Anarta Ghosh

Robustness of shape descriptors to incomplete contour representations

With inspiration from psychophysical researches of the human visual system, we propose a novel aspect and a method for performance evaluation of contour-based shape recognition algorithms regarding their robustness to incompleteness of contours. We use complete contour representations of objects as a reference (training) set. Incomplete contour representations of the same objects are used as a test set. The performance of an algorithm is reported using the recognition rate as a function of the percentage of contour retained. We call this evaluation procedure the ICR test. We consider three types of contour incompleteness, viz. segment-wise contour deletion, occlusion, and random pixel depletion. As an illustration, the robustness of two shape recognition algorithms to contour incompleteness is evaluated. These algorithms use a shape context and a distance multiset as local shape descriptors. Qualitatively, both algorithms mimic human visual perception in the sense that recognition performance monotonously increases with the degree of completeness and that they perform best in the case of random depletion and worst in the case of occluded contours. The distance multiset method performs better than the shape context method in this test framework.

Automatic segmentation and inpainting of specular highlights for endoscopic imaging

Minimally invasive medical procedures have become increasingly common in todays healthcare practice. Images taken during such procedures largely show tissues of human organs, such as the mucosa of the gastrointestinal tract. These surfaces usually have a glossy appearance showing specular highlights. For many visual analysis algorithms, these distinct and bright visual features can become a significant source of error. In this article, we propose two methods to address this problem: (a) a segmentation method based on nonlinear filtering and colour image thresholding and (b) an efficient inpainting method. The inpainting algorithm eliminates the negative effect of specular highlights on other image analysis algorithms and also gives a visually pleasing result. The methods compare favourably to the existing approaches reported for endoscopic imaging. Furthermore, in contrast to the existing approaches, the proposed segmentation method is applicable to the widely used sequential RGB image acquisition systems.

Learning vector quantization: The dynamics of winner-takes-all algorithms

Winner-Takes-All (WTA) prescriptions for learning vector quantization (LVQ) are studied in the framework of a model situation: two competing prototype vectors are updated according to a sequence of example data drawn from a mixture of Gaussians. The theory of on-line learning allows for an exact mathematical description of the training dynamics, even if an underlying cost function cannot be identified. We compare the typical behavior of several WTA schemes including basic LVQ and unsupervised vector quantization. The focus is on the learning curves, i.e. the achievable generalization ability as a function of the number of training examples.

Performance analysis of LVQ algorithms: a statistical physics approach

Learning vector quantization (LVQ) constitutes a powerful and intuitive method for adaptive nearest prototype classification. However, original LVQ has been introduced based on heuristics and numerous modifications exist to achieve better convergence and stability. Recently, a mathematical foundation by means of a cost function has been proposed which, as a limiting case, yields a learning rule similar to classical LVQ2.1. It also motivates a modification which shows better stability. However, the exact dynamics as well as the generalization ability of many LVQ algorithms have not been thoroughly investigated so far. Using concepts from statistical physics and the theory of on-line learning, we present a mathematical framework to analyse the performance of different LVQ algorithms in a typical scenario in terms of their dynamics, sensitivity to initial conditions, and generalization ability. Significant differences in the algorithmic stability and generalization ability can be found already for slightly different variants of LVQ. We study five LVQ algorithms in detail: Kohonens original LVQ1, unsupervised vector quantization (VQ), a mixture of VQ and LVQ, LVQ2.1, and a variant of LVQ which is based on a cost function. Surprisingly, basic LVQ1 shows very good performance in terms of stability, asymptotic generalization ability, and robustness to initializations and model parameters which, in many cases, is superior to recent alternative proposals.

Window-based example selection in learning vector quantization

A variety of modifications have been employed to learning vector quantization (LVQ) algorithms using either crisp or soft windows for selection of data. Although these schemes have been shown in practice to improve performance, a theoretical study on the influence of windows has so far been limited. Here we rigorously analyze the influence of windows in a controlled environment of gaussian mixtures in high dimensions. Concepts from statistical physics and the theory of online learning allow an exact description of the training dynamics, yielding typical learning curves, convergence properties, and achievable generalization abilities. We compare the performance and demonstrate the advantages of various algorithms, including LVQ 2.1, generalized LVQ (GLVQ), Learning from Mistakes (LFM) and Robust Soft LVQ (RSLVQ). We find that the selection of the window parameter highly influences the learning curves but not, surprisingly, the asymptotic performances of LVQ 2.1 and RSLVQ. Although the prototypes of LVQ 2.1 exhibit divergent behavior, the resulting decision boundary coincides with the optimal decision boundary, thus yielding optimal generalization ability.

Indistinct Frame Detection in Colonoscopy Videos

An automated system for analysis of colonoscopy videos is expected to complement the expertise and the experience of a medical professional in: (a) detecting lesions and (b) assessing the quality of a given procedure. Colonoscopy videos contain a significant number of frames which do not carry any clinical information. The presence of such frames would slow down or cause the failure of the processing steps of such an automated system. Furthermore, many existing metrics to measure the quality of the colonoscopy procedures directly involve the number of such indistinct frames present in the videos. We propose a novel algorithm to detect indistinct frames based on the wavelet analysis. The L2 norm of the detail coefficients of the wavelet decomposition of a colonoscopy image is considered as the feature vector of the proposed classification system. The algorithm was tested on a manually labeled, balanced data set. It achieved an accuracy of 99.59% in a leave-two-out cross validation procedure based on Bayesian classification. Furthermore, when applied to full colonoscopy videos, the presented algorithm detected 26.2% of the frames as indistinct, of which 92.3% were correctly classified. The proposed method outperforms the current best performing algorithm both in terms of accuracy and computation time.

Learning dynamics and robustness of vector quantization and neural gas

Various alternatives have been developed to improve the winner-takes-all (WTA) mechanism in vector quantization, including the neural gas (NG). However, the behavior of these algorithms including their learning dynamics, robustness with respect to initialization, asymptotic results, etc. has only partially been studied in a rigorous mathematical analysis. The theory of on-line learning allows for an exact mathematical description of the training dynamics in model situations. We demonstrate using a system of three competing prototypes trained from a mixture of Gaussian clusters that the NG can improve convergence speed and achieves robustness to initial conditions. However, depending on the structure of the data, the NG does not always obtain the best asymptotic quantization error.

Pilot evaluation of a ward-based automated hand hygiene training system.

A novel artificial intelligence (AI) system (SureWash; GLANTA, Dublin, Ireland) was placed on a ward with 45 staff members for two 6-day periods to automatically assess hand hygiene technique and the potential effectiveness of the automated training system. Two human reviewers assessed videos from 50 hand hygiene events with an interrater reliability (IIR) of 88% (44/50). The IIR was 88% (44/50) for the human reviewers and 80% (40/50) for the software. This study also investigated the poses missed and the impact of feedback on participation (+113%), duration (+11%), and technique (+2.23%). Our findings showed significant correlation between the human raters and the computer, demonstrating for the first time in a clinical setting the potential use of this type of AI technology in hand hygiene training.

EFFECT OF HIGH CURVATURE POINT DELETION ON THE PERFORMANCE OF TWO CONTOUR BASED SHAPE RECOGNITION ALGORITHMS

Psychophysical researches on the human visual system have shown that the points of high curvature on the contour of an object play an important role in the recognition process. Inspired by these studies we propose: (i) a novel algorithm to select points of high curvature on the contour of an object which can be used to construct a recognizable polygonal approximation, (ii) a test which evaluates the effect of deletion of contour segments containing such points on the performance of contour based object recognition algorithms. We use complete contour representations of objects as a reference (training) set. Incomplete contour representations of the same objects are used as a test set. The performance of an algorithm is reported using the recognition rate as a function of the percentage of contour retained. We consider two types of contour incompleteness obtained by deletion of contour segments of high or low curvature. We illustrate the test procedure using two shape recognition algorithms that deploy a shape context and a distance multiset as local shape descriptors. Both algorithms qualitatively mimic human visual perception in that the deletion of segments of high curvature has a stronger performance degradation effect than the deletion of other parts of the contour. This effect is more pronounced in the performance of the shape context method.

Experts vs. novices: applying eye-tracking methodologies in colonoscopy video screening for polyp search

We present in this paper a novel study aiming at identifying the differences in visual search patterns between physicians of diverse levels of expertise during the screening of colonoscopy videos. Physicians were clustered into two groups -experts and novices- according to the number of procedures performed, and fixations were captured by an eye-tracker device during the task of polyp search in different video sequences. These fixations were integrated into heat maps, one for each cluster. The obtained maps were validated over a ground truth consisting of a mask of the polyp, and the comparison between experts and novices was performed by using metrics such as reaction time, dwelling time and energy concentration ratio. Experimental results show a statistically significant difference between experts and novices, and the obtained maps show to be a useful tool for the characterisation of the behaviour of each group.