Diagnostic and Prognostic DNA-Karyometry for Cancer Diagnostics

Journal of Cancer Research Updates(2020)

引用 3|浏览0
暂无评分
摘要
Diagnostic and prognostic DNA-karyometry represents an automated computerized microscopical procedure, designed to improve cancer diagnostics at three different aspects: Screening for cancer cells, e.g. in body cavity effusions, urines or mucosal smears Specifying the risk of dysplasias or borderline lesions to progress to manifest cancer, e.g. of oral, bronchial or cervical mucosa, or the ovary Grading the malignancy of certain tumors, like prostate cancer. It combines an automated diagnostic classification of Feulgen-stained nuclei with precise nuclear DNA-measurements. DNA-aneuploidy is used as a specific marker of malignancy and its degree for grading. All types of cytological specimens can be used after (re-)staining specific for DNA according to Feulgen. Histological specimens are subjected to enzymatic cell separation before Feulgen-staining. A video-slide scanner is used for automated scanning of microscopical slides. Diagnostic nuclear classifiers have tissue-specifically been trained by an expert-cytopathologist (A. B.), based on Random Forest Classifiers, applying 18 different morphometric features. They achieve an overall accuracy of 91.1% to differentiate 8 differents types of objects/nuclei. Nuclear DNA-measurements of diploid nuclei achieve a CV of 1% morphometrically abnormal nuclei. Time needed for loading, scanning and validation of results per slide is about 10 minutes. Results of digital diagnostic nuclear classification can be verified by a cytopathologist, using image galleries. Likewise automated diagnostic interpretation of nuclear DNA-distributions can be checked on the monitor, before a pathologists validated diagnoses are issued. Screening-results are presented for body cavity effusions and urines. Evaluations of dysplasias are reported for oral, bronchial and cervical smears. Results of grading malignancy are shown for prostate cancers.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要