Deep Learning-based Classification Research Articles

Deep Learning-Based Classification of Inflammatory Arthritis by Identification of Joint Shape Patterns-How Neural Networks Can Tell Us Where to "Deep Dive" Clinically.

Objective:We investigated whether a neural network based on the shape of joints can differentiate between rheumatoid arthritis (RA), psoriatic arthritis (PsA), and healthy controls (HC), which class patients with undifferentiated arthritis (UA) are assigned to, and whether this neural network is able to identify disease-specific regions in joints.MethodsWe trained a novel neural network on 3D articular bone shapes of hand joints of RA and PsA patients as well as HC. Bone shapes were created from high-resolution peripheral-computed-tomography (HR-pQCT) data of the second metacarpal bone head. Heat maps of critical spots were generated using GradCAM. After training, we fed shape patterns of UA into the neural network to classify them into RA, PsA, or HC.ResultsHand bone shapes from 932 HR-pQCT scans of 617 patients were available. The network could differentiate the classes with an area-under-receiver-operator-curve of 82% for HC, 75% for RA, and 68% for PsA. Heat maps identified anatomical regions such as bare area or ligament attachments prone to erosions and bony spurs. When feeding UA data into the neural network, 86% were classified as “RA,” 11% as “PsA,” and 3% as “HC” based on the joint shape.ConclusionWe investigated neural networks to differentiate the shape of joints of RA, PsA, and HC and extracted disease-specific characteristics as heat maps on 3D joint shapes that can be utilized in clinical routine examination using ultrasound. Finally, unspecific diseases such as UA could be grouped using the trained network based on joint shape.

Deep Learning-Based Phenotypic Assessment of Red Cell Storage Lesions for Safe Transfusions.

This study presents a novel approach to automatically perform instant phenotypic assessment of red blood cell (RBC) storage lesion in phase images obtained by digital holographic microscopy. The proposed model combines a generative adversarial network (GAN) with marker-controlled watershed segmentation scheme. The GAN model performed RBC segmentations and classifications to develop ageing markers, and the watershed segmentation was used to completely separate overlapping RBCs. Our approach achieved good segmentation and classification accuracy with a Dice's coefficient of 0.94 at a high throughput rate of about 152 cells per second. These results were compared with other deep neural network architectures. Moreover, our image-based deep learning models recognized the morphological changes that occur in RBCs during storage. Our deep learning-based classification results were in good agreement with previous findings on the changes in RBC markers (dominant shapes) affected by storage duration. We believe that our image-based deep learning models can be useful for automated assessment of RBC quality, storage lesions for safe transfusions, and diagnosis of RBC-related diseases.

Heart Disease Detection Based on Feature Fusion Technique with Augmented Classification Using Deep Learning Technology

An accurate prediction of cardiac disease is a crucial task for medical and research organizations. Cardiac patients are usually facing heart attacks sometimes tends to death. Therefore, a prior stage of heart diagnosis is compulsory, so that model of optimal Deep learning technology is prosperous for the healthcare sector. The earlier models related to this research work are outdated, some applications cannot provide efficient outcomes. The available conventional models like the Genetic algorithm (GA), PSO (particle swarm optimization), RFO (Random Forest optimization), X-boosting. KNN and many available technologies are only dispensing abnormality information but they are not providing location, depth, and affected area dimensions. Moreover, earlier models only supported fixed scanning in radiology not supporting cloud-level deployment. The sensitivity and robustness of diagnosis are very low therefore a DCAlexNet CNN deep learning technology is needed. The deep learning-based classification is performed through the DCAlexNet CNN (convolutional Neural networks) technique. The implementing application is loading training samples from Kaggle or ANDI dataset. The uploaded image samples are pre-processed through resolution, intensity, and brightness adjustment in the python NumPy tool. The. CSV file (text file) is processed through clustering as well as dimensionality adjusting technique. The processed images are segmented through RRF (Restrictive Random Field) technology. The segmentation on images provides features that are loaded in the local server after that saved into CNN memory. Now the .csv file and trained features are applied to DCAlexNet CNN deep learning architecture. The training processing can give information about diseases in the heart and dimensionality of the affected area (depth and location). Now the application is waiting for real-time samples which is nothing but testing, in this testing part locally available affected and healthy heart ultrasound images are given to DCAlexNet CNN. The designed application can easily be identified whether the uploaded image has abnormality or not. The test-based and image-oriented feature fusion can help the application detect heart abnormalities in an easy way. To this feature fusion-based DCAlexNet CNN confusion matrix generates performance measures like accuracy 98.67%, sensitivity 97.45%, Recall 99.34%, and F1 Score 99.34%, these numerical comparison results compete with present technology and outperformance application robustness.

AOS: Anti-Obesity System With Deep Learning-based Classification Model Using a Novel Data Augmentation Technique

Obesity has been a worldwide leading cause of associated health risks for decades, yet not much is being done to solve this issue. Previously, minimal efforts have been made to address this issue in regards to mandating nutrition labels on the back of packaged foods. However, the mundane and growingly elusive packing labels have diminished the proper effects of this constitution as people rarely ever take the time nor effort to thoroughly read through the labels. Moreover, most dishes served from restaurants do not serve with any nutritional information, which places many of the people dining out in complete darkness of what they are truly consuming. To solve this problem, we propose the deep learning-based AOS (Anti-Obesity System), which analyzes images of common junk foods and unhealthy meals to classify the category of food. The proposed system consists of a deep learning-based classifier and a post-processing module that outputs relevant nutritional information regarding each category of food. In addition, we propose a novel data augmentation technique in order to make the trained model produce better results. We also conduct the ablation study to experimentally prove that the proposed method enforces the model to work better in real-world situations. We achieve an AP of 59.30% on the proposed food dataset. Code will be available at https://github.com/smkim0508/ObesityNutritionClassifier.git.

Enhancing deep learning based classifiers with inpainting anatomical side markers (L/R markers) for multi-center trials

Background and objectiveThe protocol for placing anatomical side markers (L/R markers) in chest radiographs varies from one hospital or department to another. However, the markers have strong signals that can be useful for deep learning-based classifier to predict diseases. We aimed to enhance the performance of a deep learning-based classifiers in multi-center datasets by inpainting the L/R markers. MethodsThe L/R marker was detected with using the EfficientDet detection network; only the detected regions were inpainted using a generative adversarial network (GAN). To analyze the effect of the inpainting in detail, deep learning-based classifiers were trained using original images, marker-inpainted images, and original images clipped using the min-max value of the marker-inpainted images. Binary classification, multi-class classification, and multi-task learning with segmentation and classification were developed and evaluated. Furthermore, the performances of the network on internal and external validation datasets were compared using DeLong's test for two correlated receiver operating characteristic (ROC) curves in binary classification and Stuart-Maxwell test for marginal homogeneity in multi-class classification and multi-task learning. In addition, the qualitative results of activation maps were evaluated using the gradient-class activation map (Grad-CAM). ResultsMarker-inpainting preprocessing improved the classification performances. In the binary classification based on the internal validation, the area under the curves (AUCs) and accuracies were 0.950 and 0.900 for the model trained on the min-max clipped images and 0.911 and 0.850 for the model trained on the original images, respectively (P-value=0.006). In the external validation, the AUCs and accuracies were 0.858 and 0.677 for the model trained using the inpainted images and 0.723 and 0.568 for the model trained using the original images (P-value<0.001), respectively. In addition, the models trained using the marker inpainted images showed the best performance in multi-class classification and multi-task learning. Furthermore, the activation maps obtained using the Grad-CAM improved with the proposed method. The 5-fold validation results also showed improvement trend according to the preprocessing strategies. ConclusionsInpainting an L/R marker significantly enhanced the classifier's performance and robustness, especially in internal and external studies, which could be useful in developing a more robust and accurate deep learning-based classifier for multi-center trials. The code for detection is available at: https://github.com/mi2rl/MI2RLNet. And the code for inpainting is available at: https://github.com/mi2rl/L-R-marker-inpainting.

A deep learning approach for medical waste classification

As the demand for health grows, the increase in medical waste generation is gradually outstripping the load. In this paper, we propose a deep learning approach for identification and classification of medical waste. Deep learning is currently the most popular technique in image classification, but its need for large amounts of data limits its usage. In this scenario, we propose a deep learning-based classification method, in which ResNeXt is a suitable deep neural network for practical implementation, followed by transfer learning methods to improve classification results. We pay special attention to the problem of medical waste classification, which needs to be solved urgently in the current environmental protection context. We applied the technique to 3480 images and succeeded in correctly identifying 8 kinds of medical waste with an accuracy of 97.2%; the average F1-score of five-fold cross-validation was 97.2%. This study provided a deep learning-based method for automatic detection and classification of 8 kinds of medical waste with high accuracy and average precision. We believe that the power of artificial intelligence could be harnessed in products that would facilitate medical waste classification and could become widely available throughout China.

Assessing central serous chorioretinopathy with deep learning and multiple optical coherence tomography images

Central serous chorioretinopathy (CSC) is one of the most common macular diseases that can reduce the quality of life of patients. This study aimed to build a deep learning-based classification model using multiple spectral domain optical coherence tomography (SD-OCT) images together to diagnose CSC. Our proposed system contains two modules: single-image prediction (SIP) and a final decision (FD) classifier. A total of 7425 SD-OCT images from 297 participants (109 acute CSC, 106 chronic CSC, 82 normal) were included. In the fivefold cross validation test, our model showed an average accuracy of 94.2%. Compared to other end-to-end models, for example, a 3D convolutional neural network (CNN) model and a CNN-long short-term memory (CNN-LSTM) model, the proposed system showed more than 10% higher accuracy. In the experiments comparing the proposed model and ophthalmologists, our model showed higher accuracy than experts in distinguishing between acute, chronic, and normal cases. Our results show that an automated deep learning-based model could play a supplementary role alongside ophthalmologists in the diagnosis and management of CSC. In particular, the proposed model seems clinically applicable because it can classify CSCs using multiple OCT images simultaneously.

Deep learning-based classification of retinal vascular diseases using ultra-widefield colour fundus photographs

ObjectiveTo assess the ability of a deep learning model to distinguish between diabetic retinopathy (DR), sickle cell retinopathy (SCR), retinal vein occlusions (RVOs) and healthy eyes using ultra-widefield colour fundus...

Explicit kissing scene detection in cartoon using convolutional long short-term memory

The main concern of this study is due to certain cartoon content consisting of explicit scenes such as kissing, sex, violence. That are somehow not suitable for kids and may contradict to some religions and cultures. There are some reasons the film industry does not expel the kissing scene in a cartoon movie. It is categorized as a romance sequence and love scene. These could be a double-edged weapon that will ruin an individual’s childhood through excessive exposure to explicit content. This paper proposes a deep learning-based classifier to detect the kissing scene in the cartoon by using Darknet-19 for frame-level feature extraction, while the feature aggregation in the temporal domain is using convolutional long short-term memory (conv-LSTM). This paper also has discussed a few steps related to evaluation and analysis regarding the performance of the models. Extensive experiments prove that the proposed system provides excellent results of 96.43% accuracy to detect the kissing scene in the cartoon. Due to high accuracy performance, the model is suitable to be a kissing scene filter feature in a digital video player that may able to decrease the excessive exposure to explicit content for kids.

The applications of deep learning algorithms on in silico druggable proteins identification

IntroductionThe top priority in drug development is to identify novel and effective drug targets. In vitro assays are frequently used for this purpose; however, traditional experimental approaches are insufficient for large-scale exploration of novel drug targets, as they are expensive, time-consuming and laborious. Therefore, computational methods have emerged in recent decades as an alternative to aid experimental drug discovery studies by developing sophisticated predictive models to estimate unknown drugs/compounds and their targets. The recent success of deep learning (DL) techniques in machine learning and artificial intelligence has further attracted a great deal of attention in the biomedicine field, including computational drug discovery. ObjectivesThis study focuses on the practical applications of deep learning algorithms for predicting druggable proteins and proposes a powerful predictor for fast and accurate identification of potential drug targets. MethodsUsing a gold-standard dataset, we explored several typical protein features and different deep learning algorithms and evaluated their performance in a comprehensive way. We provide an overview of the entire experimental process, including protein features and descriptors, neural network architectures, libraries and toolkits for deep learning modelling, performance evaluation metrics, model interpretation and visualization. ResultsExperimental results show that the hybrid model (architecture: CNN-RNN (BiLSTM) + DNN; feature: dictionary encoding + DC_TC_CTD) performed better than the other models on the benchmark dataset. This hybrid model was able to achieve 90.0% accuracy and 0.800 MCC on the test dataset and 84.8% and 0.703 on a nonredundant independent test dataset, which is comparable to those of existing methods. ConclusionWe developed the first deep learning-based classifier for fast and accurate identification of potential druggable proteins. We hope that this study will be helpful for future researchers who would like to use deep learning techniques to develop relevant predictive models.

Crowdsourcing for Machine Learning in Public Health Surveillance: Lessons Learned From Amazon Mechanical Turk.

BackgroundCrowdsourcing services, such as Amazon Mechanical Turk (AMT), allow researchers to use the collective intelligence of a wide range of web users for labor-intensive tasks. As the manual verification of the quality of the collected results is difficult because of the large volume of data and the quick turnaround time of the process, many questions remain to be explored regarding the reliability of these resources for developing digital public health systems.ObjectiveThis study aims to explore and evaluate the application of crowdsourcing, generally, and AMT, specifically, for developing digital public health surveillance systems.MethodsWe collected 296,166 crowd-generated labels for 98,722 tweets, labeled by 610 AMT workers, to develop machine learning (ML) models for detecting behaviors related to physical activity, sedentary behavior, and sleep quality among Twitter users. To infer the ground truth labels and explore the quality of these labels, we studied 4 statistical consensus methods that are agnostic of task features and only focus on worker labeling behavior. Moreover, to model the meta-information associated with each labeling task and leverage the potential of context-sensitive data in the truth inference process, we developed 7 ML models, including traditional classifiers (offline and active), a deep learning–based classification model, and a hybrid convolutional neural network model.ResultsAlthough most crowdsourcing-based studies in public health have often equated majority vote with quality, the results of our study using a truth set of 9000 manually labeled tweets showed that consensus-based inference models mask underlying uncertainty in data and overlook the importance of task meta-information. Our evaluations across 3 physical activity, sedentary behavior, and sleep quality data sets showed that truth inference is a context-sensitive process, and none of the methods studied in this paper were consistently superior to others in predicting the truth label. We also found that the performance of the ML models trained on crowd-labeled data was sensitive to the quality of these labels, and poor-quality labels led to incorrect assessment of these models. Finally, we have provided a set of practical recommendations to improve the quality and reliability of crowdsourced data.ConclusionsOur findings indicate the importance of the quality of crowd-generated labels in developing ML models designed for decision-making purposes, such as public health surveillance decisions. A combination of inference models outlined and analyzed in this study could be used to quantitatively measure and improve the quality of crowd-generated labels for training ML models.

Design of robust deep learning-based object detection and classification model for autonomous driving applications

Design of robust deep learning-based object detection and classification model for autonomous driving applications

Estimation of pavement crack ratio by top-view transformation of in-vehicle smartphone camera and deep learning-based classification

Estimation of pavement crack ratio by top-view transformation of in-vehicle smartphone camera and deep learning-based classification

Quantized Federated Learning Under Transmission Delay and Outage Constraints

Federated learning (FL) has been recognized as a viable distributed learning paradigm which trains a machine learning model collaboratively with massive mobile devices in the wireless edge while protecting user privacy. Although various communication schemes have been proposed to expedite the FL process, most of them have assumed ideal wireless channels which provide reliable and lossless communication links between the server and mobile clients. Unfortunately, in practical systems with limited radio resources such as constraint on the training latency and constraints on the transmission power and bandwidth, transmission of a large number of model parameters inevitably suffers from quantization errors (QE) and transmission outage (TO). In this paper, we consider such non-ideal wireless channels, and carry out the first analysis showing that the FL convergence can be severely jeopardized by TO and QE, but intriguingly can be alleviated if the clients have uniform outage probabilities. These insightful results motivate us to propose a robust FL scheme, named <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">FedTOE</monospace> , which performs joint allocation of wireless resources and quantization bits across the clients to minimize the QE while making the clients have the same TO probability. Extensive experimental results are presented to show the superior performance of <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">FedTOE</monospace> for deep learning-based classification tasks with transmission latency constraints.

Environment Adaptive Deep Learning Classification System Based on One-shot Guidance

When utilizing the deep learning models in some real applications, the distribution of the labels in the environment can be used to increase the accuracy. Generally, to compute this distribution, there should be the validation set that is labeled by the ground truths. On the other side, the dependency of ground truths limits the utilization of the distribution in various environments. In this paper, we carried out a novel system for the deep learning-based classification to solve this problem. Firstly, our system only uses one validation set with ground truths to compute some hyper parameters, which is named as one-shot guidance. Secondly, in an environment, our system builds the validation set and labels this by the prediction results, which does not need any guidance by the ground truths. Thirdly, the computed distribution of labels by the validation set selectively cooperates with the probability of labels by the output of models, which is to increase the accuracy of predict results on testing samples. We selected six popular deep learning models on three real datasets for the evaluation. The experimental results show that our system can achieve higher accuracy than state-of-art methods while reducing the dependency of labeled validation set.

A Systematic Review on Recent Advancements in Deep and Machine Learning Based Detection and Classification of Acute Lymphoblastic Leukemia

Automatic Leukemia or blood cancer detection is a challenging job and is very much required in healthcare centers. It has a significant role in early diagnosis and treatment planning. Leukemia is a hematological disorder that starts from the bone marrow and affects white blood cells (WBCs). Microscopic analysis of WBCs is a preferred approach for an early detection of Leukemia since it is cost-effective and less painful. Very few literature reviews have been done to demonstrate a comprehensive analysis of deep and machine learning-based Acute Lymphoblastic Leukemia (ALL) detection. This article presents a systematic review of the recent advancements in this knowledge domain. Here, various artificial intelligence-based ALL detection approaches are analyzed in a systematic manner with merits and demits. The review of these schemes is conducted in a structured manner. For this purpose, segmentation schemes are broadly categorized into signal and image processing-based techniques, conventional machine learning-based techniques, and deep learning-based techniques. Conventional machine learning-based ALL classification approaches are categorized into supervised and unsupervised machine learning is presented. In addition, deep learning-based classification methods are categorized into Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), and the Autoencoder. Then, CNN-based classification schemes are further categorized into conventional CNN, transfer learning, and other advancements in CNN. A brief discussion of these schemes and their importance in ALL classification are also presented. Moreover, a critical analysis is performed to present a clear idea about the recent research in this field. Finally, various challenging issues and future scopes are discussed that may assist readers in formulating new research problems in this domain.

Simply Fine-Tuned Deep Learning-Based Classification for Breast Cancer with Mammograms

A lump growing in the breast may be referred to as a breast mass related to the tumor. However, not all tumors are cancerous or malignant. Breast masses can cause discomfort and pain, depending on the size and texture of the breast. With an appropriate diagnosis, non-cancerous breast masses can be diagnosed earlier to prevent their cultivation from being malignant. With the development of the artificial neural network, the deep discriminative model, such as a convolutional neural network, may evaluate the breast lesion to distinguish benign and malignant cancers from mammogram breast masses images. This work accomplished breast masses classification relative to benign and malignant cancers using a digital database for screening mammography image datasets. A residual neural network 50 (ResNet50) model along with an adaptive gradient algorithm, adaptive moment estimation, and stochastic gradient descent optimizers, as well as data augmentations and fine-tuning methods, were implemented. In addition, a learning rate scheduler and -fold cross-validation were applied with training procedures to determine the best models. The results of training accuracy, -value, test accuracy, area under the curve, sensitivity, precision, F1-score, specificity, and kappa for adaptive gradient algorithm , , , and stochastic gradient descent fine-tunings indicate that the classifier is feasible for categorizing breast cancer into benign and malignant from the mammographic breast masses images.

Feedforward Deep Learning Optimizer-based RNA-Seq Women's cancers Detection with a hybrid Classification Models for Biomarker Discovery

Women's cancers, signified by breast adenocarcinoma and non-small-cell lung cancers, are a significant threat to women's health. Across the globe, the leading cause of death for women is a group of tumors referred to as "female-oriented cancers". The most recent researches in the classification of molecular tumors is the analysis of women's cancers using RNA-Seq data for precision cancer diagnoses. Furthermore, discovering the different genes’ patterns behaviors will lead to predict the cancer-specific biomarkers to early diagnosis and detection of cancer-specific in women. An overfit model will be resulted due to the high-dimensional data of RNA-Seq from a small samples of data. In this work, we propose a filter-based selection approach for a deep learning-based classification model. In addition, hybrid classification models have been proposed to be compared with the new modified deep learning model. The Experiments’ analysis showed that the proposed filter-based selection approach for a deep learning-based classification model performed better than the other hybrid models in terms of performance evaluation metrics, with an accuracy of 96.7% for RNA-Seq breast adenocarcinoma data and 95.5% for RNA-Seq non-small-cell lung cancer data.

Deep learning-based classification of infectious keratitis on slit-lamp images.

Infectious keratitis (IK) is an ocular emergency caused by a variety of microorganisms, including bacteria, fungi, viruses, and parasites. Culture-based methods were the gold standard for diagnosing IK, but difficult biopsy, delaying report, and low positive rate limited their clinical application. This study aims to construct a deep-learning-based auxiliary diagnostic model for early IK diagnosis. A retrospective study. IK patients with pathological diagnosis were enrolled and their slit-lamp photos were collected. Image augmentation, normalization, and histogram equalization were applied, and five image classification networks were implemented and compared. Model blending technique was used to combine the advantages of single model. The performance of combined model was validated by 10-fold cross-validation, receiver operating characteristic curves (ROC), confusion matrix, Gradient-wright class activation mapping (Grad-CAM) visualization, and t-distributed Stochastic Neighbor Embedding (t-SNE). Three experienced cornea specialists were invited and competed with the combined model on making clinical decisions. Overall, 4830 slit-lamp images were collected from patients diagnosed with IK between June 2010 and May 2021, including 1490 (30.8%) bacterial keratitis (BK), 1670 (34.6%) fungal keratitis (FK), 600 (12.4%) herpes simplex keratitis (HSK), and 1070 (22.2%) Acanthamoeba keratitis (AK). KeratitisNet, the combination of ResNext101_32x16d and DenseNet169, reached the highest accuracy 77.08%. The accuracy of KeratitisNet for diagnosing BK, FK, AK, and HSK was 70.27%, 77.71%, 83.81%, and 79.31%, and AUC was 0.86, 0.91, 0.96, and 0.98, respectively. KeratitisNet was mainly confused in distinguishing BK and FK. There were 20% of BK cases mispredicted into FK and 16% of FK cases mispredicted into BK. In diagnosing each type of IK, the accuracy of model was significantly higher than that of human ophthalmologists (p < 0.001). KeratitisNet demonstrates a good performance on clinical IK diagnosis and classification. Deep learning could provide an auxiliary diagnostic method to help clinicians suspect IK using different corneal manifestations.

A smart healthcare-based system for classification of dementia using deep learning.

ObjectivesThis study aims to develop a deep learning-based classification model for early detection of dementia using a wearable device that can measure the electrical conductivity of the skin, temperature, and movement as factors related to dementia, interlocking them with an application, and analyzing the collected data.MethodsThis study was conducted on 18 elderly individuals (5 males, 13 females) aged 65 years or older who consented to the study. The Korean Mini-Mental State Examination survey for cognitive function tests was conducted by well-trained researchers. The subjects were first grouped into high- or low-risk group for dementia based on their Korean Mini-Mental State Examination score. Data obtained by wearable devices of each subject were then used for the classification of the high- and low-risk groups of dementia through a smart healthcare-based system implementing a deep neural network with scaled principal component analysis. The correlation coefficients between the Korean Mini-Mental State Examination score and the featured data were also investigated.ResultsOur study showed that the proposed system using a deep neural network with scaled principal component analysis was effective in detecting individuals at high risk for dementia with up to 99% accuracy and which performance was better compared with commonly used classification algorithms. In addition, it was found that the electrical conductivity of skin had the closest correlation with the results of the Korean Mini-Mental State Examination score among data collected through wearable devices in this study.ConclusionsOur proposed system can contribute to effective early detection of dementia for the elderly, using a non-invasive and easy-to-wear wearable device and classification algorithms with a simple cognitive function test. In the future, we intend to have more subjects participate in the experiment, to include more relevant variables in the wearable device, and to analyze the effectiveness of the smart healthcare-based dementia classification system over the long term.