Deep Convolutional Model Research Articles

Artificial intelligence (AI) vs. human in hip fracture detection

ObjectiveThis study aimed to assess the diagnostic accuracy and sensitivity of a YOLOv4-tiny AI model for detecting and classifying hip fractures types. Materials and methodsIn this retrospective study, a dataset of 1000 hip and pelvic radiographs was divided into a training set consisting of 450 fracture and 450 normal images (900 images total) and a testing set consisting of 50 fracture and 50 normal images (100 images total). The training set images were each manually augmented with a bounding box drawn around each hip, and each bounding box was manually labeled either (1) normal, (2) femoral neck fracture, (3) intertrochanteric fracture, or (4) subtrochanteric fracture. Next, a deep convolutional neural network YOLOv4-tiny AI model was trained using the augmented training set images, and then model performance was evaluated with the testing set images. Human doctors then evaluated the same testing set images, and the performances of the model and doctors were compared. The testing set contained no crossover data. ResultsThe resulting output images revealed that the AI model produced bounding boxes around each hip region and classified the fracture and normal hip regions with a sensitivity of 96.2%, specificity of 94.6%, and an accuracy of 95%. The human doctors performed with a sensitivity ranging from 69.2 to 96.2%. Compared with human doctors, the detection rate sensitivity of the model was significantly better than a general practitioner and first-year residents and equivalent to specialist doctors. ConclusionsThis model showed hip fracture detection sensitivity comparable to well-trained radiologists and orthopedists and classified hip fractures highly accurately.

Deep learning-based fully automated intelligent and precise diagnosis for melanocytic lesions

Melanocytic lesions occur on the surface of the skin, in which the malignant type is melanoma with a high fatality rate, seriously endangering human health. The histopathological analysis is the gold standard for diagnosis of melanocytic lesions. In this study, a fully automated intelligent diagnosis method based on deep learning was proposed to classify the pathological whole slide images (WSI) of melanocytic lesions. Firstly, the color normalization based on CycleGAN neural network was performed on multi-center pathological WSI; Secondly, ResNet-152 neural network-based deep convolutional network prediction model was built using 745 WSI; Then, a decision fusion model was cascaded, which calculates the average prediction probability of each WSI; Finally, the diagnostic performance of the proposed method was verified by internal and external test sets containing 182 and 54 WSI, respectively. Experimental results showed that the overall diagnostic accuracy of the proposed method reached 94.12% in the internal test set and exceeded 90% in the external test set. Furthermore, the color normalization method adopted was superior to the traditional color statistics-based and staining separation-based methods in terms of structure preservation and artifact suppression. The results demonstrate that the proposed method can achieve high precision and strong robustness in pathological WSI classification of melanocytic lesions, which has the potential in promoting the clinical application of computer-aided pathological diagnosis.

An intelligent driven deep residual learning framework for brain tumor classification using MRI images

Brain tumor classification is an expensive complicated challenge in the sector of clinical image analysis. Machine learning algorithms enabled radiologists to accurately diagnose tumors without requiring major surgery. However, several challenges rise; first, the major challenge in designing the most accurate deep learning architecture for classifying brain tumors; and secondly, difficulty of finding an expert who is experienced in the field of classifying brain tumors using images by deep learning models. These difficulties made us motivated to propose an advanced and high accurate framework based on the concepts of deep learning and evolutionary algorithms to automatically design the ResNet architecture efficiently for classifying three types of brain tumors on a large database of MRI images. Thus, we propose an optimization-based deep convolutional ResNet model combined with a novel evolutionary algorithm to optimize the architecture and hyperparameters of deep ResNet model automatically without need of human experts as well manual architecture design which is complicated task to classify different types of brain tumor. Also, we propose an improved version of ant colony optimization (IACO) based on the concepts of differential evolution strategy and multi-population operators. These two concepts make an effective balance for solution diversity and convergence speed as well as enhancing the optimization performance and avoiding falling into the local optima for designing the deep learning-based ResNet architectures. The experimental finding revealed that our proposed framework obtained an average accuracy of 0.98694 which efficiently shows that our IACO-ResNet algorithm can help excellently with the automatic classification of brain tumors.

A novel denoising method for low SNR NMR logging echo signal based on deep learning

The T 2 spectrum obtained by inversion of the nuclear magnetic resonance logging echo signal can provide various petrophysical parameters to help technicians effectively identify the type of fluid. However, raw logging data with a low signal-to-noise ratio can cause inversion results to deviate from the truth. Therefore, a deep learning denoising method is proposed to eliminate the limitations of traditional mathematical transform methods. A one-dimensional deep convolutional generative adversarial networks model is designed to fit the noise distribution of actual logging data, then generate simulation data. A multi-scale echo denoising network (MsEDNet) is designed to adaptively learn the multi-exponential decay characteristics of the signal and the optimal transform space. The denoising dataset consists of simulation data and logging data, which is applied to train MsEDNet and improve the generalization performance. With effectiveness analysis and various denoising experiments, it is validated that the proposed method has excellent denoising performance on simulation data, logging data, and water tank data.

Equilibrium-based COVID-19 diagnosis from routine blood tests: A sparse deep convolutional model

SARS-CoV2 (COVID-19) is the virus that causes the pandemic that has severely impacted human society with a massive death toll worldwide. Hence, there is a persistent need for fast and reliable automatic tools to help health teams in making clinical decisions. Predictive models could potentially ease the strain on healthcare systems by early and reliable screening of COVID-19 patients which helps to combat the spread of the disease. Recent studies have reported some key advantages of employing routine blood tests for initial screening of COVID-19 patients. Thus, in this paper, we propose a novel COVID-19 prediction model based on routine blood tests. In this model, we depend on exploiting the real dependency among the employed feature pool by a sparsification procedure. In this sparse domain, a hybrid feature selection mechanism is proposed. This mechanism fuses the selected features from two perspectives, the first is Pearson correlation and the second is a new Minkowski-based equilibrium optimizer (MEO). Then, the selected features are fed into a new 1D Convolutional Neural Network (1DCNN) for a final diagnosis decision. The proposed prediction model is tested with a new public dataset from San Raphael Hospital, Milan, Italy, i.e., OSR dataset which has two sub-datasets. According to the experimental results, the proposed model outperforms the state-of-the-art techniques with an average testing accuracy of 98.5% while we employ only less than half the size of the feature pool, i.e., we need only less than half the given blood tests in the employed dataset to get a final diagnosis decision.

Satelight: self-attention-based model for epileptic spike detection from multi-electrode EEG

Objective. Because of the lack of highly skilled experts, automated technologies that support electroencephalogram (EEG)-based in epilepsy diagnosis are advancing. Deep convolutional neural network-based models have been used successfully for detecting epileptic spikes, one of the biomarkers, from EEG. However, a sizeable number of supervised EEG records are required for training. Approach. This study introduces the Satelight model, which uses the self-attention (SA) mechanism. The model was trained using a clinical EEG dataset labeled by five specialists, including 16 008 epileptic spikes and 15 478 artifacts from 50 children. The SA mechanism is expected to reduce the number of parameters and efficiently extract features from a small amount of EEG data. To validate the effectiveness, we compared various spike detection approaches with the clinical EEG data. Main results. The experimental results showed that the proposed method detected epileptic spikes more effectively than other models (accuracy = 0.876 and false positive rate = 0.133). Significance. The proposed model had only one-tenth the number of parameters as the other effective model, despite having such a high detection performance. Further exploration of the hidden parameters revealed that the model automatically attended to the EEG’s characteristic waveform locations of interest.

Harmonized system code prediction of import and export commodities based on Hybrid Convolutional Neural Network with Auxiliary Network

China Customs mainly uses manual inspections on the tax rates of import and export commodities, which can only cover a small part of the mass of commodities. Therefore, we investigate the natural language processing technology to determine the tax rate automatically by commodities classification. However, the unique challenge is that the structured commodity text is ambiguous, and has no continuous context and semantics, leading to difficulties for classification. In light of this challenge, we draw on the idea of the deep pyramid convolutional model and propose a Shallow Structured Convolutional Neural Network (SSCNN) with an Auxiliary Network to reduce the semantic fusion in commodity classification. When extracting shallow features, our model uses a structural token to fill in the feature boundary of structured text to prevent the feature fusion problem of adjacent features brought by the convolution operation. Auxiliary Network learns the distinguishing features of each commodity category and integrates the customs-specific knowledge to improve the classification performance of similar goods. In the empirical study on a real-world customs dataset, our model outperforms the mainstream deep learning methods including Transformer, BERT, BART and RoBERTa, which verifies the effectiveness of this method on classifying import and export commodities.

LMix: regularization strategy for convolutional neural networks

Deep convolutional neural networks perform well in the field of computer vision, but exhibit undesirable behaviors such as memorization and sensitivity to adversarial examples. Therefore, proper regularization strategies are needed to alleviate these problems. Currently, regularization strategies with mixed sample data augmentation perform very well, and these algorithms allow the network to generalize better, improve the baseline performance of the model. However, interpolation-based mixed sample data augmentation distorts the data distribution, while masking-based mixed sample data augmentation results in excessive information loss for overly regular shapes of masks. Although mixed sample data augmentation is proven to be an effective method to improve the baseline performance, generalization ability and robustness of deep convolutional models, there is still room for improvement in terms of maintaining the of image local consistency and image data distribution. In this paper, we propose a new mixed sample data augmentation-LMix, which uses random masking to increase the number of masks in the image to maintain the data distribution, and high-frequency filtering to sharpen the image to highlight recognition regions. We applied the method to train CIFAR-10, CIFAR-100, SVHN, and Tiny-ImageNet datasets under the PreAct-ResNet18 model to evaluate the method, and obtained the latest results of 96.32, 79.85, 97.01, and 64.16%, respectively, which are 1.70, 4.73, and 8.06% higher than the optimal baseline accuracy. The LMix algorithm improves the generalization ability of the state-of-the-art neural network architecture and enhances the robustness to adversarial samples.

Color Matching Generation Algorithm for Animation Characters Based on Convolutional Neural Network.

In recent years, for China, animation industry is a relatively new and mature emerging national sunrise industry after animation industry, which appears on the world stage more and more frequently and is widely concerned and valued by people from all over the world. Therefore, this paper innovatively uses the convolutional neural network algorithm to innovate the color matching generation of animation characters and improve the traditional technology of color matching for animation characters. In this paper, we mainly use Generative Adversarial Network (GAN), Deep Convolutional Generative Adversarial Network and VGG model, and multiscale discriminator theory and use ACGAN research method. And we study this paper's innovative LMV-ACGAN research method, and we have come to the conclusion that other models have higher collapse rate than this model; this model has higher color matching of anime characters. Color matching improves with the increase of convolutional neural network utilization, etc. Moreover, superior and minor reviews of this study are provided to make later researchers understand this study more rationally and objectively.

Action Recognition Model in Choral Conducting Teaching in Colleges and Universities under New Media Environment.

In current college music education, choral conducting is a required course for students. The course implementation aims to cultivate excellent and high-quality choral conductors. The requirements for choral conducting teaching in college music education under the new media environment have been further improved. First, this study gives the value of applying new media technology in choral conducting teaching in colleges and universities. Then, based on the key point that choral conductors' expression of music mainly relies on gestural language, an action recognition model in college choral conducting teaching is proposed. The model is designed with an adaptive deep graph convolution model, and a spatio-temporal convolution submodel with a small number of parameters is created using group convolution. After the trained teacher model is obtained, the spatio-temporal convolutional submodel with fewer parameters is trained using the knowledge distillation method combined with data augmentation techniques. The final action recognition fusion model is obtained using the linear fusion method. The experimental results demonstrate that the proposed model can recognize the movements in college choral conducting teaching with higher performance than other existing models, which provides effective guidance for college choral conducting teaching in the new media environment.

Dense open-set recognition based on training with noisy negative images

Deep convolutional models often produce inadequate predictions for inputs which are foreign to the training distribution. Consequently, the problem of detecting outlier images has recently been receiving a lot of attention. Unlike most previous work, we address this problem in the dense prediction context. Our approach is based on two reasonable assumptions. First, we assume that the inlier dataset is related to some narrow application field (e.g. road driving). Second, we assume that there exists a general-purpose dataset which is much more diverse than the inlier dataset (e.g. ImageNet-1 k). We consider pixels from the general-purpose dataset as noisy negative samples since most (but not all) of them are outliers. We encourage the model to recognize borders between the known and the unknown by pasting jittered negative patches over inlier training images. Our experiments target two dense open-set recognition benchmarks (WildDash 1 and Fishyscapes) and one dense open-set recognition dataset (StreetHazard). Extensive performance evaluation indicates competitive potential of the proposed approach. • Training with noisy negative images greatly improves dense open-set recognition. • Training with randomly pasted negatives improves generalization on mixed-content images. • Shared features improve outlier detection and allow for inference with a single forward pass. • State-of-the-art results on dense open-set recognition benchmarks: WildDash 1, Fishyscapes and StreetHazard.

MSFF-CDCGAN: A novel method to predict RNA secondary structure based on Generative Adversarial Network.

Access to RNA secondary structure is a prerequisite for understanding and mastering RNA function. RNA secondary structures play an important role in cells, they can cause or contribute to neurological disorders and can be applied in the medical field. However, the experimental method to obtain RNA secondary structure is costly, laborious and not universal. Although computational methods can predict RNA secondary structure more accurately for short-sequence RNAs, it cannot predict long-sequence RNAs and pseudoknot, which is the bottleneck of RNA secondary structure prediction at present. In recent years, researchers have attempted to use deep learning algorithms to predict RNA secondary structure and have achieved results. However, the small amount of data on the secondary structure of long-sequence RNAs leads to the low accuracy of deep learning methods to predict the secondary structure of RNAs across races. Similarly, RNA structure with pseudoknot is very complex and insufficient data caused the deep learning algorithm to struggle to predict the secondary structure of RNA containing pseudoknots. The RNA data are encoded into grayscale images by a unique encoding method based on the real RNA secondary structure and sequence information. Then, this paper reasonably expands the image data to increase the amount of RNA data, solves the problem of insufficient data for predicting long sequences and RNA secondary structure with pseudoknots in current deep learning methods, and provides a good data foundation for deep learning.The article proposes a multi-scale feature fusion Conditional Deep Convolutional Generative Adversarial Network prediction model (MSFF-CDCGAN) based on the improved Conditional Deep Convolutional Generative Adversarial Network (CDCGAN) model to predict RNA secondary structure. The experimental results showed that the MSFF-CDCGAN model could predict long-sequence RNAs and pseudoknots more accurately than traditional prediction methods. This paper introduces Generative Adversarial Network (GAN) to RNA secondary structure prediction for the first time. It uses a unique image encoding approach to expand the original RNA data set, thus transforming the structure prediction problem into an image analysis problem and effectively solving the bottleneck in RNA secondary structure prediction.

Olive Disease Classification Based on Vision Transformer and CNN Models.

It has been noted that disease detection approaches based on deep learning are becoming increasingly important in artificial intelligence-based research in the field of agriculture. Studies conducted in this area are not at the level that is desirable due to the diversity of plant species and the regional characteristics of many of these species. Although numerous researchers have studied diseases on plant leaves, it is undeniable that timely diagnosis of diseases on olive leaves remains a difficult task. It is estimated that people have been cultivating olive trees for 6000 years, making it one of the most useful and profitable fruit trees in history. Symptoms that appear on infected leaves can vary from one plant to another or even between individual leaves on the same plant. Because olive groves are susceptible to a variety of pathogens, including bacterial blight, olive knot, Aculus olearius, and olive peacock spot, it has been difficult to develop an effective olive disease detection algorithm. For this reason, we developed a unique deep ensemble learning strategy that combines the convolutional neural network model with vision transformer model. The goal of this method is to detect and classify diseases that can affect olive leaves. In addition, binary and multiclassification systems based on deep convolutional models were used to categorize olive leaf disease. The results are encouraging and show how effectively CNN and vision transformer models can be used together. Our model outperformed the other models with an accuracy of about 96% for multiclass classification and 97% for binary classification, as shown by the experimental results reported in this study.

3D deep convolutional neural network segmentation model for precipitate and porosity identification in synchrotron X-ray tomograms.

New developments at synchrotron beamlines and the ongoing upgrades of synchrotron facilities allow the possibility to study complex structures with a much better spatial and temporal resolution than ever before. However, the downside is that the data collected are also significantly larger (more than several terabytes) than ever before, and post-processing and analyzing these data is very challenging to perform manually. This issue can be solved by employing automated methods such as machine learning, which show significantly improved performance in data processing and image segmentation than manual methods. In this work, a 3D U-net deep convolutional neural network (DCNN) model with four layers and base-8 characteristic features has been developed to segment precipitates and porosities in synchrotron transmission X-ray micrograms. Transmission X-ray microscopy experiments were conducted on micropillars prepared from additively manufactured 316L steel to evaluate precipitate information. After training the 3D U-net DCNN model, it was used on unseen data and the prediction was compared with manual segmentation. A good agreement was found between both segmentations. An ablation study was performed and revealed that the proposed model showed better statistics than other models with lower numbers of layers and/or characteristic features. The proposed model is able to segment several hundreds of gigabytes of data in a few minutes and could be applied to other materials and tomography techniques. The code and the fitted weights are made available with this paper for any interested researcher to use for their needs (https://github.com/manasvupadhyay/erc-gamma-3D-DCNN).

Deep Learning Model for Prediction of Diffusion in Defect Substances

Actual diffusion activity function is an important metric utilized to describe the diffusion activities of a vacancy defect substance. In this paper, we propose a deep learning three-dimensional convolutional CNN model (D3-CNN). A 3D convolution has its kernel slides in three dimensions as opposed to two dimensions with 2D convolutions. 3D convolution is more suitable for three-dimensional data. We also propose an amplification learning technique to predict the actual diffusion activity of a vacancy defect substance, which is impacted by the geometrical parameters of the defect substance and the vacancy distribution function. In this model, the geometric parameters of a three-dimensional constructed vacancy defect substance are generated. The 3D dataset is obtained by the atoms diffusion defect (ADD) simulation model. The geometric parameters of the 3D vacancy defect substance are computed by the proposed amplification technique. The 3D geometric parameters and the diffusion activity values are applied to a deep learning model for training. The actual diffusion activity values of a substance with a vacancy size ranging from size 0.52 mm to 0.61 mm are used for training, and the actual diffusion activity values of substance vacancy of size between 0.41 and 1.01 are classified by the three-dimensional network. The model can realize high speed and accuracy for the actual diffusion activity value. The mean relative absolute errors between the D3-CNN and the ADD models are 0.028–7.85% with a vacancy size of 0.41 to 0.81. For a usual sample with a vacancy of size equal to 0.6, the CPU computation load required by our model is 14.2 × 10−2 h, while the time required is 15.16 h for the ADD model. These results indicate that our proposed deep learning model has a strong learning capability and can function as an influential model to classify the diffusion activity of compound vacancy defect substances.

Detecting the sources of chemicals in the Black Sea using non-target screening and deep learning convolutional neural networks

The Black Sea is an important ecosystem, which is affected by various anthropogenic pressures, such as shipping activities and wastewater inputs from large coastal cities. Significant loads of chemical pollutants are being continuously brought in by major European rivers. This study investigated the spatial distribution of chemicals in the Ukrainian shelf (the northwestern part of the Black Sea) and their main sources. Chemical occurrence data used in the study was generated within the Joint Black Sea Surveys (JBSS), which took place in 2016 and 2017 as a part of the EU/UNDP EMBLAS II project (www.emblasproject.org). During the JBSS, seawater samples were analyzed by a non-target screening workflow using liquid chromatography high-resolution mass spectrometry (LC-HRMS). Open-source algorithms were applied to generate a combined dataset of 30,489 detected chemical signals and their intensities. Out of these, 35 compounds were tentatively identified by the application of a non-target screening identification workflow based on automated matching of their mass spectra against those in available mass spectral libraries. The dataset was used to generate images, representing spatial distribution of each of the signals. These images were then used as an input to a deep learning convolutional neural network classification model. The study resulted in the development of an open-source end-to-end workflow for the estimation of the pollution load by chemicals contributed by the two major inflowing rivers (Danube and Dnieper) and other, so far unidentified, sources. A dedicated dashboard was built to facilitate data visualization per detected signal/compound. The presented model proved to be especially useful at the prioritization of signals of unknown compounds, which is of key importance for the follow up structure elucidation efforts of bulky non-target screening data. The deep learning approach for peak prioritization of unknown chemicals in the environment has been used for the first time.

Driver drowsiness detection in video sequences using hybrid selection of deep features

Monitoring driver’s drowsiness is a complex problem that involves many indicators whether behavioral or physiological. Drowsiness is a challenging problem that can lead to road disasters. Sleeping driver is more dangerous on the road than a speeding driver. Many statistics showed that one-fifth of road accidents in the world were due to driver fatigue, hence safety modules that can alert drowsy drivers in the hopes of reducing the risk of accidents are very important. This paper proposes a framework for driver drowsiness detection based on a computer vision solution. The proposed framework’s first task is to detect the driver’s face. A transfer learning is then performed for extracting the deep features from the driver’s face image using a pre-trained deep convolutional network model trained on a facial recognition dataset. The previous tasks are applied in a sliding temporal window (less than a second) in which the frames are sampled. In this work, 9 frames were the best choice. The extracted features of these frames represent the observation matrix. Then temporal feature aggregation is applied to construct the raw feature vector. To obtain the final feature vector, a proposed feature selection is applied to omit possible irrelevant features. The final feature vector is finally fed to a binary classifier to decide whether there is drowsiness or not. Extensive experiments are applied to NTHU Drowsy Driver Detection (NTHU-DDD) video dataset. The outcomes show the outperformance of the proposed approach compared with the state-of-the-art approaches.

Dyskinesia Estimation of Imbalanced Data Using a Deep-Learning Model.

The collection of Parkinson's Disease (PD) time-series data usually results in imbalanced and incomplete datasets due to the geometric distribution of PD complications' sever-ity scores. Consequently, when training deep convolutional models on these datasets, the models suffer from overfitting and lack generalizability to unseen data. In this paper, we investigated a new framework of Conditional Generative Ad-versarial Netuwoks (cGANs) as a solution to improve the extrapolation and generalizability of the regression models in such datasets. We used a real-world PD dataset to esti-mate Dyskinesia severity in patients with PD. The developed cGAN demonstrated significantly better generalizability to unseen data samples than a traditional Convolutional Neural Network with an improvement of 34%. This solution can be applied in similar imbalanced time-series data, especially in the healthcare domain, where balanced and uniformly distributed data samples are not readily available.

Coronavirus Detection using Deep Pre-trained Model on Chest X-ray Imaging

The constant increase in the number of patients suffering from the pandemic Covid-19 has overpowered the medical management system over the world. This pandemic has affected every aspect of human life all over the world. However, the limited number of testing kits is making it more challenging to deal with the global pandemic. It is near to impossible for people to maintain social distancing all the time, and lengthy testing procedure is one of the major cause of the rapid spread of the epidemic Covid-19. The polymerase chain reaction test that is done in real-time (RT-PCR) is time-consuming for each patient with gasping disease to be tested. However, the use of a chest x-ray machine tool can accelerate testing time. X-ray machines are now accessible in almost every medical management healthcare system, with instant results. The proposed research aims to detect the pandemic. The dataset was collected from an open-source repository that consists of the chest x-ray images of patients who were Covid-19 positive, and covid-19 negative which means not infected by the pandemic coronavirus but suffering from normal pneumonia. The approach of this paper uses a deep convolutional neural network model i-e VGG16, to begin the process of classifying images into positive or negative. In the proposed research VGG16, the Deep Convolutional Network model is used, this is a pre-trained model, previously trained on the ImageNet dataset that com-prises 14M images that belong to 1K different classes. This model is fine-tuned on 1202 images. The preprocessing of collected data is done by converting it into an RGB channel and resizing it to match the requirement of our model i-e VGG16. Data augmentation is performed on the preprocessed dataset. Our proposed model gives 99.1%, 100%, and 98.59% accuracies, sensitivity specificity respectively for the detection of Covid-19 on patients’ cxr images.

Kushner-Stratonovich Dice Segmented Curvelet (KSDSC) deep convolutional neural learning for heart disease prediction

Data mining is commonly used method for processing large amount of data in heart disease prediction. Many heart disease prediction researches are carried out by different authors. But the accuracy level was not improved. In order to address these issues, Kushner-Stratonovich Dice Segmented Haar Wavelet (KSDSC) Deep Convolutional Neural Learning Model is introduced. The main aim of KSDSC Model is to perform efficient heart disease prediction with five layers, namely one input layer, three hidden layers and one output layer. Initially, ultrasound images are collected as an input at input layer. An image pre-processing is carried out using Kushner-Stratonovich Filter to eliminate the noisy pixels from US image and transmitted to hidden layer 2. Sørensen–Dice Image Segmentation Process partitions the preprocessed image into number of divisions in hidden layer 2. After that in hidden layer 3, the multiple features are extracted using Curvelet transform from segmented image and sent to the output layer. The output layer uses softmax yolov4 darknet53 activation function to match extracted features for heart disease prediction. Experimental analysis is performed on parameters such as prediction accuracy, false positive rate, and prediction time with respect to a number of US images.