Raissa Relator

Segmental HOG: new descriptor for glomerulus detection in kidney microscopy image

BackgroundThe detection of the glomeruli is a key step in the histopathological evaluation of microscopic images of the kidneys. However, the task of automatic detection of the glomeruli poses challenges owing to the differences in their sizes and shapes in renal sections as well as the extensive variations in their intensities due to heterogeneity in immunohistochemistry staining.Although the rectangular histogram of oriented gradients (Rectangular HOG) is a widely recognized powerful descriptor for general object detection, it shows many false positives owing to the aforementioned difficulties in the context of glomeruli detection.ResultsA new descriptor referred to as Segmental HOG was developed to perform a comprehensive detection of hundreds of glomeruli in images of whole kidney sections. The new descriptor possesses flexible blocks that can be adaptively fitted to input images in order to acquire robustness for the detection of the glomeruli. Moreover, the novel segmentation technique employed herewith generates high-quality segmentation outputs, and the algorithm is assured to converge to an optimal solution. Consequently, experiments using real-world image data revealed that Segmental HOG achieved significant improvements in detection performance compared to Rectangular HOG.ConclusionThe proposed descriptor for glomeruli detection presents promising results, and it is expected to be useful in pathological evaluation.

Support consistency of direct sparse-change learning in Markov networks

We study the problem of learning sparse structure changes between two Markov networks P and Q. Rather than fitting two Markov networks separately to two sets of data and figuring out their differences, a recent work proposed to learn changes directly via estimating the ratio between two Markov network models. Such a direct approach was demonstrated to perform excellently in experiments, although its theoretical properties remained unexplored. In this paper, we give sufficient conditions for successful change detection with respect to the sample size np,nq, the dimension of data m, and the number of changed edges d. More specifically, we prove that the true sparse changes can be consistently identified for np = Ω(d2 log m2+ m/2) and nq = Ω(n2p/d), with an exponentially decaying upper-bound on learning error. Our theoretical guarantee can be applied to a wide range of discrete/continuous Markov networks.

Stochastic Dykstra Algorithms for Metric Learning with Positive Definite Covariance Descriptors

Recently, covariance descriptors have received much attention as powerful representations of set of points. In this research, we present a new metric learning algorithm for covariance descriptors based on the Dykstra algorithm, in which the current solution is projected onto a half-space at each iteration, and runs at \(O(n^3)\) time. We empirically demonstrate that randomizing the order of half-spaces in our Dykstra-based algorithm significantly accelerates the convergence to the optimal solution. Furthermore, we show that our approach yields promising experimental results on pattern recognition tasks.

Mahalanobis Encodings for Visual Categorization

Nowadays, the design of the representation of images is one of the most crucial factors in the performance of visual categorization. A common pipeline employed in most of recent researches for obtaining an image representa- tion consists of two steps: the encoding step and the pooling step. In this paper, we introduce the Mahalanobis metric to the two popular image patch encoding modules, Histogram Encoding and Fisher Encoding, that are used for Bag- of-Visual-Word method and Fisher Vector method, respectively. Moreover, for the proposed Fisher Vector method, a close-form approximation of Fisher Vector can be derived with the same assumption used in the original Fisher Vector, and the codebook is built without resorting to time-consuming EM (Expectation-Maximization) steps. Experimental evaluation of multi-class classification demonstrates the effectiveness of the proposed encoding methods.

Identifying statistically significant combinatorial markers for survival analysis

BackgroundSurvival analysis methods have been widely applied in different areas of health and medicine, spanning over varying events of interest and target diseases. They can be utilized to provide relationships between the survival time of individuals and factors of interest, rendering them useful in searching for biomarkers in diseases such as cancer. However, some disease progression can be very unpredictable because the conventional approaches have failed to consider multiple-marker interactions. An exponential increase in the number of candidate markers requires large correction factor in the multiple-testing correction and hide the significance.MethodsWe address the issue of testing marker combinations that affect survival by adapting the recently developed Limitless Arity Multiple-testing Procedure (LAMP), a p-value correction technique for statistical tests for combination of markers. LAMP cannot handle survival data statistics, and hence we extended LAMP for the log-rank test, making it more appropriate for clinical data, with newly introduced theoretical lower bound of the p-value.ResultsWe applied the proposed method to gene combination detection for cancer and obtained gene interactions with statistically significant log-rank p-values. Gene combinations with orders of up to 32 genes were detected by our algorithm, and effects of some genes in these combinations are also supported by existing literature.ConclusionThe novel approach for detecting prognostic markers presented here can identify statistically significant markers with no limitations on the order of interaction. Furthermore, it can be applied to different types of genomic data, provided that binarization is possible.

Computer Science Education for Primary and Lower Secondary School Students: Teaching the Concept of Automata

We explore the feasibility of early introduction to automata theory through gamification. We designed a puzzle game that players can answer correctly if they understand the fundamental concepts of automata theory. In our investigation, 90 children played the game, and their actions were recorded in play logs. An analysis of the play logs shows that approximately 60% of the children achieved correct-answer rates of at least 70%, which suggests that primary and lower secondary school students can understand the fundamental concepts of automata theory. Meanwhile, our analysis shows that most of them do not fully understand automata theory, but some of them have a good understanding of the concept.