Mathematical morphologic segmentation dedicated to quantitative immunohistochemistry.

OBJECTIVE: To develop automatic segmentation sequences for fully automated quantitative immunohistochemistry of cancer cell nuclei by image analysis. STUDY DESIGN: The study focused on the automated delineation of cancer cell lobules and nuclei, taking breast carcinoma as an example. A hierarchic segmentation was developed, employing mainly the chaining of mathematical morphology operators. The proposed sequence was tested on 22 images of various situations, collected from 18 different cases of breast carcinoma. A quality control procedure was applied, comparing the automated method with manual outlining of cancer cell foci and with manual pricking of cancer cell nuclei. RESULTS: Good concordance was found between automated and manual segmentation procedures (90% for cancer cell clumps, 97% for cancer cell nuclei on average), but the rate of false positive nuclei (small regions labeled as nuclei by the segmentation procedure) could be relatively high (11% on average, with a maximum of 35%) and can result in underestimation of the immunostaining ratio. CONCLUSION: This study examined a preliminary approach to automated immunoquantification, limited to automated segmentation without any color characterization. The automated hierarchic segmentation presented here leads to good discrimination of cancer cell nuclei at the chosen magnification.

On the number of clusters and the fuzziness index for unsupervised FCA application to BOLD fMRI time series.

The aim of this paper is to present an exploratory data-driven strategy based on Unsupervised Fuzzy Clustering Analysis (UFCA) and its potential for fMRI data analysis in the temporal domain. The a priori definition of the number of clusters is addressed and solved using heuristics. An original validity criterion is proposed taking into account data geometry and the partition Membership Functions (MFs). From our simulations, this criterion is shown to outperform other indices used in the literature. The influence of the fuzziness index was studied using simulated activation combined with real life noise data acquired from subjects under a resting state. Receiver Operating Characteristics (ROC) methodology is implemented to assess the performance of the proposed UFCA with respect to the fuzziness index. An interval of choice around 2, a value widely used in FCA, is shown to yield the best performance.

A multistep unsupervised fuzzy clustering analysis of fMRI time series.

A paradigm independent multistage strategy based on the Unsupervised Fuzzy Clustering Analysis (UFCA) and its potential for fMRI data analysis are presented. The influence of the fuzziness index is studied using Receiver Operating Characteristics (ROC) methodology and an interval of choice, around the widely used value 2, is shown to yield the best performance. The ill-balanced data problem is also overcome using a pre-processing step to reduce the number of voxels presented to the method. Statistical and anatomical criteria are proposed to exclude some voxels and enhance the UFCA sensitivity. An original postprocessing step aiming at statistically characterizing the obtained clusters is also developed. Two similarity criteria are used: the correlation coefficient on temporal profiles and a novel fuzzy overlap coefficient on membership degree maps. This final step provides a useful analysis tool to study intra-individual reproducibility of the classes across series (stimulation vs. stimulation, noise vs. noise or stimulation vs. noise). Finally, a comparison between this technique and some existing or locally developed postprocessing algorithms is presented using ROC methods. Its sensitivity and robustness is compared to the classical FCA or other techniques as a function of several parameters such as Contrast-to-Noise Ratio (CNR) and noise amplitude. Even without knowledge about the paradigm, the hemodynamic response function and the number of clusters, the performances of the proposed strategy are comparable to those of the classical approaches where extensive prior knowledge has to be added.

Brain tissue classification of magnetic resonance images using partial volume modeling.

This paper presents a fully automatic three-dimensional classification of brain tissues for Magnetic Resonance (MR) images. An MR image volume may be composed of a mixture of several tissue types due to partial volume effects. Therefore, we consider that in a brain dataset there are not only the three main types of brain tissue: gray matter, white matter, and cerebro spinal fluid, called pure classes, but also mixtures, called mixclasses. A statistical model of the mixtures is proposed and studied by means of simulations. It is shown that it can be approximated by a Gaussian function under some conditions. The D'Agostino-Pearson normality test is used to assess the risk alpha of the approximation. In order to classify a brain into three types of brain tissue and deal with the problem of partial volume effects, the proposed algorithm uses two steps: 1) segmentation of the brain into pure and mixclasses using the mixture model; 2) reclassification of the mixclasses into the pure classes using knowledge about the obtained pure classes. Both steps use Markov random field (MRF) models. The multifractal dimension, describing the topology of the brain, is added to the MRFs to improve discrimination of the mixclasses. The algorithm is evaluated using both simulated images and real MR images with different T1-weighted acquisition sequences.

Automatic morphological sieving: comparison between different methods, application to DNA ploidy measurements.

The aim of the present study is to propose alternative automatic methods to time consuming interactive sorting of elements for DNA ploidy measurements. One archival brain tumour and two archival breast carcinoma were studied, corresponding to 7120 elements (3764 nuclei, 3356 debris and aggregates). Three automatic classification methods were tested to eliminate debris and aggregates from DNA ploidy measurements (mathematical morphology (MM), multiparametric analysis (MA) and neural network (NN)). Performances were evaluated by reference to interactive sorting. The results obtained for the three methods concerning the percentage of debris and aggregates automatically removed reach 63, 75 and 85% for MM, MA and NN methods, respectively, with false positive rates of 6, 21 and 25%. Information about DNA ploidy abnormalities were globally preserved after automatic elimination of debris and aggregates by MM and MA methods as opposed to NN method, showing that automatic classification methods can offer alternatives to tedious interactive elimination of debris and aggregates, for DNA ploidy measurements of archival tumours.

Image analysis software for automatic DNA ploidy assessment of archival solid tumours.

BACKGROUND: Image cytometry has proved to provide a good alternative to flow cytometry for DNA ploidy measurement of archival tumors. However, when interactively done this technique is unable to give statistically valuable results within an acceptable time for clinical oncology. METHODS: An image cytometer was developed for fully automatic DNA ploidy quantitation, focusing efforts on speed and accuracy. Software functionalities include systematic acquisition of fields on a microscopic slide, detection, localization and sorting of nuclei, computation of the DNA content together with post-processing tools, for a deeper analysis of the DNA ploidy diagram. RESULTS: DNA ploidy analysis of archival breast carcinoma samples illustrates the accuracy of DNA ploidy measurements and the sensitivity in the detection of DNA ploidy abnormalities as a result of cell sorting. CONCLUSIONS: Fully automatic image cytometry is able to combine qualities of flow cytometry (automatic analysis of a statistically significant collection of cell nuclei) with additional advantages: sorting of unwanted events (debris, stromal and inflammatory cell nuclei) and facilities for an a posteriori control of the quality of cell selection. This method is well suited to DNA ploidy analysis of archival cancer samples.

Fully automatic identification of AC and PC landmarks on brain MRI using scene analysis.

We describe a method for identification of brain structures from MRI data sets. The bulk of the paper concerns an automatic system for finding the anterior and posterior commissures [(AC) and (PC)] in the midsagittal plane. These landmarks are key for the definition of the Talairach space, commonly used in stereotactic neurosurgery, in the definition of common coordinate systems for the pooling of functional positron emission tomography (PET) images and for neuroanatomy studies. The process works according to a step-by-step procedure: it first analyzes the skull limits. A grey-level histogram is then calculated and allows an automated selection of thresholds. Then, the interhemispheric plane is detected. Following an advanced scene analysis in the midsagittal plane for anatomical structures, the AC and the PC are identified. Experimentally, with a set of 200 patients, the process never failed. Its performances and limits are comparable to that of neuroanatomy experts. Those results are due to a high degree of robustness at each step of the program.

[Automation of the cytometric analysis of the DNA content of solid tumors]. / Automatisation de l'analyse cytodensitométrique du contenu en ADN des tumeurs solides.

An automatic machine, dedicated to solid tumor DNA ploidy quantitation has been built in order to provide pathologists with a tool usable in clinical practice. Main efforts were focused on an automation of each step of the analysis and on an elimination of any subjective choice, while preserving the quality of measurement. As the software is independent of the machine architecture, it offers performances which increase in parallel with the rapid evolution of the computers. An illustration of the various functionalities of the automaton is proposed through the study of deparaffined breast cancer samples.

Generalized region growing operator with optimal scanning: application to segmentation of breast cancer images.

Segmentation of medical images is a complex problem owing to the large variety of their characteristics. In the automated analysis of breast cancers, two image classes may be distinguished according to whether one considers the quantification of DNA (grey level images of isolated nuclei) or the detection of immunohistochemical staining (colour images of histological sections). The study of these image classes generally involves the use of largely different image processing techniques. We therefore propose a new algorithm derived from the watershed transformation enabling us to solve these two segmentation problems with the same general approach. We then present visual and quantitative results to validate our method.

[Neighborhood graphs and image processing. Contribution to images of immunohistochemical staining]. / Graphes de voisinage et traitement d'images. Application à des images de marquage immunohistochimique.

Devising an image analyzer dedicated to the automatic quantification of immunohistochemical staining for clinical oncology implies developing a method for the delimitation of tumoral cell nests, setting aside tumoral stroma, while accounting for the topology of the staining. The representation of images by neighborhood graphs can bring an answer to both requirements. In this paper, a methodological approach is presented. It consists in a preliminary study dealing with nuclear immunostaining images of breast cancer. Segmentation of the graph structure allows to separate clusters of cancer cells and the analysis of this structure can account for the focal or diffuse aspect of the staining within the tumor.