Assessment of DNA Karyometry for Automated Detection of Cancer Cells

Abstract

High throughput and sufficient diagnostic accuracy of microscopical screening of cytological samples for the presence of cancer cells necessitates the use of highly qualified professionals. Using supervised machine learning, a programme was created that can categorise Feulgen-stained nuclei into eight diagnostically different types using commercially available, automated microscope-based screeners (MotiCyte and EasyScan). The nuclear DNA content was internally calibrated, using normal cells. The nuclei of cells that seemed to be malignant were recognized morphometrically.  A blinded study was performed using oral smears from 92 patients with Fanconi anemia, revealing oral leukoplakias or erythroplakias. In a previous study, we evaluated the diagnostic accuracy of 121 samples of serous effusions. In addition, we sought to identify those whose tumours would not progress within 4 years using a blinded study with 80 prostate cancer patients who were receiving active surveillance. Applying a threshold of the presence of >4% of morphologically abnormal nuclei from oral squamous cells and DNA single-cell or stemline aneuploidy to identify samples suspected of malignancy, an overall diagnostic accuracy of 91.3% was found as compared with 75.0%, determined by conventional subjective cytological assessment using the same slides. Automated screening effusions, accuracy was 84.3%, while conventional cytology accuracy was 95.9%. Within 4.1 years, none of the prostate cancer patients under active monitoring with DNA grade 1 demonstrated disease progression. In order to identify malignant cells in various human specimen types with diagnostic accuracy on par with subjective cytological evaluation, an automated microscope-based screener was created. This automated method could detect early prostate tumours that do not spread while receiving no treatment.