Explainable Artificial Intelligence (XAI) has experienced significant growth in recent years. This can be attributed to the increasing adoption of Machine Learning (ML) models, and the need to understand the reasons behind their inferences, as most of them are “black boxes”, lacking interpretability and explainability. To deal with this issue, the most promising approaches include Fuzzy Cognitive Maps (FCMs), which due to their simple yet effective structure to knowledge modeling they are applicable to a wide variety of scientific domains. Although several studies have reviewed the various modifications and applications of FCMs, the potential of the FCMs in terms of interpretable/explainable decision-making has not been sufficiently explored. This review study considers the capacity of FCMs in the context of transparent, interpretable, and explainable ML, with four main contributions: a) it systematically reviews the most recent developments in FCMs, that have been proposed between 2018 and 2024, as well as their prospects in XAI; b) it presents recently proposed learning algorithms, used to train the FCMs; c) it reports the latest applications of FCMs, including state-of-the-art models for interpretable 1D signal and image analysis; d) it identifies limitations of the current FCM models and highlights open challenges and perspectives, indicating novel pathways for innovation towards XAI.

Several deep learning (DL) approaches have been proposed to deal with image classification tasks. However, despite their effectiveness, they lack interpretability, as they are unable to explain or justify their results. To address the challenge of interpretable image classification, this paper introduces a novel framework, named Interpretable Intuitionistic Fuzzy Cognitive Maps (I2FCMs).Intuitionistic FCMs (iFCMs) have been proposed as an extension of FCMs offering a natural mechanism to assess the quality of their output through the estimation of hesitancy, a concept resembling human hesitation in decision making. In the context of image classification, hesitancy is considered as a degree of unconfidence with which an image is categorized to a class. To the best of our knowledge this is the first time iFCMs are applied for image classification. Further novel contributions of the introduced framework include the following: a) a feature extraction process focusing on the most informative image regions; b) a learning algorithm for automatic data-driven determination of the intuitionistic fuzzy interconnections of the iFCM, thereby reducing human intervention in the definition of the graph structure; c) an inherently interpretable classification approach based on image contents, providing understandable explanations of its predictions, using linguistic terms. Furthermore, the proposed I2FCM framework can be applied to DL models, including Convolutional Neural Network (CNN), rendering them interpretable. The effectiveness of I2FCM is evaluated on publicly available datasets, and the results confirm that it can provide enhanced classification performance, while providing interpretable inferences.

Synthetic medical image data can unlock the potential of deep learning (DL)-based clinical decision support (CDS) systems through the creation of large scale, privacy-preserving, training sets. Despite the significant progress in this field, there is still a largely unanswered research question: "How can we quantitatively assess the similarity of a synthetically generated set of images with a set of real images in a given application domain". Today, answers to this question are mainly provided via user evaluation studies, inception-based measures, and the classification performance achieved on synthetic images. This paper proposes a novel measure to assess the similarity between synthetically generated and real sets of images, in terms of their utility for the development of DL-based CDS systems. Inspired by generalized neural additive models, and unlike inception-based measures, the proposed measure is interpretable (Interpretable Utility Similarity, IUS), explaining why a synthetic dataset could be more useful than another one in the context of a CDS system based on clinically relevant image features. The experimental results on publicly available benchmark datasets from various color medical imaging modalities including endoscopic, dermoscopic and fundus imaging, indicate that selecting synthetic images of high utility similarity using IUS can result in relative improvements of up to 54.6% in terms of classification performance. The generality of IUS for synthetic data assessment is demonstrated also for grayscale X-ray and ultrasound imaging modalities. IUS implementation is available at https://github.com/innoisys/ius.

Ischemic stroke is a time-critical medical emergency where rapid diagnosis is essential for improving patient outcomes. Non-contrast computed tomography (NCCT) serves as the frontline imaging tool, yet it often fails to reveal the subtle ischemic changes present in the early, hyperacute phase. This limitation can delay crucial interventions. To address this diagnostic challenge, we introduce a semi-supervised segmentation method using generative adversarial networks (GANs) to accurately delineate early ischemic stroke regions. The proposed method employs an adversarial framework to effectively learn from a limited number of annotated NCCT scans, while simultaneously leveraging a larger pool of unlabeled scans. By employing Dice loss, cross-entropy loss, a feature matching loss and a self-training loss, the model learns to identify and delineate early infarcts, even when they are faint or their size is small. Experiments on the publicly available Acute Ischemic Stroke Dataset (AISD) demonstrate the potential of the proposed method to enhance diagnostic capabilities, reduce the burden of manual annotation, and support more efficient clinical decision-making in stroke care.

Abstract Shadow removal is an important problem in computer vision, since the presence of shadows complicates core computer vision tasks, including image segmentation and object recognition. Most state-of-the-art shadow removal methods are based on complex deep learning architectures, which require training on a large amount of data. In this paper a novel and efficient methodology is proposed aiming to provide a simple solution to shadow removal, both in terms of implementation and computational cost. The proposed methodology is fully unsupervised, based solely on color image features. Initially, the shadow region is automatically extracted by a segmentation algorithm based on Electromagnetic-Like Optimization. Superpixel-based segmentation is performed and pairs of shadowed and non-shadowed regions, which are nearest neighbors in terms of their color content, are identified as parts of the same object. The shadowed part of each pair is relighted by means of histogram matching, using the content of its non-shadowed counterpart. Quantitative and qualitative experiments on well-recognized publicly available benchmark datasets are conducted to evaluate the performance of proposed methodology in comparison to state-of-the-art methods. The results validate both its efficiency and effectiveness, making evident that solving the shadow removal problem does not necessarily require complex deep learning-based solutions.

The physician’s experience is highly correlated with the content interpretation of medical images. Over time, physicians develop their ability to examine the images, and this is usually reflected on gaze patterns they follow to observe visual cues that lead them to diagnostic decisions. In the context of gaze prediction, graph and machine learning methods have been proposed for the visual saliency estimation on generic images. In this work we preset a novel and robust gaze estimation methodology based on physicians’ eye fixations, using convolutional neural networks (CNNs) trained according to a novel co-operative scheme, on medical images acquired during Wireless Capsule Endoscopy (WCE). The proposed training approach considers both the reconstruction accuracy of the estimated saliency maps, and their contribution to the classification process of normal and abnormal findings. The model that was trained with the proposed co-operative procedure was able to achieve an average score of 0.76 Judd’s Area Under the receiver operating Characteristic (AUC-J).

Image classification is a fundamental component of intelligent vision systems. Developing classifiers capable of explaining how or why a classification result occurs, in a way compatible with human perception, remains a challenge. Considering the increasing demand for such classifiers, this paper introduces a novel interpretable classification scheme based on a Fuzzy Cognitive Map (FCM), named xFCM. xFCM is a directed graph with nodes representing semantic concepts of the real world, as these are illustrated within different images. These concepts are considered as Semantic Granules (SGs) instantiated as clusters of images sharing common characteristics. The edges of the graph represent similarities between the SGs, linguistically expressed by fuzzy sets. Unlike current FCM-based classification approaches, xFCM embeds a mechanism for automatic determination of its structure from data. In addition, it is simple to implement, and it exploits cause- and-effect relationships between its concepts to derive a classification result that is interpretable by humans. The results of the experiments, using publicly available datasets, prove the effectiveness of the proposed framework, in comparison with other state-of-the-art classifiers.

Interpretable machine learning models have recently received a considerable attention for their capability to provide understandable explanations of predictions evoked from complex systems. In this paper a novel automatic interpretable classification scheme is introduced based on a Fuzzy Cognitive Map (FCM). The proposed approach aims to address the problem of image classification using high-level features, extracted from a Convolutional Neural Network (CNN). The proposed FCM constitutes a fuzzy-graph structure for representing causal reasoning concerning semantic concepts of the real world, as these are depicted within different images. An advantage over current classifiers is that the proposed FCM embeds a mechanism for automatic determination of its structure from the datasets used in each experiment. More importantly, it provides interpretable classification results, whereas it is simple to implement. Experimental results using publicly available datasets show that the proposed approach efficiently extends the application of the traditional FCMs into image classification problem.

Convolutional Neural Networks (CNNs) have demonstrated an outstanding performance on a range of image classification problems in various domains. However, their major drawback is that they are “black box” and opaque classifiers. Taking into consideration the increasing demand for interpretable classification models, this paper introduces a novel meta-feature extraction scheme. This scheme is based on fuzzy sets, and it can be applied on the feature maps of a CNN. Initially, representative image prototypes are selected based on their deep feature map representation. Then, it constructs information granules from the feature maps, describing the content of each image class. It uses fuzzy sets to linguistically characterize the similarity between the deep feature maps of an image and the deep feature maps of the image prototypes. Thus, a classification outcome can be interpreted based on the features characterizing the different image classes involved in a classification problem. The experimental evaluation of the proposed scheme is performed on five publicly available image datasets. The results indicate that the proposed scheme outperforms other state-of-the-art classifiers, while providing an understandable interpretation of the classification result.

Wireless capsule endoscopy (WCE) is a modality used for the non-invasive examination of the gastrointestinal (GI) tract. Physicians diagnose pathologies in images derived from Capsule Endoscopy (CE) using specific gaze patterns to observe pathologically related visual cues. Lately, deep learning has advanced in the domain of human eye-fixation estimation in natural images. However, the potentials of predicting the eye related patterns, such as eye fixations, in medical images has not been thoroughly investigated. In this work, we propose a CNN auto-encoder model, that is capable of predicting saliency maps estimating the gaze-patterns, in terms of eye-fixations, of physicians in CE images. The proposed model outperforms other approaches for visual saliency estimation based on physicians' eye fixation by providing an AUC-J of 0.726 among CE images depicting various pathological and normal cases.

Salient object detection (SOD) can directly improve the performance of tasks like obstacle detection, semantic segmentation and object recognition. Such tasks are important for robotic and other autonomous navigation systems. State-of-the-art SOD methodologies, provide improved performance by incorporating depth information, usually acquired using additional specialized sensors, e.g., RGB-D cameras. This introduces an overhead to the overall cost and flexibility of such systems. Nevertheless, the recent advances of machine learning, have provided models, capable of generating depth map approximations, given a single RGB image. In this work, we propose a novel monocular SOD (MonoSOD) methodology, based on a two-branch CNN autoencoder architecture capable of predicting depth maps and estimating saliency through a trainable refinement scheme. Its application on benchmark datasets, indicates that its performance is comparable to that of state-of-the-art SOD methods relying on RGB-D data. Therefore, it could be considered as a lower-cost alternative of such methods for future applications.

Convolutional neural networks (CNNs) are artificial learning systems typically based on two operations: convolution, which implements feature extraction through filtering, and pooling, which implements dimensionality reduction. The impact of pooling in the classification performance of the CNNs has been highlighted in several previous works, and a variety of alternative pooling operators have been proposed. However, only a few of them tackle with the uncertainty that is naturally propagated from the input layer to the feature maps of the hidden layers through convolutions. In this article we present a novel pooling operation based on (type-1) fuzzy sets to cope with the local imprecision of the feature maps, and we investigate its performance in the context of image classification. Fuzzy pooling is performed by fuzzification, aggregation, and defuzzification of feature map neighborhoods. It is used for the construction of a fuzzy pooling layer that can be applied as a drop-in replacement of the current, crisp, pooling layers of CNN architectures. Several experiments using publicly available datasets show that the proposed approach can enhance the classification performance of a CNN. A comparative evaluation shows that it outperforms state-of-the-art pooling approaches.

The generalization performance in deep learning is linked to the size and the variations of the samples available during training. This is apparent in the domain of computer-aided gastrointestinal tract abnormality detection, where the lesions can vary a lot from each other and the number of available samples is limited, mainly due to personal data protection legislations. In this work we present a novel approach of tackling the problem of limited training data availability by making use of artificially generated images. More specifically we trained a Generative Adversarial Network (GAN) using Wireless Capsule Endoscopy (WCE) images to generate fake but realistic images from the small bowel. The generated images were then used to train a Convolutional Neural Network (CNN) to identify inflammatory conditions on real WCE images. To evaluate the performance of our approach, in our experiments we compare the generalization performance of the same CNN architecture trained separately with real and fake images, obtaining 90.9% and 79.1% Area Under Receiver Operating Characteristic (AUC), respectively. The results show that training using solely artificially generated data can be effective in cases where real training data are inaccessible.

The interpretation of medical images depends on physicians' experience. Over time, physicians develop their ability to examine the images, and this is usually reflected on gaze patterns they follow to observe visual cues, which lead them to diagnostic decisions. In the context of gaze prediction, graph and machine learning methods have been proposed for the visual saliency estimation on generic images. In this work we preset a novel and robust gaze estimation methodology based on physicians' eye fixations, using convolutional neural networks combined with regularization methods, on medical images taken during Wireless Capsule Endoscopy (WCE). Furthermore, we present a novel dataset of physicians' eye fixation patterns which was used for the training of the neural network model. The model was able to achieve 68.5% Judd's Area Under the receiver operating Characteristic (AUC-J).

The fracture detection process is difficult and requires specialized knowledge of the anatomical structures of the area under consideration. X-ray imaging provides images of the body's internal structures. Despite the rapid developments of medical imaging by adding newer imaging techniques such as CT and MRI, the exam of choice to detect bone fractures faster and cheaper is x-ray imaging (radiography). The objective of this study is the automatic detection of fractures in bone x-ray images using an image classification method. The dataset that was used in this study consists of 300 x-ray bone images of upper and lower extremity. In this study, we propose a novel feature extraction and classification methodology for the detection of bone fractures, named Wavelet Fuzzy Phrases (WFP). WFP extracts textural information from different bands of the 2D Discrete Wavelet Transform (DWT) images, which is expressed by a set of words. Each word is represented by a fuzzy set. The words form phrases, obtained from the aggregation of the fuzzy sets, representing the image contents. The classification accuracy achieved for bone fracture detection is 84%, which is higher than that obtained by other, state-of-the-art bone fracture detection methods. The results of this work show that this method can be used to draw the attention of the physicians in areas of the x-rays that are suspicious for fracture; therefore, it could contribute in the reduction of diagnostic errors as well as the increase of the radiologists' productivity.

Robotic endoscopic systems offer a minimally invasive approach to the examination of internal body structures, and their application is rapidly extending to cover the increasing needs for accurate therapeutic interventions. In this context, it is essential for such systems to be able to perform measurements, such as measuring the distance traveled by a wireless capsule endoscope, so as to determine the location of a lesion in the gastrointestinal tract, or to measure the size of lesions for diagnostic purposes. In this paper, we investigate the feasibility of performing contactless measurements using a computer vision approach based on neural networks. The proposed system integrates a deep convolutional image registration approach and a multilayer feed-forward neural network into a novel architecture. The main advantage of this system, with respect to the state-of-the-art ones, is that it is more generic in the sense that it is 1) unconstrained by specific models, 2) more robust to nonrigid deformations, and 3) adaptable to most of the endoscopic systems and environment, while enabling measurements of enhanced accuracy. The performance of this system is evaluated under ex vivo conditions using a phantom experimental model and a robotically assisted test bench. The results obtained promise a wider applicability and impact in endoscopy in the era of big data.

The detection of abnormalities in endoscopic video frames can contribute in the early and more accurate detection of pathologic conditions. In this paper we present a novel Convolutional Neural Network (CNN) architecture for automatic detection of abnormal images in endoscopic video sequences. It features multiscale representation of the endoscopic images in its structure, and peephole connections contributing in enhanced generalization with less computational requirements. An important aspect of the proposed architecture is that it enables weakly-supervised learning, using only semantically annotated images. A novel cross-dataset experimental study is performed to investigate its generalization performance on various publicly available datasets. The results validate that the proposed architecture outperforms recent approaches, with results reaching up to 90.66% in terms of the area under the receiver operating characteristic.

This paper proposes a novel methodology for automatic detection and localization of gastrointestinal (GI) anomalies in endoscopic video frame sequences. Training is performed with weakly annotated images, using only image-level, semantic labels instead of detailed, and pixel-level annotations. This makes it a cost-effective approach for the analysis of large videoendoscopy repositories. Other advantages of the proposed methodology include its capability to suggest possible locations of GI anomalies within the video frames, and its generality, in the sense that abnormal frame detection is based on automatically derived image features. It is implemented in three phases: 1) it classifies the video frames into abnormal or normal using a weakly supervised convolutional neural network (WCNN) architecture; 2) detects salient points from deeper WCNN layers, using a deep saliency detection algorithm; and 3) localizes GI anomalies using an iterative cluster unification (ICU) algorithm. ICU is based on a pointwise cross-feature-map (PCFM) descriptor extracted locally from the detected salient points using information derived from the WCNN. Results, from extensive experimentation using publicly available collections of gastrointestinal endoscopy video frames, are presented. The data sets used include a variety of GI anomalies. Both anomaly detection and localization performance achieved, in terms of the area under receiver operating characteristic (AUC), were >80%. The highest AUC for anomaly detection was obtained on conventional gastroscopy images, reaching 96%, and the highest AUC for anomaly localization was obtained on wireless capsule endoscopy images, reaching 88%.

Various modalities are used for the examination of the gastrointestinal (GI) tract. One such modality is Wireless Capsule Endoscopy (WCE), a noninvasive technique which consists of a swallowable color camera that enables the detection of GI pathology with only minimal patient discomfort. Currently, tracking of the capsule position is estimated in the 3D abdominal space, using radio-frequency (RF) triangulation. The RF triangulation technique, however, does not provide sufficient information about the location of the capsule along the GI lumen, and consequently, the localization of any possible abnormality. Recently, we proposed a geometric visual odometry (VO) method for the localization of the capsule in the GI lumen. In this paper, we extend this state-of-art method by exploiting an artificial neural network (ANN) to augment the geometric method and achieve higher localization accuracy. The results of this novel approach are validated with an in-vitro experiment that provides ground truth information about the location of the capsule. The mean absolute error obtained, for a distance of 19.6cm, is 0.79±0.51cm.

Wireless capsule endoscopy (WCE) is performed with a swallowable miniature optical endoscope which transmits color images wirelessly during its journey in the gastrointestinal tract. In this paper we present a computationally efficient and effective approach to cope with automatic detection of possible abnormalities in the WCE videos and consequently with the reduction of the time required for the WCE inspection. It involves automatic detection of salient points based on color information and supervised classification of simple color vectors extracted from the neighborhood of each point. The experiments performed aim to determine the optimal color space components for feature extraction, and identification of abnormalities. Main advantages of this approach are its computational efficiency, its sensitivity to detect small lesions, and its generality. The results obtained from experimentation with a dataset with various types of abnormalities and non-ideal normal frames, approximate 0.9 in terms of the area under receiver operating characteristic (ROC).

Wireless capsule endoscopy (WCE) is performed by a swallowable pill capsule equipped with a camera wirelessly transmitting color video frames to an external receiver. The resulting video consists usually of several thousands of frames and its visual examination requires hours of endoscopists' undivided attention. In this paper we propose a novel visualization approach for WCE which enables faster examination of the endoscopic video, while providing a broader field of view. This is achieved by an algorithm that iteratively samples clusters of consecutive frames from the original video. The frames of each cluster are geometrically transformed, so as to generate a seamless collage subsequently projected into a new frame without any information loss. The new frames compose a new WCE video with a smaller number of frames. The video frame collage is based on homography matrix estimation from frame correspondences. The experiments show that the length of the WCE video, and therefore the reading times required can be significantly reduced.

Computational analysis of wireless capsule endoscopy (WCE) videos has already proved its potentials in the discovery or characterization of lesions and in the reduction of the time required by the endoscopists to perform the examination. An open problem that has only partially been addressed is the localization of the capsule endoscope in the gastrointestinal (GI) tract. Previous works have been based mainly on external, wearable, sensors. In this paper we propose a novel approach based solely on visual information extracted from WCE videos. This approach is based on a feature tracking method for visual odometry, which enables the estimation of both the rotation and the displacement of a capsule endoscope from reference anatomical landmarks. Its implementation is based on a novel, open access Java Video Analysis (JVA) framework, which enables quick and standardized development of intelligent video analysis applications. The experimental evaluation presented in this paper, indicates the feasibility of the proposed methodological approach and the efficiency of its implementation.

Uncertainty and imprecision characterize human cognitive and reasoning processes. Fuzzy cognitive maps (FCMs) are computationally simple yet effective structures to approximately model and simulate such processes. A limitation of current FCMs is that they are unable to model the hesitancy introduced into a complex system due to imperfect facts, missing information, and indecision. To cope with this issue, we propose a novel extension of the FCM model which is based on the theory of intuitionistic fuzzy sets. This intuitionistic FCM (iFCM) model, which is denoted as iFCM-II, inherently exploits the mathematical framework of intuitionistic fuzzy sets for the definition of the concepts constituting the cognitive map and their interrelations, as well as for reasoning. Furthermore, unlike the previous iFCM model, which is denoted as iFCM-I, it enables an intuitionistic estimation of hesitancy at the output concepts, thus offering a natural mechanism to assess the quality of its output. The advantages of the proposed iFCM model over the current FCM and iFCM models are demonstrated with reproducible numeric examples for process control and decision support applications.

Fuzzy cognitive maps have proven an exceptional means to reasoning for decision support. In this paper we propose a novel approach to hesitancy-aware reasoning based on cognitive maps and intuitionistic fuzzy logic. Intuitionistic fuzzy sets are considered for the linguistic representation of both the concepts of a cognitive map and the relations defined between them. A comparative advantage of this approach over the state of the art is an intrinsic mechanism of modeling human hesitancy, as introduced in the construction of a cognitive model, and propagated through the reasoning process to a final decision. A numerical example demonstrates its effectiveness which can extend to a variety of real-world applications.

Medical decision making can be regarded as a process, combining both analytical cognition and intuition. It involves reasoning within complex causal models of multiple concepts, usually described by uncertain, imprecise, and/or incomplete information. Aiming to model medical decision making, we propose a novel approach based on cognitive maps and intuitionistic fuzzy logic. The new model, called intuitionistic fuzzy cognitive map (iFCM), extends the existing fuzzy cognitive map (FCM) by considering the expert's hesitancy in the determination of the causal relations between the concepts of a domain. Furthermore, a modification in the formulation of the new model makes it even less sensitive than the original model to missing input data. To validate its effectiveness, an iFCM with 34 concepts representing fuzzy, linguistically expressed patient-specific data, symptoms, and multimodal measurements was constructed for pneumonia severity assessment. The results obtained reveal its comparative advantage over the respective FCM model by providing decisions that match better with the ones made by the experts. The generality of the proposed approach suggests its suitability for a variety of medical decision-making tasks.

BACKGROUND: Complementary DNA (cDNA) microarrays are a well established technology for studying gene expression. A microarray image is obtained by laser scanning a hybridized cDNA microarray, which consists of thousands of spots representing chains of cDNA sequences, arranged in a two-dimensional array. The separation of the spots into distinct cells is widely known as microarray image gridding. METHODS: In this paper we propose M3G, a novel method for automatic gridding of cDNA microarray images based on the maximization of the margin between the rows and the columns of the spots. Initially the microarray image rotation is estimated and then a pre-processing algorithm is applied for a rough spot detection. In order to diminish the effect of artefacts, only a subset of the detected spots is selected by matching the distribution of the spot sizes to the normal distribution. Then, a set of grid lines is placed on the image in order to separate each pair of consecutive rows and columns of the selected spots. The optimal positioning of the lines is determined by maximizing the margin between these rows and columns by using a maximum margin linear classifier, effectively facilitating the localization of the spots. RESULTS: The experimental evaluation was based on a reference set of microarray images containing more than two million spots in total. The results show that M3G outperforms state of the art methods, demonstrating robustness in the presence of noise and artefacts. More than 98% of the spots reside completely inside their respective grid cells, whereas the mean distance between the spot center and the grid cell center is 1.2 pixels. CONCLUSIONS: The proposed method performs highly accurate gridding in the presence of noise and artefacts, while taking into account the input image rotation. Thus, it provides the potential of achieving perfect gridding for the vast majority of the spots.

In this paper, we propose a novel scheme for efficient content-based medical image retrieval, formalized according to the PAtterns for Next generation DAtabase systems (PANDA) framework for pattern representation and management. The proposed scheme involves block-based low-level feature extraction from images followed by the clustering of the feature space to form higher-level, semantically meaningful patterns. The clustering of the feature space is realized by an expectation-maximization algorithm that uses an iterative approach to automatically determine the number of clusters. Then, the 2-component property of PANDA is exploited: the similarity between two clusters is estimated as a function of the similarity of both their structures and the measure components. Experiments were performed on a large set of reference radiographic images, using different kinds of features to encode the low-level image content. Through this experimentation, it is shown that the proposed scheme can be efficiently and effectively applied for medical image retrieval from large databases, providing unsupervised semantic interpretation of the results, which can be further extended by knowledge representation methodologies.

Thyroid nodules are solid or cystic lumps formed in the thyroid gland and may be caused by a variety of thyroid disorders. This paper presents a novel active contour model for precise delineation of thyroid nodules of various shapes according to their echogenicity and texture, as displayed in ultrasound (US) images. The proposed model, named joint echogenicity-texture (JET), is based on a modified Mumford-Shah functional that, in addition to regional image intensity, incorporates statistical texture information encoded by feature distributions. The distributions are aggregated within the functional through new log-likelihood goodness-of-fit terms. The JET model requires only a rough region of interest within the thyroid gland as input and automatically proceeds with precise delineation of the nodules, revealing their shape and size. The performance of the JET model was validated on a range of US images displaying hypoechoic and isoechoic nodules of various shapes. The quantification of the results shows that the JET model: 1) provides precise delineations of thyroid nodules as compared to "ground truth" delineations obtained by experts and 2) copes with the limitations of the previous thyroid US delineation approaches as it is capable of delineating thyroid nodules regardless of their echogenicity or shape.

This paper presents a novel method for DNA microarray gridding based on support vector machine (SVM) classifiers. It employs a set of soft-margin SVMs to estimate the lines of the DNA microarray grid by maximizing the margin between the lines and the spots. This process comprises an efficient and effective approach of separating the spots into distinct rows and columns. The classifiers are trained using the spot locations as training vectors. The results obtained from the application of the proposed method on reference microarray images illustrate its robustness in the presence of artifacts, noise and weakly expressed spots. The comparative evaluation presented reveals its advantageous performance over a state of the art gridding approach. The gridding quality achieved exceeds 95% in terms of the total number of perfectly gridded spots.

The most common radiographic manifestation of bacterial pulmonary infections are foci of consolidation which are visible as bright shadows interfering with the interior lung intensity. In critically-ill patients this interference can be severe leading to vague or invisible lung field boundaries which are difficultly distinguished even by experienced physicians. This problem is amplified if the radiographs are of low quality as obtained with a portable x-ray device, routinely used in intensive care units. This paper proposes a pioneering methodology that copes with lung field detection in both stationary and portable chest radiographs by combining statistical grey-level intensity information and directional edge maps. The boundaries of the lung fields are approximated by consecutive intuitively manipulated parametric curves. Conventional and state of the art lung field detection approaches address only stationary radiography, and only a few of them cope with pulmonary infections. The proposed methodology features unsupervised operation, it is not iterative, it is not limited by the patientspsila positioning, and it is tolerant to the presence of consolidations and boundary discontinuities of the lung fields. Its performance is validated on various stationary radiographs and on a set of portable radiographs obtained from patients with bacterial pulmonary infections.

Wireless Capsule Endoscopy (WCE) is a non-invasive colour imaging technique that has been introduced for the screening of the gastrointestinal tract and especially the small intestine. WCE is performed by a wireless swallowable endoscopic capsule that transmits more than 50,000 video frames per examination. The visual inspection of the resulting video is a highly time-consuming task even for the experienced gastroenterologist. In this paper we propose a novel WCE video summarization approach which is subsequently evaluated using real world patient data. The proposed approach aims to the reduction of the number of the video frames to be visually inspected so as to enable significant reduction in the video assessment time. It is based on clustering using symmetric non-negative matrix factorization initialized by the fuzzy c-means algorithm and supported by non-negative Lagrangian relaxation to extract a subset of video scenes containing the most representative frames from an entire examination. Real world patient data that display abnormal findings at several sites in the small intestine were annotated by expert gastroenterologists in order to experimentally evaluate the proposed approach. The results demonstrate that the suggested approach leads to significant reduction of the total number of frames in the input video without losing critical information related to the abnormal regions of the small intestine.

The robustness of textural features on speckle noise is of vital importance for ultrasound imaging. A set of novel fuzzy features for thyroid ultrasound texture representation, demonstrating noise-resistant properties, is presented, analyzed and evaluated in this study. The textural feature extraction scheme is based on the fuzzyfication of the local binary pattern approach. The proposed features are evaluated on an annotated dataset of B-mode thyroid ultrasound images acquired from 75 patients. The experimental results illustrate that these features provide accurate representation of the thyroid texture. They can be effectively utilized for thyroid nodule detection outperforming other thyroid texture representation approaches that have been recently proposed in the literature.

This paper investigates a novel computational approach to thyroid tissue characterization in ultrasound images. It is based on the hypothesis that tissues in thyroid ultrasound images may be differentiated by directionality patterns. These patterns may not be always distinguishable by the human eye because of the dominant image noise. The encoding of the directional patterns in the thyroid ultrasound images is realized by means of radon transform features. A representative set of ultrasound images, acquired from 66 patients was constructed to perform experiments that test the validity of the initial hypothesis. Supervised classification experiments showed that the proposed approach is capable of discriminating normal and nodular thyroid tissues, whereas nodular tissues can be further characterized as of high or low malignancy risk.

This paper presents a computer-aided approach for nodule delineation in thyroid ultrasound (US) images. The developed algorithm is based on a novel active contour model, named variable background active contour (VBAC), and incorporates the advantages of the level set region-based active contour without edges (ACWE) model, offering noise robustness and the ability to delineate multiple nodules. Unlike the classic active contour models that are sensitive in the presence of intensity inhomogeneities, the proposed VBAC model considers information of variable background regions. VBAC has been evaluated on synthetic images, as well as on real thyroid US images. From the quantification of the results, two major impacts have been derived: 1) higher average accuracy in the delineation of hypoechoic thyroid nodules, which exceeds 91%; and 2) faster convergence when compared with the ACWE model.

In this paper, we extend the application of four texture feature extraction methods proposed for the detection of colorectal lesions, into the discrimination of gastric polyps in endoscopic video. Support Vector Machines have been utilized for the texture classification task. The polyp discrimination performance of the surveyed schemes is compared by means of Receiver Operating Characteristics (ROC). The results advocate the feasibility of a computer-based system for polyp detection in video gastroscopy that exploits the textural characteristics of the gastric mucosa in conjunction with its color appearance.

Nodular thyroid disease is a frequent occurrence in clinical practice and it is associated with increased risk of thyroid cancer and hyperfunction. In this paper we propose a novel method for computer-aided detection of thyroid nodules in ultrasound (US) images. The proposed method is based on a level-set image segmentation approach that takes into account the inhomogeneity of the US images. This novel method was experimentally evaluated using US images acquired from 35 patients. The results show that the proposed method achieves more accurate delineation of the thyroid nodules in the US images and faster convergence than other relevant methods.

A novel active contour model named variable background active contour model is proposed and applied for the detection of thyroid nodules in ultrasound images. The new model offers edge independency, no need for smoothing, ability for topological changes and it is more accurate when compared to the active contour without edges model. Improved accuracy is achieved by introducing as background a limited image subset which appropriately changes shape to reduce the effects of background inhomogeneity. We validated the proposed model on ultrasound images acquired from 24 patients and the results demonstrate an improvement in accuracy when compared to the active contour without edges model.

We present an approach to the detection of tumors in colonoscopic video. It is based on a new color feature extraction scheme to represent the different regions in the frame sequence. This scheme is built on the wavelet decomposition. The features named as color wavelet covariance (CWC) are based on the covariances of second-order textural measures and an optimum subset of them is proposed after the application of a selection algorithm. The proposed approach is supported by a linear discriminant analysis (LDA) procedure for the characterization of the image regions along the video frames. The whole methodology has been applied on real data sets of color colonoscopic videos. The performance in the detection of abnormal colonic regions corresponding to adenomatous polyps has been estimated high, reaching 97% specificity and 90% sensitivity.

Video processing for classification applications in medical imaging is an area with great importance. In this paper a framework for classification of suspicious lesions using the video produced during an endoscopic session is presented. The proposed approach is based on a feature extraction scheme that uses second order statistical information of the wavelet transformation. These features are used as input to a multilayer feedforward neural network (MFNN) architecture, which has been trained using features of normal and tumor regions. The system uses a limited number of frames with a rather small population of training vectors. The classification results are promising, since the system has been proven to be capable to classify and locate regions, that correspond to lesions with a success of 94 up to 99%, in a sequence of the video-frames. The proposed methodology can be used as a valuable diagnostic tool that may assist physicians to identify possible tumor regions or malignant formations.

A few approaches have been presented in the literature towards the discrimination of texture in medical images. Medical experts proposed that the more valuable information for discriminating among normal and suspicious cancer regions in endoscopic images is the texture of the examined tissue. Texture can be encoded by a number of mathematical descriptors. Three well-known textural descriptors, as well as a new wavelet-based one are used in this paper for an accurate study and evaluation of the methodologies encountered. Experiments conducted include tests with various images from the Brodatz album, as well as interpretation of tissue regions in endoscopic image. In all cases the recognition task is supported by multilayer perceptron type neural network architectures.