Annalisa Barla

Histogram intersection kernel for image classification

In this paper we address the problem of classifying images, by exploiting global features that describe color and illumination properties, and by using the statistical learning paradigm. The contribution of this paper is twofold. First, we show that histogram intersection has the required mathematical properties to be used as a kernel function for support vector machines (SVMs). Second, we give two examples of how a SVM, equipped with such a kernel, can achieve very promising results on image classification based on color information.

Building kernels from binary strings for image matching

In the statistical learning framework, the use of appropriate kernels may be the key for substantial improvement in solving a given problem. In essence, a kernel is a similarity measure between input points satisfying some mathematical requirements and possibly capturing the domain knowledge. We focus on kernels for images: we represent the image information content with binary strings and discuss various bitwise manipulations obtained using logical operators and convolution with nonbinary stencils. In the theoretical contribution of our work, we show that histogram intersection is a Mercers kernel and we determine the modifications under which a similarity measure based on the notion of Hausdorff distance is also a Mercers kernel. In both cases, we determine explicitly the mapping from input to feature space. The presented experimental results support the relevance of our analysis for developing effective trainable systems.

Algebraic stability indicators for ranked lists in molecular profiling

MOTIVATION We propose a method for studying the stability of biomarker lists obtained from functional genomics studies. It is common to adopt resampling methods to tune and evaluate marker-based diagnostic and prognostic systems in order to prevent selection bias. Such caution promotes honest estimation of class prediction, but leads to alternative sets of solutions. In microarray studies, the difference in lists may be bewildering, also due to the presence of modules of functionally related genes. Methods for assessing stability understand the dependency of the markers on the data or on the predictors type and help selecting solutions. RESULTS A computational framework for comparing sets of ranked biomarker lists is presented. Notions and algorithms are based on concepts from permutation group theory. We introduce several algebraic indicators and metric methods for symmetric groups, including the Canberra distance, a weighted version of Spearmans footrule. We also consider distances between partial lists and an aggregation of sets of lists into an optimal list based on voting theory (Borda count). The stability indicators are applied in practical situations to several synthetic, cancer microarray and proteomics datasets. The addressed issues are predictive classification, presence of modules, comparison of alternative biomarker lists, outlier removal, control of selection bias by randomization techniques and enrichment analysis. AVAILABILITY Supplementary Material and software are available at the address http://biodcv.fbk.eu/listspy.html

The l1-l2 regularization framework unmasks the hypoxia signature hidden in the transcriptome of a set of heterogeneous neuroblastoma cell lines

BackgroundGene expression signatures are clusters of genes discriminating different statuses of the cells and their definition is critical for understanding the molecular bases of diseases. The identification of a gene signature is complicated by the high dimensional nature of the data and by the genetic heterogeneity of the responding cells. The l1-l2 regularization is an embedded feature selection technique that fulfills all the desirable properties of a variable selection algorithm and has the potential to generate a specific signature even in biologically complex settings. We studied the application of this algorithm to detect the signature characterizing the transcriptional response of neuroblastoma tumor cell lines to hypoxia, a condition of low oxygen tension that occurs in the tumor microenvironment.ResultsWe determined the gene expression profile of 9 neuroblastoma cell lines cultured under normoxic and hypoxic conditions. We studied a heterogeneous set of neuroblastoma cell lines to mimic the in vivo situation and to test the robustness and validity of the l1-l2 regularization with double optimization. Analysis by hierarchical, spectral, and k-means clustering or supervised approach based on t-test analysis divided the cell lines on the bases of genetic differences. However, the disturbance of this strong transcriptional response completely masked the detection of the more subtle response to hypoxia. Different results were obtained when we applied the l1-l2 regularization framework. The algorithm distinguished the normoxic and hypoxic statuses defining signatures comprising 3 to 38 probesets, with a leave-one-out error of 17%. A consensus hypoxia signature was established setting the frequency score at 50% and the correlation parameter ε equal to 100. This signature is composed by 11 probesets representing 8 well characterized genes known to be modulated by hypoxia.ConclusionWe demonstrate that l1-l2 regularization outperforms more conventional approaches allowing the identification and definition of a gene expression signature under complex experimental conditions. The l1-l2 regularization and the cross validation generates an unbiased and objective output with a low classification error. We feel that the application of this algorithm to tumor biology will be instrumental to analyze gene expression signatures hidden in the transcriptome that, like hypoxia, may be major determinant of the course of the disease.

Multi-output learning via spectral filtering

In this paper we study a class of regularized kernel methods for multi-output learning which are based on filtering the spectrum of the kernel matrix. The considered methods include Tikhonov regularization as a special case, as well as interesting alternatives such as vector-valued extensions of L2 boosting and other iterative schemes. Computational properties are discussed for various examples of kernels for vector-valued functions and the benefits of iterative techniques are illustrated. Generalizing previous results for the scalar case, we show a finite sample bound for the excess risk of the obtained estimator, which allows to prove consistency both for regression and multi-category classification. Finally, we present some promising results of the proposed algorithms on artificial and real data.

Hausdorff Kernel for 3D Object Acquisition and Detection

Learning one class at a time can be seen as an effective solution to classification problems in which only the positive examples are easily identifiable. A kernel method to accomplish this goal consists of a representation stage - which computes the smallest sphere in feature space enclosingthe positive examples - and a classification stage - which uses the obtained sphere as a decision surface to determine the positivity of new examples. In this paper we describe a kernel well suited to represent, identify, and recognize 3D objects from unconstrained images. The kernel we introduce, based on Hausdorff distance, is tailored to deal with grey-level image matching. The effectiveness of the proposed method is demonstrated on several data sets of faces and objects of artistic relevance, like statues.

Effect of Size and Heterogeneity of Samples on Biomarker Discovery: Synthetic and Real Data Assessment

Motivation The identification of robust lists of molecular biomarkers related to a disease is a fundamental step for early diagnosis and treatment. However, methodologies for the discovery of biomarkers using microarray data often provide results with limited overlap. These differences are imputable to 1) dataset size (few subjects with respect to the number of features); 2) heterogeneity of the disease; 3) heterogeneity of experimental protocols and computational pipelines employed in the analysis. In this paper, we focus on the first two issues and assess, both on simulated (through an in silico regulation network model) and real clinical datasets, the consistency of candidate biomarkers provided by a number of different methods. Methods We extensively simulated the effect of heterogeneity characteristic of complex diseases on different sets of microarray data. Heterogeneity was reproduced by simulating both intrinsic variability of the population and the alteration of regulatory mechanisms. Population variability was simulated by modeling evolution of a pool of subjects; then, a subset of them underwent alterations in regulatory mechanisms so as to mimic the disease state. Results The simulated data allowed us to outline advantages and drawbacks of different methods across multiple studies and varying number of samples and to evaluate precision of feature selection on a benchmark with known biomarkers. Although comparable classification accuracy was reached by different methods, the use of external cross-validation loops is helpful in finding features with a higher degree of precision and stability. Application to real data confirmed these results.

Old fashioned state-of-the-art image classification

In this paper we present a statistical learning scheme for image classification based on a mixture of old fashioned ideas and state of the art learning tools. We represent input images through large dimensional and usually sparse histograms which, depending on the task, are either color histograms or co-occurrence matrices. Support vector machines are trained on these sparse inputs directly, to solve problems like indoor/outdoor classification and cityscape retrieval from image databases. The experimental results indicate that the use of a kernel function derived from the computer vision literature leads to better recognition results than off the shelf kernels. According to our findings, it appears that image classification problems can be addressed with no need of explicit feature extraction or dimensionality reduction stages. We argue that this might be used as the starting point for developing image classification systems which can be easily tuned to a number of different tasks.

Comparison of Clinical Outcomes for Patients with Large Choroidal Melanoma after Primary Treatment with Enucleation or Proton Beam Radiotherapy

Purpose: To evaluate survival and clinical outcome for patients with a large uveal melanoma treated by either enucleation or proton beam radiotherapy (PBRT). Procedures: This retrospective non-randomized study evaluated 132 consecutive patients with T3 and T4 choroidal melanoma classified according to TNM stage grouping. Results: Cumulative all-cause mortality, melanoma-related mortality and metastasis-free survival were not statistically different between the two groups (log-rank test, p = 0.56, p = 0.99 and p = 0.25, respectively). Eye retention of the tumours treated with PBRT at 5 years was 74% (SD 6.2%). In these patients at diagnosis, 73% of eyes had a best-corrected visual acuity (BCVA) of 0.1 or better. After 12 and 60 months, BCVA of 0.1 or better was observed in 47.5 and 32%, respectively. Conclusion and Message: Although enucleation is the most common primary treatment for large uveal melanomas, PBRT is an eye-preserving option that may be considered for some patients.

Proton beam radiotherapy of uveal melanoma: Italian patients treated in Nice, France.

Purpose To evaluate the results of 15 years of experience with proton beam radiotherapy in the treatment of intraocular melanoma, and to determine univariate and multivariate risk factors for local failure, eye retention, and survival. Methods A total of 368 cases of intraocular melanoma were treated with proton beam radiotherapy at Centre Lacassagne Cyclotron Biomedical of Nice, France, between 1991 and 2006. Actuarial methods were used to evaluate rate of local tumor control, eye retention, and survival after proton beam radiotherapy. Cox regression models were extracted to evaluate univariate risk factors, while regularized least squares algorithm was used to have a multivariate classification model to better discriminate risk factors. Results Tumor relapse occurred in 8.4% of the eyes, with a median recurrence time of 46 months. Enucleation was performed on 11.7% of the eyes after a median time of 49 months following proton beam; out of these, 29 eyes were enucleated due to relapse and 16 due to other causes. The univariate regression analysis identified tumor height and diameter as primary risk factors for enucleation. Regularized least squares analysis demonstrated the higher effectiveness of a multivariate model of five risk factors (macula distance, optic disc distance, tumor height, maximum diameter, and age) in discriminating relapsed vs nonrelapsed patients. Conclusions This data set, which is the largest in Italy with relatively long-term follow-up, demonstrates that a high rate of tumor control, survival, and eye retention were achieved after proton beam irradiation, as in other series.