Neural Network Based Filtering Method for Cancer Detection

Objective: The growth of anomalous cells in the human body in an uncontrolled manner is characterized as cancer. The detection of cancer is a multi-stage process in the clinical examination. Methods: It is mainly involved with the assistance of radiological imaging. The imaging technique is used to identify the spread of cancer in the human body. This imaging-based detection can be improved by incorporating the Image Processing methodologies. In image processing, the preprocessing is applied at the lower-level abstraction. It removes the unwanted noise pixel present in the image, which also distributes the pixel values based on the specific distribution method. Results: Neural Network is a learning and processing engine, which is mainly used to create cognitive intelligence in various domains. In this work, the Neural Network (NN) based filtering approach is developed to improve the preprocessing operation in the cancer detection process. Conclusion: The performance of the proposed filtering method is compared with the existing linear and non-linear filters in terms of Mean Squared Error (MSE), Peak Signal to Noise Ratio (PSNR) and Image Enhancement Factor (IEF).

Normalization by orientation anisotropy in human V1-V3

An influential account of neuronal responses in primary visual cortex is the normalized energy model. This model is often implemented as a two-stage computation. The first stage is the extraction of contrast energy, whereby a complex cell computes the squared and summed outputs of a pair of linear filters in quadrature phase. The second stage is normalization, in which a local population of complex cells mutually inhibit one another. Because the population includes cells tuned to a range of orientations and spatial frequencies, the result is that the responses are effectively normalized by the local stimulus contrast. Here, using evidence from human functional MRI, we show that the classical model fails to account for the relative responses to two classes of stimuli: straight, parallel, band-passed contours (gratings), and curved, band-passed contours (snakes). The snakes elicit fMRI responses that are about twice as large as the gratings, yet traditional energy models, including normalized energy models, predict responses that are about the same. Here, we propose a computational model, in which responses are normalized not by the sum of the contrast energy, but by the orientation anisotropy, computed as the variance in contrast energy across orientation channels. We first show that this model accounts for differential responses to these two classes of stimuli. We then show that the model successfully generalizes to other band-pass textures, both in V1 and in extrastriate cortex (V2 and V3). We speculate that high anisotropy in the orientation responses leads to larger outputs in downstream areas, which in turn normalizes responses in these later visual areas, as well as in V1 via feedback.

Gaussian Processes for Regression and Optimisation

<p>Gaussian processes have proved to be useful and powerful constructs for the purposes of regression. The classical method proceeds by parameterising a covariance function, and then infers the parameters given the training data. In this thesis, the classical approach is augmented by interpreting Gaussian processes as the outputs of linear filters excited by white noise. This enables a straightforward definition of dependent Gaussian processes as the outputs of a multiple output linear filter excited by multiple noise sources. We show how dependent Gaussian processes defined in this way can also be used for the purposes of system identification. Onewell known problem with Gaussian process regression is that the computational complexity scales poorly with the amount of training data. We review one approximate solution that alleviates this problem, namely reduced rank Gaussian processes. We then show how the reduced rank approximation can be applied to allow for the efficient computation of dependent Gaussian processes. We then examine the application of Gaussian processes to the solution of other machine learning problems. To do so, we review methods for the parameterisation of full covariance matrices. Furthermore, we discuss how improvements can be made by marginalising over alternative models, and introduce methods to perform these computations efficiently. In particular, we introduce sequential annealed importance sampling as a method for calculating model evidence in an on-line fashion as new data arrives. Gaussian process regression can also be applied to optimisation. An algorithm is described that uses model comparison between multiple models to find the optimum of a function while taking as few samples as possible. This algorithm shows impressive performance on the standard control problem of double pole balancing. Finally, we describe how Gaussian processes can be used to efficiently estimate gradients of noisy functions, and numerically estimate integrals.</p>

Gaussian Processes for Regression and Optimisation

<p>Gaussian processes have proved to be useful and powerful constructs for the purposes of regression. The classical method proceeds by parameterising a covariance function, and then infers the parameters given the training data. In this thesis, the classical approach is augmented by interpreting Gaussian processes as the outputs of linear filters excited by white noise. This enables a straightforward definition of dependent Gaussian processes as the outputs of a multiple output linear filter excited by multiple noise sources. We show how dependent Gaussian processes defined in this way can also be used for the purposes of system identification. Onewell known problem with Gaussian process regression is that the computational complexity scales poorly with the amount of training data. We review one approximate solution that alleviates this problem, namely reduced rank Gaussian processes. We then show how the reduced rank approximation can be applied to allow for the efficient computation of dependent Gaussian processes. We then examine the application of Gaussian processes to the solution of other machine learning problems. To do so, we review methods for the parameterisation of full covariance matrices. Furthermore, we discuss how improvements can be made by marginalising over alternative models, and introduce methods to perform these computations efficiently. In particular, we introduce sequential annealed importance sampling as a method for calculating model evidence in an on-line fashion as new data arrives. Gaussian process regression can also be applied to optimisation. An algorithm is described that uses model comparison between multiple models to find the optimum of a function while taking as few samples as possible. This algorithm shows impressive performance on the standard control problem of double pole balancing. Finally, we describe how Gaussian processes can be used to efficiently estimate gradients of noisy functions, and numerically estimate integrals.</p>

Generation of Two Correlated Stationary Gaussian Processes

Since correlated stochastic processes are often presented in practical problems, feasible methods to model and generate correlated processes appropriately are needed for analysis and simulation. The present paper systematically presents three methods to generate two correlated stationary Gaussian processes. They are (1) the method of linear filters, (2) the method of series expansion with random amplitudes, and (3) the method of series expansion with random phases. All three methods intend to match the power spectral density for each process but use information of different levels of correlation. The advantages and disadvantages of each method are discussed.

A robust receptive field code for optic flow detection and decomposition during self-motion

The perception of optic flow is essential for any visually guided behavior of a moving animal. To mechanistically predict behavior and understand the emergence of self-motion perception in vertebrate brains, it is essential to systematically characterize the motion receptive fields (RFs) of optic flow processing neurons. Here, we present the fine-scale RFs of thousands of motion-sensitive neurons studied in the diencephalon and the midbrain of zebrafish. We found neurons that serve as linear filters and robustly encode directional and speed information of translation-induced optic flow. These neurons are topographically arranged in pretectum according to translation direction. The unambiguous encoding of translation enables the decomposition of translational and rotational self-motion information from mixed optic flow. In behavioral experiments, we successfully demonstrated the predicted decomposition in the optokinetic and optomotor responses. Together, our study reveals the algorithm and the neural implementation for self-motion estimation in a vertebrate visual system.

Differences in olfactory bulb mitral cell spiking with ortho- and retronasal stimulation revealed by data-driven models

The majority of olfaction studies focus on orthonasal stimulation where odors enter via the front nasal cavity, while retronasal olfaction, where odors enter the rear of the nasal cavity during feeding, is understudied. The coding of retronasal odors via coordinated spiking of neurons in the olfactory bulb (OB) is largely unknown despite evidence that higher level processing is different than orthonasal. To this end, we use multi-electrode array in vivo recordings of rat OB mitral cells (MC) in response to a food odor with both modes of stimulation, and find significant differences in evoked firing rates and spike count covariances (i.e., noise correlations). Differences in spiking activity often have implications for sensory coding, thus we develop a single-compartment biophysical OB model that is able to reproduce key properties of important OB cell types. Prior experiments in olfactory receptor neurons (ORN) showed retro stimulation yields slower and spatially smaller ORN inputs than with ortho, yet whether this is consequential for OB activity remains unknown. Indeed with these specifications for ORN inputs, our OB model captures the salient trends in our OB data. We also analyze how first and second order ORN input statistics dynamically transfer to MC spiking statistics with a phenomenological linear-nonlinear filter model, and find that retro inputs result in larger linear filters than ortho inputs. Finally, our models show that the temporal profile of ORN is crucial for capturing our data and is thus a distinguishing feature between ortho and retro stimulation, even at the OB. Using data-driven modeling, we detail how ORN inputs result in differences in OB dynamics and MC spiking statistics. These differences may ultimately shape how ortho and retro odors are coded.

Low level visual features support robust material perception in the judgement of metallicity

The human visual system is able to rapidly and accurately infer the material properties of objects and surfaces in the world. Yet an inverse optics approach—estimating the bi-directional reflectance distribution function of a surface, given its geometry and environment, and relating this to the optical properties of materials—is both intractable and computationally unaffordable. Rather, previous studies have found that the visual system may exploit low-level spatio-chromatic statistics as heuristics for material judgment. Here, we present results from psychophysics and modeling that supports the use of image statistics heuristics in the judgement of metallicity—the quality of appearance that suggests an object is made from metal. Using computer graphics, we generated stimuli that varied along two physical dimensions: the smoothness of a metal object, and the evenness of its transparent coating. This allowed for the exploration of low-level image statistics, whilst ensuring that each stimulus was a naturalistic, physically plausible image. A conjoint-measurement task decoupled the contributions of these dimensions to the perception of metallicity. Low-level image features, as represented in the activations of oriented linear filters at different spatial scales, were found to correlate with the dimensions of the stimulus space, and decision-making models using these activations replicated observer performance in perceiving differences in metal smoothness and coating bumpiness, and judging metallicity. Importantly, the performance of these models did not deteriorate when objects were rotated within their simulated scene, with corresponding changes in image properties. We therefore conclude that low-level image features may provide reliable cues for the robust perception of metallicity.

Performance Analysis of Different Restoration Techniques for Degraded 3D Cervical Spine Mri Image

In the proposed work, noise models such as Salt & Pepper, Gaussian and Poisson are considered in order to corrupt the image.Image restoration is still challenging task to recover an original image using a degradation and restoration model. In the paper, Gaussian, Average and Wiener linear image restoration techniques are used to recover the original MRI image. Median filter, Min and Max non-linear filters are also used to obtain uncorrupt image in the paper.Mean square error (MSE), Peak Signal to Noise Ratio (PSNR) and Cross Correlation(CC) performance analysis criteria are used to compare different restoration technique so that better performance in a clinical diagnosis can be achieved. In the paper it can be found that wiener filter with 5 x 5 window for Gaussian, speckle and Poisson noise provides best performance in terms of MSE and PSNR. Also,median filter with 5 x 5 window gives better accuracy of results to restore 3D salt & pepper noised image in terms of MSE and CC.

The Impact of Linear Filter Preprocessing in the Interpretation of Permutation Entropy

Permutation Entropy (PE) is a powerful tool for measuring the amount of information contained within a time series. However, this technique is rarely applied directly on raw signals. Instead, a preprocessing step, such as linear filtering, is applied in order to remove noise or to isolate specific frequency bands. In the current work, we aimed at outlining the effect of linear filter preprocessing in the final PE values. By means of the Wiener–Khinchin theorem, we theoretically characterize the linear filter’s intrinsic PE and separated its contribution from the signal’s ordinal information. We tested these results by means of simulated signals, subject to a variety of linear filters such as the moving average, Butterworth, and Chebyshev type I. The PE results from simulations closely resembled our predicted results for all tested filters, which validated our theoretical propositions. More importantly, when we applied linear filters to signals with inner correlations, we were able to theoretically decouple the signal-specific contribution from that induced by the linear filter. Therefore, by providing a proper framework of PE linear filter characterization, we improved the PE interpretation by identifying possible artifact information introduced by the preprocessing steps.