Browsing by Subject "Histopathology"

Now showing 1 - 5 of 5

Open Access
Automated cancer stem cell recognition in H and E stained tissue using convolutional neural networks and color deconvolution
(SPIE, 2017) Aichinger, W.; Krappe, S.; Çetin, A. Enis; Çetin-Atalay, R.; Üner, A.; Benz, M.; Wittenberg, T.; Stamminger, M.; Münzenmayer, C.
The analysis and interpretation of histopathological samples and images is an important discipline in the diagnosis of various diseases, especially cancer. An important factor in prognosis and treatment with the aim of a precision medicine is the determination of so-called cancer stem cells (CSC) which are known for their resistance to chemotherapeutic treatment and involvement in tumor recurrence. Using immunohistochemistry with CSC markers like CD13, CD133 and others is one way to identify CSC. In our work we aim at identifying CSC presence on ubiquitous Hematoxilyn and Eosin (HE) staining as an inexpensive tool for routine histopathology based on their distinct morphological features. We present initial results of a new method based on color deconvolution (CD) and convolutional neural networks (CNN). This method performs favorably (accuracy 0.936) in comparison with a state-of-the-art method based on 1DSIFT and eigen-analysis feature sets evaluated on the same image database. We also show that accuracy of the CNN is improved by the CD pre-processing.
Open Access
Implantable microelectromechanical sensors for diagnostic monitoring and post-surgical prediction of bone fracture healing
(John Wiley and Sons Inc., 2015) McGilvray, K. C.; Ünal, E.; Troyer, K. L.; Santoni, B. G.; Palmer, R. H.; Easley, J. T.; Demir, Hilmi Volkan; Puttlitz, C. M.
The relationship between modern clinical diagnostic data, such as from radiographs or computed tomography, and the temporal biomechanical integrity of bone fracture healing has not been well-established. A diagnostic tool that could quantitatively describe the biomechanical stability of the fracture site in order to predict the course of healing would represent a paradigm shift in the way fracture healing is evaluated. This paper describes the development and evaluation of a wireless, biocompatible, implantable, microelectromechanical system (bioMEMS) sensor, and its implementation in a large animal (ovine) model, that utilized both normal and delayed healing variants. The in vivo data indicated that the bioMEMS sensor was capable of detecting statistically significant differences (p-value <0.04) between the two fracture healing groups as early as 21 days post-fracture. In addition, post-sacrifice micro-computed tomography, and histology data demonstrated that the two model variants represented significantly different fracture healing outcomes, providing explicit supporting evidence that the sensor has the ability to predict differential healing cascades. These data verify that the bioMEMS sensor can be used as a diagnostic tool for detecting the in vivo course of fracture healing in the acute post-treatment period. © 2015 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
Open Access
p53 codon 72 polymorphism in bladder cancer-No evidence of association with increased risk or invasiveness
(Springer, 2001) Törüner, G. A.; Uçar, A.; Tez, M.; Çetinkaya, M.; Özen, H.; Özçelik, T.
We studied the effect of the p53 gene Arg72Pro polymorphism on bladder cancer susceptibility in a case control study of 121 bladder cancer patients and 114 age-sex matched controls to determine whether this polymorphism is a biomarker for the risk and how aggressive the disease is. Genomic DNA was obtained from venous blood samples for genotype determination by PCR and restriction digestion. The genotype frequencies in the patient group were Arg/Arg: 0.3553, Arg/Pro: 0.4711, Pro/Pro: 0.1736, and in the control group Arg/Arg: 0.3684, Arg/Pro: 0.4825, Pro/Pro: 0.1491. The distribution of genotypes between the two groups was not statistically different (χ2 = 0.260, df: 2, P = 0.878). The patient group was subdivided into two groups as superficial bladder cancer (n = 88) and invasive bladder cancer (n = 33), according to the presence of muscle invasion. The distribution of genotypes in the superficial group was Arg/Arg: 0.3409, Arg/Pro: 0.5114, Pro/Pro: 0.1477 and in the invasive group Arg/Arg: 0.3940, Arg/Pro: 0.3636, Pro/Pro: 0.2424. No association was observed with the invasiveness of the tumor (χ2 = 2.542, df: 2, P = 0.281). Stratification of the data by tobacco exposure did not result in a significant difference in genotype frequencies. These data do not support an association between the p53 Arg72Pro polymorphism and bladder cancer.
Open Access
Quantification of SLIT-ROBO transcripts in hepatocellular carcinoma reveals two groups of genes with coordinate expression
(BioMed Central, 2008) Avci, M. E.; Konu, O.; Yagci, T.
Background: SLIT-ROBO families of proteins mediate axon pathfinding and their expression is not solely confined to nervous system. Aberrant expression of SLIT-ROBO genes was repeatedly shown in a wide variety of cancers, yet data about their collective behavior in hepatocellular carcinoma (HCC) is missing. Hence, we quantified SLIT-ROBO transcripts in HCC cell lines, and in normal and tumor tissues from liver. Methods: Expression of SLIT-ROBO family members was quantified by real-time qRT-PCR in 14 HCC cell lines, 8 normal and 35 tumor tissues from the liver. ANOVA and Pearson's correlation analyses were performed in R environment, and different clinicopathological subgroups were pairwise compared in Minitab. Gene expression matrices of cell lines and tissues were analyzed by Mantel's association test. Results: Genewise hierarchical clustering revealed two subgroups with coordinate expression pattern in both the HCC cell lines and tissues: ROBO1, ROBO2, SLIT1 in one cluster, and ROBO4, SLIT2, SLIT3 in the other, respectively. Moreover, SLIT-ROBO expression predicted AFP-dependent subgrouping of HCC cell lines, but not that of liver tissues. ROBO1 and ROBO2 were significantly up-regulated, whereas SLIT3 was significantly down-regulated in cell lines with high-AFP background. When compared to normal liver tissue, ROBO1 was found to be significantly overexpressed, while ROBO4 was down-regulated in HCC. We also observed that ROBO1 and SLIT2 differentiated histopathological subgroups of liver tissues depending on both tumor staging and differentiation status. However, ROBO4 could discriminate poorly differentiated HCC from other subgroups. Conclusion: The present study is the first in comprehensive and quantitative evaluation of SLIT-ROBO family gene expression in HCC, and suggests that the expression of SLIT-ROBO genes is regulated in hepatocarcinogenesis. Our results implicate that SLIT-ROBO transcription profile is bi-modular in nature, and that each module shows intrinsic variability. We also provide quantitative evidence for potential use of ROBO1, ROBO4 and SLIT2 for prediction of tumor stage and differentiation status.
Open Access
Two-tier tissue decomposition for histopathological image representation and classification
(Institute of Electrical and Electronics Engineers, 2015) Gultekin, T.; Koyuncu, C. F.; Sokmensuer, C.; Gunduz Demir, C.
In digital pathology, devising effective image representations is crucial to design robust automated diagnosis systems. To this end, many studies have proposed to develop object-based representations, instead of directly using image pixels, since a histopathological image may contain a considerable amount of noise typically at the pixel-level. These previous studies mostly employ color information to define their objects, which approximately represent histological tissue components in an image, and then use the spatial distribution of these objects for image representation and classification. Thus, object definition has a direct effect on the way of representing the image, which in turn affects classification accuracies. In this paper, our aim is to design a classification system for histopathological images. Towards this end, we present a new model for effective representation of these images that will be used by the classification system. The contributions of this model are twofold. First, it introduces a new two-tier tissue decomposition method for defining a set of multityped objects in an image. Different than the previous studies, these objects are defined combining texture, shape, and size information and they may correspond to individual histological tissue components as well as local tissue subregions of different characteristics. As its second contribution, it defines a new metric, which we call dominant blob scale, to characterize the shape and size of an object with a single scalar value. Our experiments on colon tissue images reveal that this new object definition and characterization provides distinguishing representation of normal and cancerous histopathological images, which is effective to obtain more accurate classification results compared to its counterparts.