Multi-class Convolutional Neural Network Research Articles

Search for core collapse supernovae signals in LIGO’s third observation run using a network of gravitational wave detectors integrated with a multiclass convolutional neural network

Search for core collapse supernovae signals in LIGO’s third observation run using a network of gravitational wave detectors integrated with a multiclass convolutional neural network

Automatic segmentation of the spinal cord nerve rootlets

Abstract Precise identification of spinal nerve rootlets is relevant to delineate spinal levels for the study of functional activity in the spinal cord. The goal of this study was to develop an automatic method for the semantic segmentation of spinal nerve rootlets from T2-weighted magnetic resonance imaging (MRI) scans. Images from two open-access 3T MRI datasets were used to train a 3D multi-class convolutional neural network using an active learning approach to segment C2-C8 dorsal nerve rootlets. Each output class corresponds to a spinal level. The method was tested on 3T T2-weighted images from three datasets unseen during training to assess inter-site, inter-session, and inter-resolution variability. The test Dice score was 0.67 ± 0.16 (mean ± standard deviation across testing images and rootlets levels), suggesting a good performance. The method also demonstrated low inter-vendor and inter-site variability (coefficient of variation ≤ 1.41%), as well as low inter-session variability (coefficient of variation ≤ 1.30%), indicating stable predictions across different MRI vendors, sites, and sessions. The proposed methodology is open-source and readily available in the Spinal Cord Toolbox (SCT) v6.2 and higher.

Application of multi-label convolutional neural networks on enhancement of reliability of classification, a case study on bearing fault diagnostics

In contrast to the mutli-class classification where one and only one class would be assigned to each instance, multi-label classification allows for more flexible classification and possibility of having multiple labels for instance. This property enables us to model the dependencies between the classes such as effect of one label on the probability of occurrence of the other label. Considering these connections between the classes can help us to enhance the reliability of classifier and handle the contradictory prediction. Much like how we can assign multiple classes to an instance, we can also have an instance that remains unlabeled. This class is typically referred to as the "neutral" label, allowing us to bluntly state that we lack sufficient information to label the instance with adequate certainty. Leveraging these properties, a 1D convolutional neural network (CNN) has been proposed to classify common faults occurring in the bearings. Network has the capability to assign a Confidence level between 0 to 1 to each class independently. Later, a threshold has been set on each class to identify the instance labels if any. Subsequently, instances with contradictory predicted label have been flagged as undetermined and removed from the predicted result. thus, the remaining classified instances can be labeled with sufficient confident. In the end, the multi-label CNN model performance has been compared with equivalent multi-class CNN model. Technique has been evaluated using MAFAULDA open dataset which is composed of 1951 multivariate time-series acquired by a triaxial accelerometer, a microphone, and a tachometer using SpectraQuest’s Machinery Fault Simulator (MFS) experimentally simulating different bearing states including normal operation as well as inner and outer race bearing faults. A data reduction scheme has been performed on the dataset to make the diagnostic more challenging and evaluate the performance of CNN in more realistic condition with insufficient data.

Deep Learning-Based Multi-Class Segmentation of the Paranasal Sinuses of Sinusitis Patients Based on Computed Tomographic Images.

Accurate paranasal sinus segmentation is essential for reducing surgical complications through surgical guidance systems. This study introduces a multiclass Convolutional Neural Network (CNN) segmentation model by comparing four 3D U-Net variations-normal, residual, dense, and residual-dense. Data normalization and training were conducted on a 40-patient test set (20 normal, 20 abnormal) using 5-fold cross-validation. The normal 3D U-Net demonstrated superior performance with an F1 score of 84.29% on the normal test set and 79.32% on the abnormal set, exhibiting higher true positive rates for the sphenoid and maxillary sinus in both sets. Despite effective segmentation in clear sinuses, limitations were observed in mucosal inflammation. Nevertheless, the algorithm's enhanced segmentation of abnormal sinuses suggests potential clinical applications, with ongoing refinements expected for broader utility.

Diabetic Retinopathy Detection Using InceptionResnet-V2 and Densenet121

This project addresses the global health challenge posed by the prevalence of diabetic retinopathy (DR) by developing an efficient automated diagnostic system. The dataset, consisting of diverse high-resolution retinal images, underwent preprocessing to categorize images into No DR (0) and DR (1-4) classes. The First initial binary classification model using a Convolutional Neural Network (CNN) discriminated between healthy and diseased retinas. Subsequently, The second multi-class CNN model was designed to predict the severity of diabetic retinopathy (DR) across a spectrum from mild (1) to proliferative DR (4), enabling a fine-grained analysis for early identification of cases requiring urgent intervention. To address real-world complexities, potential noise in the dataset, including artifacts and exposure variations, was acknowledged. The CNN models were designed to exhibit resilience to these challenges, ensuring robust performance in clinical settings. Preprocessing is considered the common occurrence of image inversion in retinal imaging by incorporating anatomical features, such as macula position and notches, to correctly identify image orientation and enhance result interpretability. The proposed automated analysis system demonstrated promising results in accurately categorizing retinal images into No DR and DR, as well as assigning severity scores for diabetic retinopathy. This project contributes significantly to computer-aided diagnostics, Supplying a dependable instrument for promptly identifying and addressing cases of diabetic retinopathy.

Multi-Class Deep Learning Model for Detecting Pediatric Distal Forearm Fractures Based on the AO/OTA Classification.

Common pediatric distal forearm fractures necessitate precise detection. To support prompt treatment planning by clinicians, our study aimed to create a multi-class convolutional neural network (CNN) model for pediatric distal forearm fractures, guided by the AO Foundation/Orthopaedic Trauma Association (AO/ATO) classification system for pediatric fractures. The GRAZPEDWRI-DX dataset (2008-2018) of wrist X-ray images was used. We labeled images into four fracture classes (FRM, FUM, FRE, and FUE with F, fracture; R, radius; U, ulna; M, metaphysis; and E, epiphysis) based on the pediatric AO/ATO classification. We performed multi-class classification by training a YOLOv4-based CNN object detection model with 7006 images from 1809 patients (80% for training and 20% for validation). An 88-image test set from 34 patients was used to evaluate the model performance, which was then compared to the diagnosis performances of two readers-an orthopedist and a radiologist. The overall mean average precision levels on the validation set in four classes of the model were 0.97, 0.92, 0.95, and 0.94, respectively. On the test set, the model's performance included sensitivities of 0.86, 0.71, 0.88, and 0.89; specificities of 0.88, 0.94, 0.97, and 0.98; and area under the curve (AUC) values of 0.87, 0.83, 0.93, and 0.94, respectively. The best performance among the three readers belonged to the radiologist, with a mean AUC of 0.922, followed by our model (0.892) and the orthopedist (0.830). Therefore, using the AO/OTA concept, our multi-class fracture detection model excelled in identifying pediatric distal forearm fractures.

Enhanced Skin Disease Detection and Classification System Using Deep Learning Technique

Even in this era of advanced technology, physicians still encounter challenges in accurately identifying skin diseases. Severe skin conditions necessitate urgent hospitalization, yet the diagnostic process involves costly tests with delayed results, exacerbating the worsening of infections. In rural areas, patients often ignore initial symptoms due to limited medical services, allowing conditions to progress unchecked. This underscores the pressing need for highly accurate automated skin disease detection systems. In response to this demand, we developed a multiclass convolutional neural network model aimed at discerning between healthy and diseased skin. Employing Python, we constructed a program utilizing this convolutional neural network with pre-trained networks- AlexNet and VGG19 to address the issue effectively. The primary objective of our project was to categorize five different skin illnesses using input images. By providing our model with pictures of affected skin regions, we aimed to detect ailments like Eczema, Nail fungus, Melanoma, Bullous, and Vascular Tumor. Upon processing the input image, our model would accurately identify and output the specific illness affecting the skin. This streamlined approach significantly aids in timely and accurate diagnosis, crucial for early intervention and treatment of various skin conditions

Multi-organ segmentation of CT via convolutional neural network: impact of training setting and scanner manufacturer

Objective. Automated organ segmentation on CT images can enable the clinical use of advanced quantitative software devices, but model performance sensitivities must be understood before widespread adoption can occur. The goal of this study was to investigate performance differences between Convolutional Neural Networks (CNNs) trained to segment one (single-class) versus multiple (multi-class) organs, and between CNNs trained on scans from a single manufacturer versus multiple manufacturers. Methods. The multi-class CNN was trained on CT images obtained from 455 whole-body PET/CT scans (413 for training, 42 for testing) taken with Siemens, GE, and Phillips PET/CT scanners where 16 organs were segmented. The multi-class CNN was compared to 16 smaller single-class CNNs trained using the same data, but with segmentations of only one organ per model. In addition, CNNs trained on Siemens-only (N = 186) and GE-only (N = 219) scans (manufacturer-specific) were compared with CNNs trained on data from both Siemens and GE scanners (manufacturer-mixed). Segmentation performance was quantified using five performance metrics, including the Dice Similarity Coefficient (DSC). Results. The multi-class CNN performed well compared to previous studies, even in organs usually considered difficult auto-segmentation targets (e.g., pancreas, bowel). Segmentations from the multi-class CNN were significantly superior to those from smaller single-class CNNs in most organs, and the 16 single-class models took, on average, six times longer to segment all 16 organs compared to the single multi-class model. The manufacturer-mixed approach achieved minimally higher performance over the manufacturer-specific approach. Significance. A CNN trained on contours of multiple organs and CT data from multiple manufacturers yielded high-quality segmentations. Such a model is an essential enabler of image processing in a software device that quantifies and analyzes such data to determine a patient’s treatment response. To date, this activity of whole organ segmentation has not been adopted due to the intense manual workload and time required.

Ensembles of spectral-spatial convolutional neural network models for classifying soil types in hyperspectral images

The paper presents a study of various approaches to the classification of soil covers based on neural network algorithms using hyperspectral remote and proximal sensing of the Earth. The spectral distributions were recorded in the laboratory using an Offner imaging scanning hyperspectrometer. Spectral-spatial characteristics of nine soil samples from various parts of a farming land in the Samara region were experimentally studied. Using a method of energy dispersion microanalysis, the correspondence between the hyperspectral data and the chemical composition of the samples taken was established. Based on the data obtained, a neural network-aided classification of soil samples was implemented depending on the content of constituent elements such as carbon and calcium. A normalized spectral-spatial convolutional neural network was used as a classifier. As a result of the work, an approach to the classification of high-resolution hyper-spectral images based on the refinement of a multiclass convolutional neural network using an ensemble of binary classifiers is proposed. It is shown that the classification of soil samples by carbon and calcium content is carried out with an accuracy of 0.96.

A plum selection system that uses a multi-class Convolutional Neural Network (CNN)

Cultivation of the Horvin plum is one of the main economic activities in the region of Márquez, in the department of Boyacá, in Colombia. However, its selection process, which needs to be optimized and improved, is still done manually and leads to delays, high labor costs and errors in its classification due to long working days. In this paper, an electromechanical system for plum classification into three classes –morphologically defined by the experience of farmers in the region— is proposed, in nutshell: The first category is the largest and most desirable, the second is smaller in size compared to the first, and the third exhibits visual impairments, rendering it undesirable. For this, 1928 samples, distributed in three categories, were entered into a convolutional neural network using the low-cost architecture available for the research. Initially, a classification accuracy of 80% was obtained for the validation data. Ultimately, following a regularization process that included DropOut and data augmentation, the algorithm achieved a 90% accuracy rate on samples not encountered during the training phase. In addition, through the Grad-CAM algorithm, the operation of the CNN was ratified, evidencing the patterns learned in each of the convolutional layers. Considering the above, it was possible to implement a machine vision system capable of classifying plums into three categories using low-cost devices.

Subregional pharyngeal changes after orthognathic surgery in skeletal Class III patients analyzed by convolutional neural networks-based segmentation

ObjectivesTo evaluate the accuracy of fully automatic segmentation of pharyngeal volume of interests (VOIs) before and after orthognathic surgery in skeletal Class III patients using a convolutional neural network (CNN) model and to investigate the clinical applicability of artificial intelligence for quantitative evaluation of treatment changes in pharyngeal VOIs. Methods310 cone-beam computed tomography (CBCT) images were divided into a training set (n = 150), validation set (n = 40), and test set (n = 120). The test datasets comprised matched pairs of pre- and post-treatment images of 60 skeletal Class III patients (mean age 23.1 ± 5.0 years; ANB<-2⁰) who underwent bimaxillary orthognathic surgery with orthodontic treatment. A 3D U-Net CNNs model was applied for fully automatic segmentation and measurement of subregional pharyngeal volumes of pre-treatment (T0) and post-treatment (T1) scans. The model's accuracy was compared to semi-automatic segmentation outcomes by humans using the dice similarity coefficient (DSC) and volume similarity (VS). The correlation between surgical skeletal changes and model accuracy was obtained. ResultsThe proposed model achieved high performance of subregional pharyngeal segmentation on both T0 and T1 images, representing a significant T1–T0 difference of DSC only in the nasopharynx. Region-specific differences amongst pharyngeal VOIs, which were observed at T0, disappeared on the T1 images. The decreased DSC of nasopharyngeal segmentation after treatment was weakly correlated with the amount of maxillary advancement. There was no correlation between the mandibular setback amount and model accuracy. ConclusionsThe proposed model offers fast and accurate subregional pharyngeal segmentation on both pre-treatment and post-treatment CBCT images in skeletal Class III patients. Clinical significanceWe elucidated the clinical applicability of the CNNs model to quantitatively evaluate subregional pharyngeal changes after surgical-orthodontic treatment, which offers a basis for developing a fully integrated multiclass CNNs model to predict pharyngeal responses after dentoskeletal treatments.

Analysing cerebrospinal fluid with explainable deep learning: From diagnostics to insights.

Analysis of cerebrospinal fluid (CSF) is essential for diagnostic workup of patients with neurological diseases and includes differential cell typing. The current gold standard is based on microscopic examination by specialised technicians and neuropathologists, which is time-consuming, labour-intensive and subjective. We, therefore, developed an image analysis approach based on expert annotations of 123,181 digitised CSF objects from 78 patients corresponding to 15 clinically relevant categories and trained a multiclass convolutional neural network (CNN). The CNN classified the 15 categories with high accuracy (mean AUC 97.3%). By using explainable artificial intelligence (XAI), we demonstrate that the CNN identified meaningful cellular substructures in CSF cells recapitulating human pattern recognition. Based on the evaluation of 511 cells selected from 12 different CSF samples, we validated the CNN by comparing it with seven board-certified neuropathologists blinded for clinical information. Inter-rater agreement between the CNN and the ground truth was non-inferior (Krippendorff's alpha 0.79) compared with the agreement of seven human raters and the ground truth (mean Krippendorff's alpha 0.72, range 0.56-0.81). The CNN assigned the correct diagnostic label (inflammatory, haemorrhagic or neoplastic) in 10 out of 11 clinical samples, compared with 7-11 out of 11 by human raters. Our approach provides the basis to overcome current limitations in automated cell classification for routine diagnostics and demonstrates how a visual explanation framework can connect machine decision-making with cell properties and thus provide a novel versatile and quantitative method for investigating CSF manifestations of various neurological diseases.

High-throughput 3DRA segmentation of brain vasculature and aneurysms using deep learning

Background and Objectives:Automatic segmentation of the cerebral vasculature and aneurysms facilitates incidental detection of aneurysms. The assessment of aneurysm rupture risk assists with pre-operative treatment planning and enables in-silico investigation of cerebral hemodynamics within and in the vicinity of aneurysms. However, ensuring precise and robust segmentation of cerebral vessels and aneurysms in neuroimaging modalities such as three-dimensional rotational angiography (3DRA) is challenging. The vasculature constitutes a small proportion of the image volume, resulting in a large class imbalance (relative to surrounding brain tissue). Additionally, aneurysms and vessels have similar image/appearance characteristics, making it challenging to distinguish the aneurysm sac from the vessel lumen.Methods:We propose a novel multi-class convolutional neural network to tackle these challenges and facilitate the automatic segmentation of cerebral vessels and aneurysms in 3DRA images. The proposed model is trained and evaluated on an internal multi-center dataset and an external publicly available challenge dataset.Results:On the internal clinical dataset, our method consistently outperformed several state-of-the-art approaches for vessel and aneurysm segmentation, achieving an average Dice score of 0.81 (0.15 higher than nnUNet) and an average surface-to-surface error of 0.20 mm (less than the in-plane resolution (0.35 mm/pixel)) for aneurysm segmentation; and an average Dice score of 0.91 and average surface-to-surface error of 0.25 mm for vessel segmentation. In 223 cases of a clinical dataset, our method accurately segmented 190 aneurysm cases.Conclusions:The proposed approach can help address class imbalance problems and inter-class interference problems in multi-class segmentation. Besides, this method performs consistently on clinical datasets from four different sources and the generated results are qualified for hemodynamic simulation. Code available at https://github.com/cistib/vessel-aneurysm-segmentation.

Multiclass convolutional neural network based classification for the diagnosis of brain MRI images

Magnetic Resonance Imaging (MRI) modality is commonly used by radiologists in the medical field to diagnose an abnormality in the brain. The manual diagnosis process of grade identification is a challenging task that may lead to a False Negative, False Positive and it is also difficult to diagnose the early stages of abnormality. Manual misprediction may reduce the chance of survival in people with brain abnormalities. Hence, computer-aided diagnosis in additional supports the radiologist to identify the abnormality accurately even in the early stages of a brain tumor. In our proposed system, we have developed a Multi Class Convolutional Neural Network model (MCCNN) to identify the tumor in Brain MRI. In this work, BRATS 2015 and Figshare Data are used. After pre-processing, the feature vector is constructed from MRI using the convolution and pooling layers of CNN. Finally, the softmax layer of CNN identifies the tumor. For analyzing the performance of the proposed MCCNN model, two experiments such as Experiment I and Experiment II are conducted. The designed MCCNN provides less complexity when compared to pre-trained models and other CNN-based networks in literature. It decreases loss value, False Negative Rate, False Positive Rate and increases the accuracy of classification. The proposed MCCNN-based model attained a noteworthy performance with an accuracy of 99% in Experiment I and an accuracy of 96% in Experiment II.

Evaluation of Decision Fusions for Classifying Karst Wetland Vegetation Using One-Class and Multi-Class CNN Models with High-Resolution UAV Images

Combining deep learning and UAV images to map wetland vegetation distribution has received increasing attention from researchers. However, it is difficult for one multi-classification convolutional neural network (CNN) model to meet the accuracy requirements for the overall classification of multi-object types. To resolve these issues, this paper combined three decision fusion methods (Majority Voting Fusion, Average Probability Fusion, and Optimal Selection Fusion) with four CNNs, including SegNet, PSPNet, DeepLabV3+, and RAUNet, to construct different fusion classification models (FCMs) for mapping wetland vegetations in Huixian Karst National Wetland Park, Guilin, south China. We further evaluated the effect of one-class and multi-class FCMs on wetland vegetation classification using ultra-high-resolution UAV images and compared the performance of one-class classification (OCC) and multi-class classification (MCC) models for karst wetland vegetation. The results highlight that (1) the use of additional multi-dimensional UAV datasets achieved better classification performance for karst wetland vegetation using CNN models. The OCC models produced better classification results than MCC models, and the accuracy (average of IoU) difference between the two model types was 3.24–10.97%. (2) The integration of DSM and texture features improved the performance of FCMs with an increase in accuracy (MIoU) from 0.67% to 8.23% when compared to RGB-based karst wetland vegetation classifications. (3) The PSPNet algorithm achieved the optimal pixel-based classification in the CNN-based FCMs, while the DeepLabV3+ algorithm produced the best attribute-based classification performance. (4) Three decision fusions all improved the identification ability for karst wetland vegetation compared to single CNN models, which achieved the highest IoUs of 81.93% and 98.42% for Eichhornia crassipes and Nelumbo nucifera, respectively. (5) One-class FCMs achieved higher classification accuracy for karst wetland vegetation than multi-class FCMs, and the highest improvement in the IoU for karst herbaceous plants reached 22.09%.

Computer Aided Diagnosis of Melanoma Using Deep Neural Networks and Game Theory: Application on Dermoscopic Images of Skin Lesions.

Early detection of melanoma remains a daily challenge due to the increasing number of cases and the lack of dermatologists. Thus, AI-assisted diagnosis is considered as a possible solution for this issue. Despite the great advances brought by deep learning and especially convolutional neural networks (CNNs), computer-aided diagnosis (CAD) systems are still not used in clinical practice. This may be explained by the dermatologist's fear of being misled by a false negative and the assimilation of CNNs to a "black box", making their decision process difficult to understand by a non-expert. Decision theory, especially game theory, is a potential solution as it focuses on identifying the best decision option that maximizes the decision-maker's expected utility. This study presents a new framework for automated melanoma diagnosis. Pursuing the goal of improving the performance of existing systems, our approach also attempts to bring more transparency in the decision process. The proposed framework includes a multi-class CNN and six binary CNNs assimilated to players. The players' strategies is to first cluster the pigmented lesions (melanoma, nevus, and benign keratosis), using the introduced method of evaluating the confidence of the predictions, into confidence level (confident, medium, uncertain). Then, a subset of players has the strategy to refine the diagnosis for difficult lesions with medium and uncertain prediction. We used EfficientNetB5 as the backbone of our networks and evaluated our approach on the public ISIC dataset consisting of 8917 lesions: melanoma (1113), nevi (6705) and benign keratosis (1099). The proposed framework achieved an area under the receiver operating curve (AUROC) of 0.93 for melanoma, 0.96 for nevus and 0.97 for benign keratosis. Furthermore, our approach outperformed existing methods in this task, improving the balanced accuracy (BACC) of the best compared method from 77% to 86%. These results suggest that our framework provides an effective and explainable decision-making strategy. This approach could help dermatologists in their clinical practice for patients with atypical and difficult-to-diagnose pigmented lesions. We also believe that our system could serve as a didactic tool for less experienced dermatologists.

Automated bow shock and magnetopause boundary detection with Cassini using threshold and deep learning methods

Two algorithms set for automatic detection of bow shock (BS) and magnetopause (MP) boundaries at Saturn using in situ magnetic field and plasma data acquired by the Cassini spacecraft are presented. Traditional threshold-based and modern deep learning algorithms were investigated for the task of boundary detection. Sections of Cassini’s orbits were pre-selected based on empirical BS and MP boundary models, and from outlier detection in magnetic field data using an autoencoder neural network. The threshold method was applied to pre-selected magnetic field and plasma data independently to compute parameters to which a threshold was applied to determine the presence of a boundary. The deep learning method used a type of convolutional neural network (CNN) called ResNet on images of magnetic field time series data and electron energy-time spectrograms to classify the presence of boundaries. 2012 data were held out of the training data to test and compare the algorithms on unseen data. The comparison showed that the CNN method applied to plasma data outperformed the threshold method. A final multiclass CNN classifier trained on plasma data obtained F1 scores of 92.1% ± 1.4% for BS crossings and 84.7% ± 1.9% for MP crossings on a corrected test dataset (from use of a bootstrap method). Reliable automated detection of boundary crossings could enable future spacecraft experiments like the PEP instrument on the upcoming JUICE spacecraft mission to dynamically adapt the best observing mode based on rapid classification of the boundary crossings as soon as it appears, thus yielding higher quality data and improved potential for scientific discovery.

Offline Handwritten Signature Verification Using Deep Neural Networks

Prior to the implementation of digitisation processes, the handwritten signature in an attendance sheet was the preferred way to prove the presence of each student in a classroom. The method is still preferred, for example, for short courses or places where other methods are not implemented. However, human verification of handwritten signatures is a tedious process. The present work describes two methods for classifying signatures in an attendance sheet as valid or not. One method based on Optical Mark Recognition is general but determines only the presence or absence of a signature. The other method uses a multiclass convolutional neural network inspired by the AlexNet architecture and, after training with a few pieces of genuine training data, shows over 85% of precision and recall recognizing the author of the signatures. The use of data augmentation and a larger number of genuine signatures ensures higher accuracy in validating the signatures.

In‐line mura detection using machine learning and subspace method in display manufacturing

AbstractThis paper discusses the automatic detection of mura, non‐uniformity of brightness or color, which has been a long‐standing challenge in the display industries. Our purpose is to develop a method using machine learning, which automatically detects and classifies mura in the front‐end process. This will enable prompt feedback to the manufacturing process and contribute to improvement of the productivity. We propose “Progressive Hybrid model,” which is based on the human visual perception and consists of a multiclass CNN (Convolutional Neural Network), a 2‐class residual neural network, and a 2‐class CNN. The two 2‐class models based on the subspace method to reproduce the boundary‐samples in the human visible test are for accurate classification between Normal displays and weak mura. To reproduce the appropriate test‐images used in the human visible test, we enhance contrast of 8000 images of the dataset obtained from an OLED (Organic Light Emitting Diode) manufacturing line. The evaluation results show improvement of the accuracy from 0.830 in the general application of CNN to 0.884 in the proposed model. The proposed model is particularly effective to improve classification of Normal displays, which is usually the most numerous in the manufacturing, and thus further beneficial in the practical use.

A multiclass CNN cascade model for the clinical detection support of cardiac arrhythmia based on subject-exclusive ECG dataset

The accurate analysis of Electrocardiogram waveform plays a crucial role for supporting cardiologist in detecting and diagnosing the heartbeat disorders. To improve their detection accuracy, this work is devoted to the design of a novel classification algorithm which is composed of a cascade of two convolutional neural network (CNN), i.e a Binary CNN allowing the detection of the arrhythmic heartbeat and a Multiclass CNN able to recognize the specific disorder. Moreover, by combining the cascade architecture solution with a rule-based data splitting, which leverages the subject-exclusive and balances among the classes criteria, it is possible predicting the health status of unseen patients. Numerical results, carried out considering Massachusetts Institute of Technology-Beth Israel Hospital arrhythmia database, disclose a classification accuracy of 96.2%. Finally, a cross-database performance evaluation and a comparison analysis w.r.t. the current state-of-art further disclose the effectiveness and the efficiency of the proposed solution, as well as its benefits in terms of patient health status prediction.