Prostate Cancer (PC) has been shown to become an epidemic among men in the world. Early detection of PC is essential for treatment. Biopsies are often done to determine the Gleason score of PC which helps to predict the aggressiveness of PC. As biopsies may cause harm especially for old people, machine learning can be used to predict the Gleason grade of PC from clinical biomarkers that are typically structured in a table. In this paper, we present a comparative study of various machine learning methods to detect the Gleason grade of PC from tabular data. We also investigate the performance of advanced deep learning architectures specialized to deal with tabular data, such as TabNet, for this purpose. Moreover, we propose to build an ensemble of the best performing classifiers to grade PC with a promising performance.

Linear regression classification (LRC) has proven to be a successful recognition tool in recent years. LRC depends on using the least square algorithm to get the solution of the linear regression equation. To improve the performance of the LRC algorithm, in this paper, we extend the LRC strategy to both quaternion and reduced biquaternion domains to consider image color information. We derive closed-form solutions from the properties of both domains.We also improve on the accuracy of the closed-form solutions using nonlinear optimization. Our experiments on three benchmark color face recognition databases demonstrate the effectiveness of the proposed methods for recognizing color faces.

Kidney segmentation from Dynamic Contrast Enhanced Magnetic Resonance Images (DCE-MRI) is a fundamental step for the early detection of transplanted kidney function. This paper presents an accurate and automatic DCE-MRI kidney segmentation method which combines fuzzy c-means (FCM) algorithm and geometric deformable model (level set) method. In order to precisely extract the kidney from its background, the evolution of the level set contour in the proposed method is controlled by the fuzzy memberships of the pixels and both population-based and patient-specific shape model. The FCM algorithm is used to initially divide the input image into kidney and background clusters. The obtained fuzzy clustering membership is used to define the initial contour of the level set method. For segmenting the kidney of a specific patient, a number of high contrast time-point images are segmented constraining the evolution of the level set contour by the population-based shape model constructed from different subjects. As more images are segmented, the patient-specific shape model is built from the obtained segmentation results and gradually used to guide the evolution of the level set contour. The performance of the proposed method is evaluated on 40 subjects. Experimental results demonstrate the efficiency, consistency, and accuracy of the proposed method especially for low contrast images.

The purpose of this work is to develop a computer-aided diagnosis (CAD) system for detecting and localizing prostate cancer from diffusion-weighted magnetic resonance imaging (DWI) acquired at five distinct b-values. The first step in the proposed system depends on nonnegative matrix factorization (NMF) to fuse intensity features of prostate voxels, spatial features of neighboring voxels, and shape prior features to guide the evolution of a level set function for accurate prostate segmentation. The second step in the proposed system involves calculating the apparent diffusion coefficient (ADC) maps of the segmented prostate regions as a discriminating feature between malignant and healthy cases. These ADC maps are used in the last step of the CAD system to train a convolutional neural network (CNN)-based model to identify the ADC maps with malignant tumors. To evaluate the accuracy of the system, 50% of the ADC maps are randomly chosen to train the CNN-model while the second 50% of the ADC maps are used to evaluate the accuracy of the trained model. The proposed CAD system resulted in an average area under the receiver operating characteristic curve (AUC) of 0.93 at the five b-values.

Non-invasive evaluation of renal transplant function is essential to minimize and manage renal rejection. A computer-assisted diagnostic (CAD) system was developed to evaluate kidney function post-transplantation. The developed CAD system utilizes the amount of blood-oxygenation extracted from 3D (2D + time) blood oxygen level-dependent magnetic resonance imaging (BOLD-MRI) to estimate renal function. BOLD-MRI scans were acquired at five different echo-times (2, 7, 12, 17, and 22) ms from 15 transplant patients. The developed CAD system first segments kidneys using the level-sets method followed by estimation of the amount of deoxyhemoglobin, also known as apparent relaxation rate (R2*). These R2* estimates were used as discriminatory features (global features (mean R2*) and local features (pixel-wise R2*)) to train and test state-of-the-art machine learning classifiers to differentiate between non-rejection (NR) and acute renal rejection. Using a leave-one-out cross-validation approach along with an artificial neural network (ANN) classifier, the CAD system demonstrated 93.3% accuracy, 100% sensitivity, and 90% specificity in distinguishing AR from non-rejection . These preliminary results demonstrate the efficacy of the CAD system to detect renal allograft status non-invasively.

Kidney segmentation from Dynamic Contrast Enhanced Magnetic Resonance Images (DCE-MRI) is an important step for the early detection of transplanted kidney rejection. In this paper, an accurate kidney segmentation method from DCE-MRI is proposed. In the proposed method, fuzzy c-means (FCM) algorithm is combined with a geometric deformable model (level set) method to accurately extract the kidney from its background. The FCM algorithm is applied to the input image and the obtained result is used as the initial contour for the level set method. The evolution of the level set boundary is controlled using the kidney shape prior model and the memberships of the pixels computed using the FCM algorithm. The proposed method has been tested on 40 subjects, and experimental results confirm the efficiency, reliability, and accuracy of the proposed method.

The following topics are dealt with: learning (artificial intelligence); feature extraction; image classification; feedforward neural nets; neural nets; convolution; object detection; image segmentation; face recognition; image representation.

This paper proposes an approach for modeling textured 3D non-rigid models based on Weighted Heat Kernel Signature(W-HKS). As a first contribution, we show how to include photometric information as a weight over the shape manifold, we also propose a novel formulation for heat diffusion over weighted manifolds. As a second contribution we present a new discretization method for the proposed equation using finite element approximation. Finally, the weighted heat kernel signature is used as a shape descriptor. The proposed descriptor encodes both the photometric, and geometric information based on the solution of one equation. We also propose a new method to introduce the scale invariance for the weighted heat kernel signature. The performance is tested on two benchmark datasets. The results have indeed confirmed the high performance of the proposed approach on the textured shape retrieval problem, and showed that the proposed method is useful in coping with different challenges of shape analysis where pure geometric and pure photometric methods fail.

The basic idea of shape from shading is to infer the shape of a surface from its shading information in a single image. Since this problem is ill-posed, a number of simplifying assumptions have been often used. However they rarely hold in practice. This paper presents a simple shading-correction algorithm that transforms the image to a new image that better satisfies the assumptions typically needed by existing algorithms, thus improving the accuracy of shape recovery. The algorithm takes advantage of some local shading measures that have been driven under these assumptions. The method is successfully evaluated on real data of human teeth with ground-truth 3D shapes.

Image formation involves understanding the sensors characteristics and object reflectance. In dentistry, for example an accurate three‐dimensional (3D) representation of the human jaw may be used for diagnostic and treatment purposes. Photogrammetry can offer a flexible, cost‐effective solution in that regard. Nonetheless there are several challenges, such as non‐friendly image acquisition environment inside the human mouth, problems with lighting (specularity effects because of saliva, gum discolourisation, and occlusion because of the tongue in the lower jaw), and errors because of the data acquisition sensors (e.g. camera calibration errors, lens distortion and so on). In this study, the authors focus on the 3D surface reconstruction aspect for human jaw modelling based on physical surface characteristics and sensor properties. Owing to apparent lens distortion imposed by near‐field imaging, the authors propose a new flexible calibration for lens radial distortion based on a single image of a sphere. The authors propose a non‐Lambertian shape‐from‐shading (SFS) algorithm under perspective projection which benefits from camera calibration parameters. Our experiments provide quantitative metric results for the proposed approach. The reflectance of the tooth surface is modelled by the Oren–Nayar reflectance model for rough surfaces whose roughness parameter is physically computed from an optical surface profiler measurements. As compared to state‐of‐the‐art SFS approaches, our approach is able to recover geometric details of tooth occlusal surface. This work is fundamental for establishing an optical‐based approach for reconstructing the human jaw, that is inexpensive and does not use ionising radiation.

This paper addresses the problem of fitting a functional model to data corrupted with outliers using a multilayered feed-forward neural network. Although it is of high importance in practical applications, this problem has not received careful attention from the neural network research community. One recent approach to solving this problem is to use a neural network training algorithm based on the random sample consensus (RANSAC) framework. This paper proposes a new algorithm that offers two enhancements over the original RANSAC algorithm. The first one improves the algorithm accuracy and robustness by employing an M-estimator cost function to decide on the best estimated model from the randomly selected samples. The other one improves the time performance of the algorithm by utilizing a statistical pretest based on Wald's sequential probability ratio test. The proposed algorithm is successfully evaluated on synthetic and real data, contaminated with varying degrees of outliers, and compared with existing neural network training algorithms.

This paper addresses the problem of lens distortion calibration from a single image of a smooth, textureless surface. To the best of our knowledge, this has not been addressed before in the literature. We show that this is possible, both theoretically and practically, taking advantage of some local shading measures that vary nonlinearly as a function of lens distortion. The proposed method is easy to use and requires no specific feature extraction from images. Experiments on simulations and real data are reported and compared to well-known techniques.

Polarization imaging can give information about surface shape, and roughness. Polarization has been used for shape recovery, but with convex/concave reconstruction ambiguity. In this paper, we present a direct method to shape recovery using both polarization and shading that resolves this ambiguity, without the need for nonlinear optimization routines. Several experiments on synthetic and real datasets are reported to evaluate the proposed method. The method consistently outperforms some well-known methods based on polarization information alone.

This paper introduces a fuzzy framework for the simultaneous segmentation and registration in addition to bias field correction of MRI datasets. The framework utilizes prior information which may be available about the tissues' mean intensities and tissues' distribution through the datasets. Moreover it works on the given, original image intensities without any logarithmic transformation and thus produces more accurate results and faster performance. The algorithm is evaluated using simulated and real brain MRI data. The results show that the algorithm has indeed improved the segmentation accuracy.

This paper addresses the problem of fitting a functional model to data corrupted with outliers using a multilayered feed-forward neural network. The importance of this problem stems from the vast, diverse, practical applications of neural networks as data-driven function approximator or model estimator. Yet, the challenges raised by the presence of outliers in the data have not received the same careful attention from the neural network research community. The paper proposes an enhanced algorithm to train neural networks for robust function approximation in a random sample consensus (RANSAC) framework. The new algorithm follows the same strategy of the original RANSAC algorithm, but employs an M-estimator cost function to decide the best estimated model. The proposed algorithm is evaluated on synthetic data, contaminated with varying degrees of outliers, and compared to existing neural network training algorithms.

This paper addresses the problem of fitting a functional model to data corrupted with outliers using a multilayered feed-forward neural network (MFNN). Almost all previous efforts to solve this problem have focused on using a training algorithm that minimizes an M-estimator based error criterion. However the robustness gained from M-estimators is still low. Using a training algorithm based on the RANdom SAmple Consensus (RANSAC) framework improves significantly the robustness of the algorithm. However the algorithm typically requires prolonged period of time before a final solution is reached. In this paper, we propose a new strategy to improve the time performance of the RANSAC algorithm for training MFNNs. A statistical pre-test based on Wald's sequential probability ratio test (SPRT) is performed on each randomly generated sample to decide whether it deserves to be used for model estimation. The proposed algorithm is evaluated on synthetic data, contaminated with varying degrees of outliers, and have demonstrated faster performance compared to the original RANSAC algorithm with no significant sacrifice of the robustness.

This paper introduces a new formula for the objective function of the famous fuzzy C-means algorithm. Two weighted terms are added to the objective function to reflect any available information about the class center and class pixels distribution throughout the datasets. The algorithm is evaluated for the task of the segmentation of medical MRI brain volume. The results show that the algorithm has a considerable robustness against noise and partial volume effects, and it needs a smaller number of iterations to reach convergence compared with other similar algorithms.

This paper addresses the issue of assessing the quality of the clusters found by fuzzy and hard clustering algorithms. In particular, it seeks an answer to the question on how well cluster validity indexes can automatically determine the appropriate number of clusters that represent the data. The paper surveys several key existing solutions for cluster validity in the domain of image segmentation. In addition, it suggests two new indexes. The first one is based on Akaike's information criterion (AIC). While AIC was devoted to other domains such as statistical estimation of model fitting, it is implemented here for the first time as a validation index. The second index is developed from the well-established idea of cross-validation. The existing and new indexes are evaluated and compared on several synthetic images corrupted with noise of varying levels and volumetric MR data.

In this paper, we propose a variational method to segment image objects, which have a given parametric shape based on a level-set formulation of the Mumford-Shah functional, and the shape parameters. We define an energy functional composed by two complementary terms. The first one detects object boundaries using a Chan-Vese-like method. The second term constrains the contour to find a shape compatible with the parametric shape. The segmentation of the object of interest is given by the minimum of our energy functional. This minimum is computed with the calculus of variation and the gradient descent method that provide a system of evolution equations solved with the well-known level set method. We focus in this paper on the parametric category of image linear objects. Applications of the proposed model are presented on synthetic and real images.

Since the pioneer work of Horn, a considerable amount of computer vision research has been done on shape from shading (SFS). The basic idea of SFS is to infer the shape of an object from its shading information in a single image. Since this problem is ill-posed, a number of assumptions have been used extensively in the computer vision community for the SFS problem, such as orthographic projection, Lambertian reflectance model, single light source, and constant surface albedo. In this paper, starting with this typical set of assumptions, we derive new image intrinsic values based on image shading information. We derive these values for a number of reflectance models such as the linear and quadratic models in addition to the popular Lambertian model. We validated the obtained intrinsic values on hundreds of real and synthetic images.

This paper focuses on evaluating a number of statistical and neural methods for supervised, pixel-wise remote-sensing image classification and decision fusion. Despite the enormous progress in the analysis of remote sensing imagery over the past three decades, still much is desired in the area of image classification as no specific algorithm is known to provide accurate results under all circumstances. Decision fusion may be pursued to combine the outputs of different classifiers applied on the same data, in the hope of combining the best of what each approach provides. We report the results of the comparison between several classification and fusion methods on two real datasets, one of which is the standard benchmark Satimage dataset. It is shown that the fusion approaches can indeed outperform the performance of the best classifier.

We address the problem of calibrating camera lens distortion, which can be significant in medium to wide angle lenses. While almost all existing nonmetric distortion calibration methods need user involvement in one form or another, we present an automatic approach based on the robust the-least-median-of-squares (LMedS) estimator. Our approach is thus less sensitive to erroneous input data such as image curves that are mistakenly considered as projections of 3D linear segments. Our approach uniquely uses fast, closed-form solutions to the distortion coefficients, which serve as an initial point for a nonlinear optimization algorithm to straighten imaged lines. Moreover we propose a method for distortion model selection based on geometrical inference. Successful experiments to evaluate the performance of this approach on synthetic and real data are reported.

Correspondence is one of the major problems that must be solved in stereo vision. Correlation has been commonly used in the past for this problem. However, most classical linear correlation methods fail near depth discontinuities and in the presence of occlusions. Many robust methods have been proposed that claim to effectively deal with some or all of these issues. Many of these robust methods are transformation-based, however, other robust methods are non-transformation based. This paper gives five requirements that should be met by a transformation-based robust correlation method. We compare some of the robust correspondence methods and demonstrate their utility on different data sets. Based on these results, we propose a solution to the correspondence problem which represents a compromise between the speed of classical correlation and the improved results obtained from a more robust correspondence method. Also, we propose a median filtering technique that removes noise from the disparity maps while preserving certain image features usually removed by ordinary median filtering.