Automated plaque classification using computed tomography angiography and Gabor transformations

Cardiovascular diseases are the primary cause of death globally. These are often associated with atherosclerosis. This inflammation process triggers important variations in the coronary arteries (CA) and can lead to coronary artery disease (CAD). The presence of CA calcification (CAC) has recently been shown to be a strong predictor of CAD. In this clinical setting, computed tomography angiography (CTA) has begun to play a crucial role as a non-intrusive imaging method to characterize and study CA plaques. Herein, we describe an automated algorithm to classify plaque as either normal, calcified, or non-calcified using 2646 CTA images acquired from 73 patients. The automated technique is based on various features that are extracted from the Gabor transform of the acquired CTA images. Specifically, seven features are extracted from the Gabor coefficients : energy, and Kapur, Max, Rényi, Shannon, Vajda, and Yager entropies. The features were then ordered based on the F-value and input to numerous classification methods to achieve the best classification accuracy with the least number of features. Moreover, two well-known feature reduction techniques were employed, and the features acquired were also ranked according to F-value and input to several classifiers. The best classification results were obtained using all computed features without the employment of feature reduction, using a probabilistic neural network. An accuracy, positive predictive value, sensitivity, and specificity of 89.09%, 91.70%, 91.83% and 83.70% was obtained, respectively. Based on these results, it is evident that the technique can be helpful in the automated classification of plaques present in CTA images, and may become an important tool to reduce procedural costs and patient radiation dose. This could also aid clinicians in plaque diagnostics.

Addressing challenges of quantitative methodologies and event interpretation in the study of atrial fibrillation

Atrial fibrillation (AF) is the commonest arrhythmia, yet the mechanisms of its onset and persistence are incompletely known. Although techniques for quantitative assessment have been investigated, there have been few attempts to integrate this information to advance disease treatment protocols. In this review, key quantitative methods for AF analysis are described, and suggestions are provided for the coordination of the available information, and to develop foci and directions for future research efforts. Quantitative biologists may have an interest in this topic in order to develop machine learning and tools for arrhythmia characterization, but they may perhaps have a minimal background in the clinical methodology and in the types of observed events and mechanistic hypotheses that have thus far been developed. We attempt to address these issues via exploration of the published literature. Although no new data is presented in this review, examples are shown of current lines of investigation, and in particular, how electrogram analysis and whole-chamber quantitative modeling of the left atrium may be useful to characterize fibrillatory patterns of activity, so as to propose avenues for more efficacious acquisition and interpretation of AF data.

Automated detection of schizophrenia using nonlinear signal processing methods

Examination of the brain's condition with the Electroencephalogram (EEG) can be helpful to predict abnormality and cerebral activities. The purpose of this study was to develop an Automated Diagnostic Tool (ADT) to investigate and classify the EEG signal patterns into normal and schizophrenia classes. The ADT implements a sequence of events, such as EEG series splitting, non-linear features mining, t-test assisted feature selection, classification and validation. The proposed ADT is employed to evaluate a 19-channel EEG signal collected from normal and schizophrenia class volunteers. A dataset was created by splitting the raw 19-channel EEG into a sequence of 6250 sample points, which was helpful to produce 1142 features of normal and schizophrenia class patterns. Non-linear feature extraction was then implemented to mine 157 features from each EEG pattern, from which 14 of the principal features were identified based on significance. Finally, a signal classification practice with Decision-Tree (DT), Linear-Discriminant analysis (LD), k-Nearest-Neighbour (KNN), Probabilistic-Neural-Network (PNN), and Support-Vector-Machine (SVM) with various kernels was implemented. The experimental outcome showed that the SVM with Radial-Basis-Function (SVM-RBF) offered a superior average performance value of 92.91% on the considered EEG dataset, as compared to other classifiers implemented in this work.

Application of nonlinear methods to discriminate fractionated electrograms in paroxysmal versus persistent atrial fibrillation

BACKGROUND AND OBJECTIVE

Complex fractionated atrial electrograms (CFAE) may contain information concerning the electrophysiological substrate of atrial fibrillation (AF); therefore they are of interest to guide catheter ablation treatment of AF. Electrogram signals are shaped by activation events, which are dynamical in nature. This makes it difficult to establish those signal properties that can provide insight into the ablation site location. Nonlinear measures may improve information. To test this hypothesis, we used nonlinear measures to analyze CFAE.

METHODS

CFAE from several atrial sites, recorded for a duration of 16 s, were acquired from 10 patients with persistent and 9 patients with paroxysmal AF. These signals were appraised using non-overlapping windows of 1-, 2- and 4-s durations. The resulting data sets were analyzed with Recurrence Plots (RP) and Recurrence Quantification Analysis (RQA). The data was also quantified via entropy measures.

RESULTS

RQA exhibited unique plots for persistent versus paroxysmal AF. Similar patterns were observed to be repeated throughout the RPs. Trends were consistent for signal segments of 1 and 2 s as well as 4 s in duration. This was suggestive that the underlying signal generation process is also repetitive, and that repetitiveness can be detected even in 1-s sequences. The results also showed that most entropy metrics exhibited higher measurement values (closer to equilibrium) for persistent AF data. It was also found that Determinism (DET), Trapping Time (TT), and Modified Multiscale Entropy (MMSE), extracted from signals that were acquired from locations at the posterior atrial free wall, are highly discriminative of persistent versus paroxysmal AF data.

CONCLUSIONS

Short data sequences are sufficient to provide information to discern persistent versus paroxysmal AF data with a significant difference, and can be useful to detect repeating patterns of atrial activation.

A review of automated sleep stage scoring based on physiological signals for the new millennia

BACKGROUND AND OBJECTIVE

Sleep is an important part of our life. That importance is highlighted by the multitude of health problems which result from sleep disorders. Detecting these sleep disorders requires an accurate interpretation of physiological signals. Prerequisite for this interpretation is an understanding of the way in which sleep stage changes manifest themselves in the signal waveform. With that understanding it is possible to build automated sleep stage scoring systems. Apart from their practical relevance for automating sleep disorder diagnosis, these systems provide a good indication of the amount of sleep stage related information communicated by a specific physiological signal.

METHODS

This article provides a comprehensive review of automated sleep stage scoring systems, which were created since the year 2000. The systems were developed for Electrocardiogram (ECG), Electroencephalogram (EEG), Electrooculogram (EOG), and a combination of signals.

RESULTS

Our review shows that all of these signals contain information for sleep stage scoring.

CONCLUSIONS

The result is important, because it allows us to shift our research focus away from information extraction methods to systemic improvements, such as patient comfort, redundancy, safety and cost.

A new approach for arrhythmia classification using deep coded features and LSTM networks

BACKGROUND AND OBJECTIVE

For diagnosis of arrhythmic heart problems, electrocardiogram (ECG) signals should be recorded and monitored. The long-term signal records obtained are analyzed by expert cardiologists. Devices such as the Holter monitor have limited hardware capabilities. For improved diagnostic capacity, it would be helpful to detect arrhythmic signals automatically. In this study, a novel approach is presented as a candidate solution for these issues.

METHODS

A convolutional auto-encoder (CAE) based nonlinear compression structure is implemented to reduce the signal size of arrhythmic beats. Long-short term memory (LSTM) classifiers are employed to automatically recognize arrhythmias using ECG features, which are deeply coded with the CAE network.

RESULTS

Based upon the coded ECG signals, both storage requirement and classification time were considerably reduced. In experimental studies conducted with the MIT-BIH arrhythmia database, ECG signals were compressed by an average 0.70% percentage root mean square difference (PRD) rate, and an accuracy of over 99.0% was observed.

CONCLUSIONS

One of the significant contributions of this study is that the proposed approach can significantly reduce time duration when using LSTM networks for data analysis. Thus, a novel and effective approach was proposed for both ECG signal compression, and their high-performance automatic recognition, with very low computational cost.

Computer-aided diagnosis of congestive heart failure using ECG signals - A review

The heart muscle pumps blood to vital organs, which is indispensable for human life. Congestive heart failure (CHF) is characterized by the inability of the heart to pump blood adequately throughout the body without an increase in intracardiac pressure. The symptoms include lung and peripheral congestion, leading to breathing difficulty and swollen limbs, dizziness from reduced delivery of blood to the brain, as well as arrhythmia. Coronary artery disease, myocardial infarction, and medical co-morbidities such as kidney disease, diabetes, and high blood pressure all take a toll on the heart and can impair myocardial function. CHF prevalence is growing worldwide. It afflicts millions of people globally, and is a leading cause of death. Hence, proper diagnosis, monitoring and management are imperative. The importance of an objective CHF diagnostic tool cannot be overemphasized. Standard diagnostic tests for CHF include chest X-ray, magnetic resonance imaging (MRI), nuclear imaging, echocardiography, and invasive angiography. However, these methods are costly, time-consuming, and they can be operator-dependent. Electrocardiography (ECG) is inexpensive and widely accessible, but ECG changes are typically not specific for CHF diagnosis. A properly designed computer-aided detection (CAD) system for CHF, based on the ECG, would potentially reduce subjectivity and provide quantitative assessment for informed decision-making. Herein, we review existing CAD for automatic CHF diagnosis, and highlight the development of an ECG-based CAD diagnostic system that employs deep learning algorithms to automatically detect CHF.

Automated diagnosis of celiac disease by video capsule endoscopy using DAISY Descriptors

Celiac disease is a genetically determined disorder of the small intestine, occurring due to an immune response to ingested gluten-containing food. The resulting damage to the small intestinal mucosa hampers nutrient absorption, and is characterized by diarrhea, abdominal pain, and a variety of extra-intestinal manifestations. Invasive and costly methods such as endoscopic biopsy are currently used to diagnose celiac disease. Detection of the disease by histopathologic analysis of biopsies can be challenging due to suboptimal sampling. Video capsule images were obtained from celiac patients and controls for comparison and classification. This study exploits the use of DAISY descriptors to project two-dimensional images onto one-dimensional vectors. Shannon entropy is then used to extract features, after which a particle swarm optimization algorithm coupled with normalization is employed to select the 30 best features for classification. Statistical measures of this paradigm were tabulated. The accuracy, positive predictive value, sensitivity and specificity obtained in distinguishing celiac versus control video capsule images were 89.82%, 89.17%, 94.35% and 83.20% respectively, using the 10-fold cross-validation technique. When employing manual methods rather than the automated means described in this study, technical limitations and inconclusive results may hamper diagnosis. Our findings suggest that the computer-aided detection system presented herein can render diagnostic information, and thus may provide clinicians with an important tool to validate a diagnosis of celiac disease.

Heterogeneity of the action potential duration is required for sustained atrial fibrillation

Atrial fibrillation (AF) is the most common cardiac arrhythmia and accounts for substantial morbidity and mortality. Recently, we created a mouse model with spontaneous and sustained AF caused by a mutation in the NaV1.5 channel (F1759A) that enhances persistent Na+ current, thereby enabling the investigation of molecular mechanisms that cause AF and the identification of novel treatment strategies. The mice have regional heterogeneity of action potential duration of the atria similar to observations in patients with AF. In these mice, we found that the initiation and persistence of the rotational reentrant AF arrhythmias, known as spiral waves or rotors, were dependent upon action potential duration heterogeneity. The centers of the rotors were localized to regions of greatest heterogeneity of the action potential duration. Pharmacologically attenuating the action potential duration heterogeneity reduced both spontaneous and pacing-induced AF. Computer-based simulations also demonstrated that the action potential duration heterogeneity is sufficient to generate rotors that manifest as AF. Taken together, these findings suggest that action potential duration heterogeneity in mice and humans is one mechanism by which AF is initiated and that reducing action potential duration heterogeneity can lessen the burden of AF.

Persistent Economic Burden of the Gluten Free Diet

Gluten free (GF) products have been reported to be more expensive and less available than their gluten containing counterparts. We examined the current U.S. cost and availability of GF products and made comparisons to the marketplace over a decade ago. Cost, determined by price per ounce and availability of a “market basket” of regular and GF products across four venues and five geographic regions was compared using a student’s t test. GF products were more expensive (overall 183%), and in all regions and venues (p < 0.001). GF products from mass-market producers were 139% more expensive than the wheat-based version of the same product. Availability of GF products was greatest (66%) in the health food and upscale venues. In contrast to the results of the 2006 study, the cost of GF products has declined from 240% to 183% (adjusted for inflation). The introduction of mass-market production of GF products may have influenced the increase in availability and overall reduction of cost since 2006. The extent to which the cost of GF products impacts dietary adherence and quality of life for those on a GFD warrants exploration.

Automated detection and classification of liver fibrosis stages using contourlet transform and nonlinear features

BACKGROUND AND OBJECTIVE

Liver fibrosis is a type of chronic liver injury that is characterized by an excessive deposition of extracellular matrix protein. Early detection of liver fibrosis may prevent further growth toward liver cirrhosis and hepatocellular carcinoma. In the past, the only method to assess liver fibrosis was through biopsy, but this examination is invasive, expensive, prone to sampling errors, and may cause complications such as bleeding. Ultrasound-based elastography is a promising tool to measure tissue elasticity in real time; however, this technology requires an upgrade of the ultrasound system and software. In this study, a novel computer-aided diagnosis tool is proposed to automatically detect and classify the various stages of liver fibrosis based upon conventional B-mode ultrasound images.

METHODS

The proposed method uses a 2D contourlet transform and a set of texture features that are efficiently extracted from the transformed image. Then, the combination of a kernel discriminant analysis (KDA)-based feature reduction technique and analysis of variance (ANOVA)-based feature ranking technique was used, and the images were then classified into various stages of liver fibrosis.

RESULTS

Our 2D contourlet transform and texture feature analysis approach achieved a 91.46% accuracy using only four features input to the probabilistic neural network classifier, to classify the five stages of liver fibrosis. It also achieved a 92.16% sensitivity and 88.92% specificity for the same model. The evaluation was done on a database of 762 ultrasound images belonging to five different stages of liver fibrosis.

CONCLUSIONS

The findings suggest that the proposed method can be useful to automatically detect and classify liver fibrosis, which would greatly assist clinicians in making an accurate diagnosis.

Use of an automaton model to suggest methods for cessation of intractable fibrillatory activity

BACKGROUND

Atrial fibrillation (AF) is the most common heart arrhythmia, and permanent AF is an intractable medical problem. If cessation of permanent AF were possible, via extensive substrate ablation or multisite stimulation, it could significantly improve the public health.

METHOD

A cellular automaton composed of 576 × 576 computerized grid nodes, described in detail previously, was used to test hypotheses concerning the cessation of fibrillatory electrical activity. A refractory period gradient across the grid, and addition of randomly located nonconducting fibers, were utilized as conditions leading to fibrillatory activity. A premature S1-S2 stimulus was applied to one grid corner, resulting in unidirectional conduction block at some locations, followed by rotational activity and random propagation of activation wavelets throughout the grid, none of which terminated spontaneously. Simulated ablation lesions of dimension 20 × 20 grid nodes, imparted at core locations of rotational activity, and multisite electrode stimulation (MES) applied at nodes where recovery of excitability had occurred, were used in attempts to terminate fibrillatory activity. Six impositions of random fiber location were utilized in separate trials.

RESULTS

Simulated ablation lesions eliminated the targeted swirling vortices; however, additional vortices then often appeared at other locations. After ablating approximately one third of the grid area, localized vortices were eliminated, but individual wavelets continued to propagate about longer viable pathways forming at ablation lesions. Thus extensive ablation was unsuccessful in terminating arrhythmia. However, MES applied uniformly throughout the grid, with a coupling interval slightly longer than the maximum refractory period, terminated fibrillatory activity in some trials. More efficaciously, application of MES with a coupling interval half the maximum refractory period of the grid succeeded in capture of activation at all nodes, and when followed by a doubling of the MES coupling interval, resulted in cessation of all fibrillatory activity.

CONCLUSIONS

It is possible to terminate simulated fibrillatory activity in a computerized grid that would otherwise be intractable, using multisite stimulation with a coupling interval related to the maximum refractory period of the substrate. If each MES stimulating electrode could be individually controlled, it would be possible to apply a stimulation pattern mimicking the normal heart activation sequence.

Computer-aided diagnosis of glaucoma using fundus images: A review

BACKGROUND AND OBJECTIVES

Glaucoma is an eye condition which leads to permanent blindness when the disease progresses to an advanced stage. It occurs due to inappropriate intraocular pressure within the eye, resulting in damage to the optic nerve. Glaucoma does not exhibit any symptoms in its nascent stage and thus, it is important to diagnose early to prevent blindness. Fundus photography is widely used by ophthalmologists to assist in diagnosis of glaucoma and is cost-effective.

METHODS

The morphological features of the disc that is characteristic of glaucoma are clearly seen in the fundus images. However, manual inspection of the acquired fundus images may be prone to inter-observer variation. Therefore, a computer-aided detection (CAD) system is proposed to make an accurate, reliable and fast diagnosis of glaucoma based on the optic nerve features of fundus imaging. In this paper, we reviewed existing techniques to automatically diagnose glaucoma.

RESULTS

The use of CAD is very effective in the diagnosis of glaucoma and can assist the clinicians to alleviate their workload significantly. We have also discussed the advantages of employing state-of-art techniques, including deep learning (DL), when developing the automated system. The DL methods are effective in glaucoma diagnosis.

CONCLUSIONS

Novel DL algorithms with big data availability are required to develop a reliable CAD system. Such techniques can be employed to diagnose other eye diseases accurately.

Effects of refractory gradients and ablation on fibrillatory activity

BACKGROUND

The mechanisms involved in onset, maintenance, and termination of atrial fibrillation are not well understood. A biophysical model could be useful to determine how the events unfold.

METHOD

A two-dimensional cellular automaton consisting of 576 × 576 grid nodes was implemented to demonstrate the types of electrical activity that may occur in compromised atrial substrate. Electrical activation between nodes was made anisotropic (2:1), and the refractory period (RP) was adjusted from 74 to 192 ms in the spatial domain. Presence of collagen fibers were simulated as short lines of conduction block at many random grid sites, while ablation lesions were delineated as longer lines of block. An S1-S2 pulse from one grid corner was utilized to initiate simulated electrical activity. Simulations were done in which 1. no ablation lines, 2. random ablation lines, and 3. parallel ablation lines were added to the grid to determine how this affected the formation and annihilation of rotational activity after S1-S2 stimulation.

RESULTS

As the premature (S2) wavefront traversed the grid, rotational activity formed near boundaries where wavefronts propagated from shorter to longer refractory regions, causing unidirectional block, and were anchored by fiber clusters. Multiple wavelets appeared when wavefronts originating from different driving rotational features collided, and/or by their encounter with RP discontinuities. With the addition of randomly orientated simulated ablation lesions, followed by reinduction of fibrillatory activity, mean activation interval (AI) prolonged from a baseline level of 144.2 ms-160.3 ms (p < 0.001 in most comparisons). During fibrillatory activity, when parallel ablation lines were added to short RP regions, AI prolonged to 150.4 ms (p < 0.001), and when added to long RP regions, AI prolonged to 185.3 ms (p < 0.001). In all cases, AI prolongation after simulated ablation resulted from reduced number and/or from the isolation of local drivers, so that distant drivers in short RP regions activated long RP regions N:1, while distant drivers in long RP regions activated short RP regions at a relatively slow rate.

CONCLUSIONS

An automaton model was found useful to generate and test hypotheses concerning fibrillatory activity, which can then be validated in the clinical electrophysiology laboratory.

Source-Sink Mismatch Causing Functional Conduction Block in Re-Entrant Ventricular Tachycardia

Ventricular tachycardia (VT) caused by a re-entrant circuit is a life-threatening arrhythmia that at present cannot always be treated adequately. A realistic model of re-entry would be helpful to accurately guide catheter ablation for interruption of the circuit. In this review, models of electrical activation wavefront propagation during onset and maintenance of re-entrant VT are discussed. In particular, the relationship between activation mapping and maps of transition in infarct border zone thickness, which results in source-sink mismatch, is considered in detail and supplemented with additional data. Based on source-sink mismatch, the re-entry isthmus can be modeled from its boundary properties. Isthmus boundary segments with large transitions in infarct border zone thickness have large source-sink mismatch, and functional block forms there during VT. These alternate with segments having lesser thickness change and therefore lesser source-sink mismatch, which act as gaps, or entrance and exit points, to the isthmus during VT. Besides post-infarction substrates, the source-sink model is likely applicable to other types of volumetric changes in the myocardial conducting medium, such as when there is presence of fibrosis or dissociation of muscle fibers.

Cardiovascular involvement in celiac disease

Celiac disease (CD) is an autoimmune response to ingestion of gluten protein, which is found in wheat, rye, and barley grains, and results in both small intestinal manifestations, including villous atrophy, as well as systemic manifestations. The main treatment for the disease is a gluten-free diet (GFD), which typically results in the restoration of the small intestinal villi, and restoration of other affected organ systems, to their normal functioning. In an increasing number of recently published studies, there has been great interest in the occurrence of alterations in the cardiovascular system in untreated CD. Herein, published studies in which CD and cardiovascular terms appear in the title of the study were reviewed. The publications were categorized into one of several types: (1) articles (including cohort and case-control studies); (2) reviews and meta-analyses; (3) case studies (one to three patient reports); (4) letters; (5) editorials; and (6) abstracts (used when no full-length work had been published). The studies were subdivided as either heart or vascular studies, and were further characterized by the particular condition that was evident in conjunction with CD. Publication information was determined using the Google Scholar search tool. For each publication, its type and year of publication were tabulated. Salient information from each article was then compiled. It was determined that there has been a sharp increase in the number of CD - cardiovascular studies since 2000. Most of the publications are either of the type "article" or "case study". The largest number of documents published concerned CD in conjunction with cardiomyopathy (33 studies), and there have also been substantial numbers of studies published on CD and thrombosis (27), cardiovascular risk (17), atherosclerosis (13), stroke (12), arterial function (11), and ischemic heart disease (11). Based on the published research, it can be concluded that many types of cardiovascular issues can occur in untreated CD patients, but that most tend to resolve on a GFD, often in conjunction with the healing of small intestinal villous atrophy. However, in some cases the alterations are irreversible, underscoring the need for CD screening and treatment when cardiovascular issues arise of unknown etiology.

Use of shape-from-shading to characterize mucosal topography in celiac disease videocapsule images

AIM

To use a computerized shape-from-shading technique to characterize the topography of the small intestinal mucosa.

METHODS

Videoclips comprised of 100-200 images each were obtained from the distal duodenum in 10 celiac and 10 control patients. Images with high texture were selected from each videoclip and projected from two to three dimensions by using grayscale pixel brightness as the Z-axis spatial variable. The resulting images for celiac patients were then ordered using the Marsh score to estimate the degree of villous atrophy, and compared with control data.

RESULTS

Topographic changes in celiac patient three-dimensional constructs were often more variable as compared to controls. The mean absolute derivative in elevation was 2.34 ± 0.35 brightness units for celiacs vs 1.95 ± 0.28 for controls (P = 0.014). The standard deviation of the derivative in elevation was 4.87 ± 0.35 brightness units for celiacs vs 4.47 ± 0.36 for controls (P = 0.023). Celiac patients with Marsh IIIC villous atrophy tended to have the largest topographic changes. Plotted in two dimensions, celiac data could be separated from controls with 80% sensitivity and specificity.

CONCLUSION

Use of shape-from-shading to construct three-dimensional projections approximating the actual spatial geometry of the small intestinal substrate is useful to observe features not readily apparent in two-dimensional videocapsule images. This method represents a potentially helpful adjunct to detect areas of pathology during videocapsule analysis.

Coeliac disease and the videocapsule: what have we learned till now

Celiac disease is diagnosed in part by finding areas of pathology in the small bowel (SB) mucosa. This can often be difficult because the pathologic alterations, including atrophy of the small intestinal villi, can often be sparse and subtle. Some of the quantitative methods for detecting and measuring the presence of villous atrophy from videocapsule endoscopy (VCE) images are presented and discussed. These methods consist of static features of measurement including texture, gray level, and presence and abundance of fissures contained within each acquired image. The methods also consist of dynamic measurements including spectral analysis, and determining motion from a sequence of endoscopic images as obtained from a VCE clip. Thus far, several methods have been found useful to characterize the SB mucosa of untreated celiac disease patients versus control patients lacking villous atrophy, which have revealed significant differences in texture, frequency, and motion on analysis of VCE. In untreated celiac patients undergoing endoscopy, there tends to be greater magnitude of changes and spatial differences in textural descriptors, longer periodic components, indicating slower periodic activity, and differences in feature location, suggesting alterations in motility at areas of pathology as compared to patients without villous atrophy. Improvements in the quantitative analysis of VCE imaging in celiac patients is important to detect pathology in suspected patients, so that biopsies can be obtained from pertinent regions of the small intestinal mucosa. Improvements are also necessary so that patients with celiac disease can be monitored to evaluate the progress of mucosal healing after onset of treatment.

Quantitative analysis of patients with celiac disease by video capsule endoscopy: A deep learning method

BACKGROUND

Celiac disease is one of the most common diseases in the world. Capsule endoscopy is an alternative way to visualize the entire small intestine without invasiveness to the patient. It is useful to characterize celiac disease, but hours are need to manually analyze the retrospective data of a single patient. Computer-aided quantitative analysis by a deep learning method helps in alleviating the workload during analysis of the retrospective videos.

METHOD

Capsule endoscopy clips from 6 celiac disease patients and 5 controls were preprocessed for training. The frames with a large field of opaque extraluminal fluid or air bubbles were removed automatically by using a pre-selection algorithm. Then the frames were cropped and the intensity was corrected prior to frame rotation in the proposed new method. The GoogLeNet is trained with these frames. Then, the clips of capsule endoscopy from 5 additional celiac disease patients and 5 additional control patients are used for testing. The trained GoogLeNet was able to distinguish the frames from capsule endoscopy clips of celiac disease patients vs controls. Quantitative measurement with evaluation of the confidence was developed to assess the severity level of pathology in the subjects.

RESULTS

Relying on the evaluation confidence, the GoogLeNet achieved 100% sensitivity and specificity for the testing set. The t-test confirmed the evaluation confidence is significant to distinguish celiac disease patients from controls. Furthermore, it is found that the evaluation confidence may also relate to the severity level of small bowel mucosal lesions.

CONCLUSIONS

A deep convolutional neural network was established for quantitative measurement of the existence and degree of pathology throughout the small intestine, which may improve computer-aided clinical techniques to assess mucosal atrophy and other etiologies in real-time with videocapsule endoscopy.

Development of an automaton model of rotational activity driving atrial fibrillation

BACKGROUND

Atrial fibrillation (AF) is difficult to treat effectively, owing to uncertainty in where to best ablate to eliminate arrhythmogenic substrate. A model providing insight into the electrical activation events would be useful to guide catheter ablation strategy. Method A two-dimensional, 576×576 node automaton was developed to simulate atrial electrical activity. The substrate field was altered by the presence of differing refractory period at varying locations. Fibrosis was added in the form of short, randomly positioned lines of conduction block. Larger areas of block were used to simulate ablation lesions. Anisotropy was imposed in a 2:1 ratio. A premature electrical impulse from one of four grid corners was utilized to initiate activation.

RESULTS

Rotational activity was uninducible when refractory patch dimensions were less than 20×20mm. For larger refractory regions, a single premature stimulus was capable of inducing an average of 1.19±1.10 rotors, which often formed near the patch edges. A maximum of 5 rotors formed when refractory patch dimensions approached the size of the entire left atrial virtual field. Rotors formed along a refractory patch edge, after wavefront arrival was delayed at turning points or due to the presence of a fiber cluster of sufficient size. However, rotational activity could also occur around a large fiber cluster without the need of spatially variable refractoriness. When obstacles to conduction were lacking in size, nascent rotors drifted and either extinguished, or stabilized upon anchoring at a sufficiently large fiber cluster elsewhere in the field. Transient rotors terminated when traversing a region with differing refractory periods, if no obstacle to conduction was present to sufficiently delay wavefront arrival beyond the longest refractory period. Other rotors were annihilated when a nearby rotor with faster spin rate gradually interrupted the activation pathway. Elimination of anchors by removal, or by simulated ablation over a sufficient region, prevented rotor onset at a particular location where it would otherwise form.

CONCLUSIONS

The presence of obstacles to conduction and spatial differences in refractory period are important parameters for initiating and maintaining rotational activity in this simulation of an atrial substrate.

Recommendations to quantify villous atrophy in video capsule endoscopy images of celiac disease patients

AIM

To quantify the presence of villous atrophy in endoscopic images for improved automation.

METHODS

There are two main categories of quantitative descriptors helpful to detect villous atrophy: (1) Statistical and (2) Syntactic. Statistical descriptors measure the small intestinal substrate in endoscope-acquired images based on mathematical methods. Texture is the most commonly used statistical descriptor to quantify villous atrophy. Syntactic descriptors comprise a syntax, or set of rules, for analyzing and parsing the substrate into a set of objects with boundaries. The syntax is designed to identify and distinguish three-dimensional structures based on their shape.

RESULTS

The variance texture statistical descriptor is useful to describe the average variability in image gray level representing villous atrophy, but does not determine the range in variability and the spatial relationships between regions. Improved textural descriptors will incorporate these factors, so that areas with variability gradients and regions that are orientation dependent can be distinguished. The protrusion syntactic descriptor is useful to detect three-dimensional architectural components, but is limited to identifying objects of a certain shape. Improvement in this descriptor will require incorporating flexibility to the prototypical template, so that protrusions of any shape can be detected, measured, and distinguished.

CONCLUSION

Improved quantitative descriptors of villous atrophy are being developed, which will be useful in detecting subtle, varying patterns of villous atrophy in the small intestinal mucosa of suspected and known celiac disease patients.

Extraction and processing of videocapsule data to detect and measure the presence of villous atrophy in celiac disease patients

BACKGROUND

Videocapsule endoscopy is a relative new method to analyze the gastrointestinal tract for the presence of pathologic features. It is of relevance to detect villous atrophy in the small bowel, which is a defining symptom of celiac disease.

METHOD

In this tutorial, methods to extract and process videocapsule endoscopy data are elucidated. The algorithms, computer code, and paradigms to analyze image series are described in detail. The topics covered include extraction of data, analysis of texture, eigenanalysis, spectral analysis, three-dimensional projection, and estimation of motility. The basic paradigms to implement these processes are provided.

RESULTS

Examples of successful quantitative analysis implementations for selected untreated celiac disease patients with villous atrophy versus control patients with normal villi were illustrated. Based on the implementations, it was evident that celiac patients tended to have a rougher small intestinal texture as compared with control patients. From three-dimensional projection, celiac patients exhibited larger surface protrusions emanating from the small intestinal mucosa, which may represent clumps of atrophied villi. The periodicity of small intestinal contractions tends to be slower when villous atrophy is present, and the estimated degree of motility is reduced as compared with control image series. Basis image construction suggested that fissuring and mottling of the mucosal surface is predominant in untreated celiac patients, and mostly absent in controls.

CONCLUSIONS

Implementation of computerized methods, as described in this tutorial, will likely be useful for the automated detection and measurement of villous atrophy, and to map its extent along the small intestine of celiac patients.

Reprint of 'Model of unidirectional block formation leading to reentrant ventricular tachycardia in the infarct border zone of postinfarction canine hearts'

BACKGROUND

When the infarct border zone is stimulated prematurely, a unidirectional block line (UBL) can form and lead to double-loop (figure-of-eight) reentrant ventricular tachycardia (VT) with a central isthmus. The isthmus is composed of an entrance, center, and exit. It was hypothesized that for certain stimulus site locations and coupling intervals, the UBL would coincide with the isthmus entrance boundary, where infarct border zone thickness changes from thin-to-thick in the travel direction of the premature stimulus wavefront.

METHOD

A quantitative model was developed to describe how thin-to-thick changes in the border zone result in critically convex wavefront curvature leading to conduction block, which is dependent upon coupling interval. The model was tested in 12 retrospectively analyzed postinfarction canine experiments. Electrical activation was mapped for premature stimulation and for the first reentrant VT cycle. The relationship of functional conduction block forming during premature stimulation to functional block during reentrant VT was quantified.

RESULTS

For an appropriately placed stimulus, in accord with model predictions: 1. The UBL and reentrant VT isthmus lateral boundaries overlapped (error: 4.8±5.7mm). 2. The UBL leading edge coincided with the distal isthmus where the center-entrance boundary would be expected to occur. 3. The mean coupling interval was 164.6±11.0ms during premature stimulation and 190.7±20.4ms during the first reentrant VT cycle, in accord with model calculations, which resulted in critically convex wavefront curvature and functional conduction block, respectively, at the location of the isthmus entrance boundary and at the lateral isthmus edges.

DISCUSSION

Reentrant VT onset following premature stimulation can be explained by the presence of critically convex wavefront curvature and unidirectional block at the isthmus entrance boundary when the premature stimulation interval is sufficiently short. The double-loop reentrant circuit pattern is a consequence of wavefront bifurcation around this UBL followed by coalescence, and then impulse propagation through the isthmus. The wavefront is blocked from propagating laterally away from the isthmus by sharp increases in border zone thickness, which results in critically convex wavefront curvature at VT cycle lengths.

Atrial Tachycardias After Atrial Fibrillation Ablation Manifest Different Waveform Characteristics: Implications for Characterizing Tachycardias

INTRODUCTION

Atrial fibrillation (AF) ablation patients often manifest atrial tachycardias (AT) with atypical ECG morphologies that preclude accurate localization and mechanism. Diagnostic maneuvers used to define ATs during electrophysiology studies can be limited by tachycardia termination or transformation. Additional methods of characterizing post-AF ablation ATs are required.

METHODS AND RESULTS

We evaluated the utility of noninvasive ECG signal analytics in postablation AF patients for the following features: (1) Localization of ATs (i.e., right vs. left atrium), and (2) Identification of common left AT mechanisms (i.e., focal vs. macroreentrant). Atrial waveforms from the surface ECG were used to analyze (1) spectral organization, including dominant amplitude (DA) and mean spectral profile (MP), and (2) temporospatial variability, using temporospatial correlation coefficients. We studied 94 ATs in 71 patients who had undergone prior pulmonary vein isolation for AF and returned for a second ablation: (1) right atrial cavotricuspid-isthmus dependent (CTI) ATs (n = 21); (2) left atrial macroreentrant ATs (n = 41) and focal ATs (n = 32). Right CTI ATs manifested higher DAs and lower MPs than left ATs, indicative of greater stability and less complexity in the frequency spectrum. Left macroreentrant ATs possessed higher temporospatial organization than left focal ATs.

CONCLUSIONS

Noninvasively recorded atrial waveform signal analyses show that right ATs possess more stable activation properties than left ATs, and left macroreentrant ATs manifest higher temporospatial organization than left focal ATs. Further prospective analyses evaluating the role these novel ECG-derived tools can play to help localize and identify mechanisms of common ATs in AF ablation patients are warranted.

Model of unidirectional block formation leading to reentrant ventricular tachycardia in the infarct border zone of postinfarction canine hearts

BACKGROUND

When the infarct border zone is stimulated prematurely, a unidirectional block line (UBL) can form and lead to double-loop (figure-of-eight) reentrant ventricular tachycardia (VT) with a central isthmus. The isthmus is composed of an entrance, center, and exit. It was hypothesized that for certain stimulus site locations and coupling intervals, the UBL would coincide with the isthmus entrance boundary, where infarct border zone thickness changes from thin-to-thick in the travel direction of the premature stimulus wavefront.

METHOD

A quantitative model was developed to describe how thin-to-thick changes in the border zone result in critically convex wavefront curvature leading to conduction block, which is dependent upon coupling interval. The model was tested in 12 retrospectively analyzed postinfarction canine experiments. Electrical activation was mapped for premature stimulation and for the first reentrant VT cycle. The relationship of functional conduction block forming during premature stimulation to functional block during reentrant VT was quantified.

RESULTS

For an appropriately placed stimulus, in accord with model predictions: (1) The UBL and reentrant VT isthmus lateral boundaries overlapped (error: 4.8±5.7mm). (2) The UBL leading edge coincided with the distal isthmus where the center-entrance boundary would be expected to occur. (3) The mean coupling interval was 164.6±11.0ms during premature stimulation and 190.7±20.4ms during the first reentrant VT cycle, in accord with model calculations, which resulted in critically convex wavefront curvature with functional conduction block, respectively, at the location of the isthmus entrance boundary and at the lateral isthmus edges.

DISCUSSION

Reentrant VT onset following premature stimulation can be explained by the presence of critically convex wavefront curvature and unidirectional block at the isthmus entrance boundary when the premature stimulation interval is sufficiently short. The double-loop reentrant circuit pattern is a consequence of wavefront bifurcation around this UBL followed by coalescence, and then impulse propagation through the isthmus. The wavefront is blocked from propagating laterally away from the isthmus by sharp increases in border zone thickness, which results in critically convex wavefront curvature at VT cycle lengths.

Use of basis images for detection and classification of celiac disease

Celiac disease commonly occurs in approximately 1% of populations, but it can be difficult to diagnose. The standard method to diagnose celiac disease includes analysis of endoscopy images of the small intestinal mucosa to detect presence of villous atrophy, which can be subtle. We have devised a means to improve the image-based detection of villous atrophy and other abnormality in videocapsule endoscopy by means of incorporating basis images. Basis images were extracted from a series of 200 consecutive image frames acquired over 100 seconds at the level of the duodenal bulb in 13 celiac patients and in 13 controls. They were converted from color to 256 grayscale levels (gsl; 0 = black, 255 = white). Eight basis images were used for analysis. A histogram was constructed for each basis image, and the mean and standard deviation of the histogram values were tabulated. The significance of the difference in histogram mean level for celiacs versus controls was determined. Then the histogram mean was plotted versus the standard deviation, separately for all eight basis images, and also averaged for all bases combined. The mean histogram level for celiacs was 127.59+6.05 gsl versus 129.25+5.53 gsl for controls (p< 0.05). Thus celiac basis images tended to be darker and also more variable as compared with controls. For nonlinear classification, using the average of combined basis images, the sensitivity was 84.6% while the specificity was 92.3%. Using the single most important basis image for nonlinear classification, the sensitivity was 84.6% while the specificity was 76.9%. Construction of basis images can be useful to condense videocapsule image series into salient information, for detection of differences in grayscale level mean and variation in celiac versus control image series, and for classification of celiac versus control videoclips with nonlinear discriminant functions.

Methods to quantitate videocapsule endoscopy images in celiac disease

In this work, bioengineering methods that can be used to quantitatively analyze videocapsule endoscopy images that have been acquired from celiac patients versus controls are described. For videocapsule endoscopic analysis, each patient swallows a capsule which contains an imaging device and light source. In celiac and control patients, images are acquired and analyzed at the level of the small intestine. The data used for videocapsule analysis consisted of high resolution images of dimension 576 × 576 pixels, acquired twice per second. The goal of the quantitative analysis is to detect abnormality in celiac patient images as compared with controls. Several types of abnormality can exist at the level of the small intestine in celiac patients. In untreated patients, and often even after treatment with a gluten-free diet, there can be villous atrophy, as well as presence of fissures and a mottled appearance. To detect and discern these abnormalities, several methods of statistical and structural feature extraction and selection are described. It was found that there is a significantly greater variation in image texture and average brightness level in celiac patients as compared with controls (p < 0.05). Celiac patients have a longer dominant period as compared with controls, averaging 6.4 ± 2.6 seconds versus 4.7 ± 1.6 seconds in controls (p = 0.001). This suggests that overall motility is slower in the celiac patients. Furthermore, the mean number of villous protrusions per image was found to be 402.2 ± 15.0 in celiac patients versus 420.8 ± 24.0 in control patients (p < 0.001). The average protrusion width was 14.66 ± 1.04 pixels in celiacs versus 13.91 ± 1.47 pixels in controls (p = 0.01). The mean protrusion height was 3.10 ± 0.26 grayscale levels for celiacs versus 2.70 ± 0.43 grayscale levels for controls (p < 0.001). Thus celiac patients tended to have fewer protrusions, and these were more varied in shape, tending to be blunted, as compared with controls, which more often had fine, uniform protrusions. A variety of computerized methods are now available to quantitate videocapsule images for comparison of celiac versus control patients. Since these methods are based on computer algorithms, they can be automated and there is no variation in the results due to observer bias. These methods readily lend themselves to automation, so that it may be possible to map the entire small intestine for presence of abnormality in real-time. It is also possible to develop an automated, quantitative clinical score which can be displayed with real-time update during the procedure. This would be useful to determine progress in celiac patients on a gluten-free diet, and to better understand the properties of the healing process in these patients.

Suggestions for automatic quantitation of endoscopic image analysis to improve detection of small intestinal pathology in celiac disease patients

Although many groups have attempted to develop an automated computerized method to detect pathology of the small intestinal mucosa caused by celiac disease, the efforts have thus far failed. This is due in part to the occult presence of the disease. When pathological evidence of celiac disease exists in the small bowel it is visually often patchy and subtle. Due to presence of extraneous substances such as air bubbles and opaque fluids, the use of computerized automation methods have only been partially successful in detecting the hallmarks of the disease in the small intestine-villous atrophy, fissuring, and a mottled appearance. By using a variety of computerized techniques and assigning a weight or vote to each technique, it is possible to improve the detection of abnormal regions which are indicative of celiac disease, and of treatment progress in diagnosed patients. Herein a paradigm is suggested for improving the efficacy of automated methods for measuring celiac disease manifestation in the small intestinal mucosa. The suggestions are applicable to both standard and videocapsule endoscopic imaging, since both methods could potentially benefit from computerized quantitation to improve celiac disease diagnosis.

Atrial electrogram discordance during baseline vs reinduced atrial fibrillation: Potential ramifications for ablation procedures

BACKGROUND

There are scant data comparing the electrogram (EGM) signal characteristics of atrial fibrillation (AF) at baseline vs electrically induced states during ablation procedures.

OBJECTIVE

The purpose of this study was to use novel intracardiac signal analysis techniques to gain insights into the effects of catheter ablation and AF reinduction on AF EGMs in patients with persistent AF.

METHODS

We collected left atrial EGMs in patients undergoing first ablation for persistent AF at 3 time intervals: (1) AF at baseline; (2) AF after pulmonary vein isolation (PVI); and (3) AF after post-PVI cardioversion and subsequent reinduction. We analyzed 2 EGM spectral characteristics: (1) dominant frequency and (2) spectral complexity; and 2 EGM morphologic characteristics: (1) morphology variation and (2) pattern repetitiveness.

RESULTS

There were no differences in AF dominant frequency, dominant amplitude, spectral complexity, or metrics of EGM morphology or repetitiveness at baseline vs after PVI. However, dominant frequency, dominant amplitude, and spectral complexity differed significantly after direct current cardioversion and reinduction of AF.

CONCLUSION

The frequency, spectral complexity, and local EGM morphologies of AF do not significantly change over the course of a PVI procedure in patients with persistent AF. However, reinduction of AF after direct current cardioversion results in different dominant frequency and spectral complexity, consistent with a change in the characteristics of the perpetuating source(s) of the newly induced AF. These data suggest that AF properties can vary significantly between baseline and reinduced AF, with potential clinical ramifications for outcomes of catheter ablation procedures.

Quantitative image analysis of celiac disease

We outline the use of quantitative techniques that are currently used for analysis of celiac disease. Image processing techniques can be useful to statistically analyze the pixular data of endoscopic images that is acquired with standard or videocapsule endoscopy. It is shown how current techniques have evolved to become more useful for gastroenterologists who seek to understand celiac disease and to screen for it in suspected patients. New directions for focus in the development of methodology for diagnosis and treatment of this disease are suggested. It is evident that there are yet broad areas where there is potential to expand the use of quantitative techniques for improved analysis in suspected or known celiac disease patients.

Optimization of novel spectral estimator for fractionated electrogram analysis is helpful to discern atrial fibrillation type

INTRODUCTION

Paroxysmal versus persistent atrial fibrillation (AF) can be distinguished based on differences in the spectral parameters of fractionated atrial electrograms. Maximization of these differences would improve characterization of the arrhythmogenic substrate. A novel spectral estimator (NSE) has been shown previously to provide greater distinction in AF spectral parameters as compared with the Fourier transform estimator. Herein, it is described how the differences in NSE spectral parameters can be further improved.

METHOD

In 10 persistent and 9 paroxysmal AF patients undergoing electrophysiologic study, fractionated electrograms were acquired from the distal bipolar ablation electrode. A total of 204 electrograms were recorded from the pulmonary vein (PV) antra and from the anterior and posterior left atrial free wall. The following spectral parameters were measured: the dominant frequency (DF), which reflects local activation rate, the DF amplitude (DA), and the mean spectral profile (MP), which represents background electrical activity. To optimize differences in parameters between paroxysmal versus persistent AF patients, the NSE was varied by selectively removing subharmonics, using a threshold. The threshold was altered in steps to determine the optimal subharmonics removal.

RESULTS

At the optimal threshold level, mean differences in persistent versus paroxysmal AF spectral parameters were: ΔDA=+0.371 mV, ΔDF=+0.737 Hz, and ΔMP=-0.096 mV. When subharmonics were not removed, the differences were substantially less: ΔDA=+0.301 mV, ΔDF=+0.699 Hz, and ΔMP=-0.063 mV.

CONCLUSIONS

NSE optimization produces greater spectral parameter difference between persistent versus paroxysmal AF data. Quantifying spectral parameter differences can be assistive in characterizing the arrhythmogenic substrate.

Software algorithm and hardware design for real-time implementation of new spectral estimator

Background

Real-time spectral analyzers can be difficult to implement for PC computer-based systems because of the potential for high computational cost, and algorithm complexity. In this work a new spectral estimator (NSE) is developed for real-time analysis, and compared with the discrete Fourier transform (DFT).

Method

Clinical data in the form of 216 fractionated atrial electrogram sequences were used as inputs. The sample rate for acquisition was 977 Hz, or approximately 1 millisecond between digital samples. Real-time NSE power spectra were generated for 16,384 consecutive data points. The same data sequences were used for spectral calculation using a radix-2 implementation of the DFT. The NSE algorithm was also developed for implementation as a real-time spectral analyzer electronic circuit board.

Results

The average interval for a single real-time spectral calculation in software was 3.29 μs for NSE versus 504.5 μs for DFT. Thus for real-time spectral analysis, the NSE algorithm is approximately 150× faster than the DFT. Over a 1 millisecond sampling period, the NSE algorithm had the capability to spectrally analyze a maximum of 303 data channels, while the DFT algorithm could only analyze a single channel. Moreover, for the 8 second sequences, the NSE spectral resolution in the 3-12 Hz range was 0.037 Hz while the DFT spectral resolution was only 0.122 Hz. The NSE was also found to be implementable as a standalone spectral analyzer board using approximately 26 integrated circuits at a cost of approximately $500. The software files used for analysis are included as a supplement, please see the Additional files 1 and 2.

Conclusions

The NSE real-time algorithm has low computational cost and complexity, and is implementable in both software and hardware for 1 millisecond updates of multichannel spectra. The algorithm may be helpful to guide radiofrequency catheter ablation in real time.

Model of Bipolar Electrogram Fractionation and Conduction Block Associated With Activation Wavefront Direction at Infarct Border Zone Lateral Isthmus Boundaries

Background—

Improved understanding of the mechanisms underlying infarct border zone electrogram fractionation may be helpful to identify arrhythmogenic regions in the postinfarction heart. We describe the generation of electrogram fractionation from changes in activation wavefront curvature in experimental canine infarction.

Methods and Results—

A model was developed to estimate the extracellular signal shape that would be generated by wavefront propagation parallel to versus perpendicular to the lateral boundary (LB) of the reentrant ventricular tachycardia (VT) isthmus or diastolic pathway. LBs are defined as locations where functional block forms during VT, and elsewhere they have been shown to coincide with sharp thin-to-thick transitions in infarct border zone thickness. To test the model, bipolar electrograms were acquired from infarct border zone sites in 10 canine heart experiments 3 to 5 days after experimental infarction. Activation maps were constructed during sinus rhythm and during VT. The characteristics of model-generated versus actual electrograms were compared. Quantitatively expressed VT fractionation (7.6±1.2 deflections; 16.3±8.9-ms intervals) was similar to model-generated values with wavefront propagation perpendicular to the LB (9.4±2.4 deflections; 14.4±5.2-ms intervals). Fractionation during sinus rhythm (5.9±1.8 deflections; 9.2±4.4-ms intervals) was similar to model-generated fractionation with wavefront propagation parallel to the LB (6.7±3.1 deflections; 7.1±3.8-ms intervals). VT and sinus rhythm fractionation sites were adjacent to LBs ≈80% of the time.

Conclusions—

The results suggest that in a subacute canine infarct model, the LBs are a source of activation wavefront discontinuity and electrogram fractionation, with the degree of fractionation being dependent on activation rate and wavefront orientation with respect to the LB.

Temporal stability in the spectral representation of complex fractionated atrial electrograms

BACKGROUND

Although local electrograms during atrial fibrillation (AF) are often spectrally analyzed over 8-second (8s) intervals, changes may be common over intervals as short as 2s. We sought to determine whether averaged 2s measurements of electrogram spectral parameters were similar to 8s measurements, and whether the 2s intervals could provide an estimate of the temporal stability of the signal frequency content in paroxysmal versus persistent AF.

METHODS

Complex fractionated atrial electrograms (CFAEs) were acquired outside the pulmonary vein ostia and from free wall sites in nine paroxysmal and 10 longstanding persistent AF patients. Using a 2s sliding calculation window, a frequency spectrum was computed every 100 ms over an interval of 8.4 seconds (82 spectra in total). The dominant frequency (DF), the dominant amplitude (DA), and the mean spectral profile (MP) were measured. The 2s measurements were compared to single 8.4-second interval measurements. Coefficients of variation (COV) were computed from the 82 spectra for each CFAE recording to determine temporal variability of parameters.

RESULTS

Over the sliding 2s computation intervals, as for fixed 8.4-second computation intervals, mean DA and DF were significantly higher in longstanding persistent AF while MP was significantly higher in paroxysmal AF (P ≤ 0.001). The COV was significantly higher for the DF parameter in paroxysmal AF (P < 0.001) and significantly higher for the MP parameter in persistent AF (P < 0.02).

CONCLUSIONS

For both paroxysmal and persistent AF data, the 2s sliding window averages provide similar results to single 8.4-second intervals, and information regarding temporal stability was additionally obtained in the process.