Automated recognition of atypical nuclei in breast cancer cytology specimens by iterative image transformations

Automated detection of cancer cells in effusion specimens by DNA karyometry

Background

The average sensitivity of conventional cytology for the identification of cancer cells in effusion specimens is only approximately 58%. DNA image cytometry (DNA‐ICM), which exploits the DNA content of morphologically suspicious nuclei measured on digital images, has a sensitivity of up to 91% for the detection of cancer cells. However, when performed manually, to our knowledge to date, an expert needs approximately 60 minutes for the analysis of a single slide.

Methods

In the current study, the authors present a novel method of supervised machine learning for the automated identification of morphologically suspicious mesothelial and epithelial nuclei in Feulgen‐stained effusion specimens. The authors compared this with manual DNA‐ICM and a gold standard cytological diagnosis for 121 cases. Furthermore, the authors retrospectively analyzed whether the amount of morphometrically abnormal mesothelial or epithelial nuclei detected by the digital classifier could be used as an additional diagnostic marker.

Results

The presented semiautomated DNA karyometric solution identified more diagnostically relevant abnormal nuclei compared with manual DNA‐ICM, which led to a higher sensitivity (76.4% vs 68.5%) at a specificity of 100%. The ratio between digitally abnormal and all mesothelial nuclei was found to identify cancer cell–positive slides at 100% sensitivity and 70% specificity. The time effort for an expert therefore is reduced to the verification of a few nuclei with exceeding DNA content, which to our knowledge can be accomplished within 5 minutes.

Conclusions

The authors have created and validated a computer‐assisted bimodal karyometric approach for which both nuclear morphology and DNA are quantified from a Feulgen‐stained slide. DNA karyometry thus increases the diagnostic accuracy and reduces the workload of an expert when compared with manual DNA‐ICM.

An automated procedure for the detection of cancer cells in effusion specimens is presented. It provides a sensitivity of 76.4% and a specificity of 100% at a time effort for clinicians of approximately 5 minutes per specimen.

Segment and fit thresholding: a new method for image analysis applied to microarray and immunofluorescence data

Experiments involving the high-throughput quantification of image data require algorithms for automation. A challenge in the development of such algorithms is to properly interpret signals over a broad range of image characteristics, without the need for manual adjustment of parameters. Here we present a new approach for locating signals in image data, called Segment and Fit Thresholding (SFT). The method assesses statistical characteristics of small segments of the image and determines the best-fit trends between the statistics. Based on the relationships, SFT identifies segments belonging to background regions; analyzes the background to determine optimal thresholds; and analyzes all segments to identify signal pixels. We optimized the initial settings for locating background and signal in antibody microarray and immunofluorescence data and found that SFT performed well over multiple, diverse image characteristics without readjustment of settings. When used for the automated analysis of multicolor, tissue-microarray images, SFT correctly found the overlap of markers with known subcellular localization, and it performed better than a fixed threshold and Otsu's method for selected images. SFT promises to advance the goal of full automation in image analysis.

DNA image cytometry on machine-selected breast cancer cells and a comparison between flow cytometry and scanning cytophotometry

A DNA image cytometry method, implemented on the LEYTAS image processing system, has been applied to acriflavine-Feulgen-stained breast cancer cytology specimens. An essential feature of the LEYTAS image cytometry method (LCM) is the automated selection of single nuclei according to predetermined specifications. Visual interaction has been used to reject remaining artefacts like overlapping nuclei. DNA profiles obtained with LCM have been compared with DNA profiles obtained by scanning cytophotometry (SCM) or flow cytometry (FCM). The resolution of DNA profiles obtained with LCM is similar to that from SCM but lower than that from FCM. However, a high correlation is found for the DNA indices measured with LCM and FCM (r = 0.97). The LCM profiles of aneuploid tumours generally showed lower accessory diploid fractions than FCM profiles due to the automated rejection of leukocyte nuclei. Also, LCM profiles frequently showed the presence of minor subpopulations of highly aneuploid/polyploid tumour cells that could not be identified by FCM. Therefore, LCM appears to be supplementary to FCM for studying tumour cell stemline heterogeneity.