columnar epithelial cells per case were imaged with a 40× objective using a KEYENCE microscopy system. We analyzed chromatin contact (b1MAX value) and chromatin density (b1MAX/nuclear area value) with Cell Checker. Results: The b1MAX value was lower in small cell carcinoma, and there was a significant difference between small cell and non-small cell carcinoma. Significant differences in b1MAX/nuclear area values were found between adenocarcinoma and squamous cell carcinoma. These results were similar to those obtained with images from the 100× objective. Moreover, significant differences in b1MAX and b1MAX/nuclear area values were observed between benign and malignant cases. Conclusion: We have overcome the problem associated with using images obtained with the 100× objective and shown that the CHP method can be used to differentiate between benign and malignant cases. The CHP method could be used as a diagnostic support system for respiratory cytology.
