Performance Of Deep Learning Models Research Articles

Performance Enhancement of Alzheimer's Disease Diagnosis Using Generative Adversarial Network

Insufficient medical images have always been a challenge for deep learning-based Alzheimer's disease classification and detection tasks. The availability of Magnetic Resonance Imaging (MRI) data is limited due to patients' privacy issues. Access to individual medical records is strongly protected by the law, and appropriate consent is needed to utilize them for research purposes. Besides that, publicly available databases often experience imbalanced classes, further contributing to the issue of insufficient data. Thus, the performance of deep learning models used to diagnose Alzheimer's disease is often hindered by this issue. Basic data augmentation methods such as geometrical augmentation techniques also had limited applications to medical data. Hence, this study proposes a deep learning technique of Generative Adversarial Network (GAN) to expand the MRI dataset and improve the classification model performance. MRI images from the Open Access Series of Imaging Studies (OASIS) database are used in this study to perform experiments and validate hypotheses. After applying GAN to expand the dataset, a pre-trained Convolutional Neural Network (CNN) model is used to classify the data into multiple classes of Alzheimer's and the model's performance is measured. As a result, an improvement in accuracy for the classification task can be observed, indicating that the GAN is a solution for overcoming the challenge of insufficient data for Alzheimer's diagnosis.

Prediction of Atrial Fibrillation with Single-Lead Mobile ECG during Normal Sinus Rhythm using Deep Learning

Abstract Background Single-lead ECG (electrocardiogram) acquisition from mobile devices provides more convenient way to identify arrhythmia. Prediction of atrial fibrillation (AF) by artificial intelligence facilitates AF screening. However, little is known about the prediction of AF with a single-lead ECG. Therefore, we proposed a method for predicting AF using single-lead mobile ECG during normal sinus rhythm (NSR). Methods Three deep learning models RNN, LSTM, and Resnet50 were used. Out of 13,509 ECGs from a total of 6,719 patients, we utilized a total of 10,287 NSR ECGs from 5,170 patients. Single-lead mobile ECGs were adjusted with noise filtering and segmented every 10 seconds. A random under-sampling was applied to reduce bias from data imbalance. Finally, 31,767 ECG segments composed of 15,157 segments labeled as potential AF and 16,610 segments labeled as healthy were entered for final analysis. Results The ResNet50 showed recall of 79.3%, precision of 65.8%, F1-score of 71.9%, accuracy of 70.5%, and AUC of 0.79 for the prediction of AF with NSR ECGs. These scores were higher than the other two models followed by RNN at 0.75 and LSTM at 0.74. For the external validation set, ResNet50 showed an F1-score of 64.1%, recall of 68.9%, precision of 60.0%, accuracy of 63.4%, and AUC of 0.68 for the prediction of AF with NSR ECGs. Conclusion A deep learning model using a single-lead mobile ECG during NSR predicted AF at risk in future. However, further study is needed to improve the performance of deep learning models for applying in the clinical setting.

Multi-centre benchmarking of deep learning models for COVID-19 detection in chest x-rays

IntroductionThis study is a retrospective evaluation of the performance of deep learning models that were developed for the detection of COVID-19 from chest x-rays, undertaken with the goal of assessing the suitability of such systems as clinical decision support tools.MethodsModels were trained on the National COVID-19 Chest Imaging Database (NCCID), a UK-wide multi-centre dataset from 26 different NHS hospitals and evaluated on independent multi-national clinical datasets. The evaluation considers clinical and technical contributors to model error and potential model bias. Model predictions are examined for spurious feature correlations using techniques for explainable prediction.ResultsModels performed adequately on NHS populations, with performance comparable to radiologists, but generalised poorly to international populations. Models performed better in males than females, and performance varied across age groups. Alarmingly, models routinely failed when applied to complex clinical cases with confounding pathologies and when applied to radiologist defined “mild” cases.DiscussionThis comprehensive benchmarking study examines the pitfalls in current practices that have led to impractical model development. Key findings highlight the need for clinician involvement at all stages of model development, from data curation and label definition, to model evaluation, to ensure that all clinical factors and disease features are appropriately considered during model design. This is imperative to ensure automated approaches developed for disease detection are fit-for-purpose in a clinical setting.

Performance Evaluation of Deep Learning Models for Classifying Cybersecurity Attacks in IoT Networks

The Internet of Things (IoT) presents great potential in various fields such as home automation, healthcare, and industry, among others, but its infrastructure, the use of open source code, and lack of software updates make it vulnerable to cyberattacks that can compromise access to data and services, thus making it an attractive target for hackers. The complexity of cyberattacks has increased, posing a greater threat to public and private organizations. This study evaluated the performance of deep learning models for classifying cybersecurity attacks in IoT networks, using the CICIoT2023 dataset. Three architectures based on DNN, LSTM, and CNN were compared, highlighting their differences in layers and activation functions. The results show that the CNN architecture outperformed the others in accuracy and computational efficiency, with an accuracy rate of 99.10% for multiclass classification and 99.40% for binary classification. The importance of data standardization and proper hyperparameter selection is emphasized. These results demonstrate that the CNN-based model emerges as a promising option for detecting cyber threats in IoT environments, supporting the relevance of deep learning in IoT network security.

From simple to complex: a sequential method for enhancing time series forecasting with deep learning

Abstract Time series forecasting is a well-known deep learning application field in which previous data are used to predict the future behavior of the series. Recently, several deep learning approaches have been proposed in which several nonlinear functions are applied to the input to obtain the output. In this paper, we introduce a novel method to improve the performance of deep learning models in time series forecasting. This method divides the model into hierarchies or levels from simpler to more complex ones. Simpler levels handle smoothed versions of the input, whereas the most complex level processes the original time series. This method follows the human learning process where general/simpler tasks are performed first, and afterward, more precise/harder ones are accomplished. Our proposed methodology has been applied to the LSTM architecture, showing remarkable performance in various time series. In addition, a comparison is reported including a standard LSTM and novel methods such as DeepAR, Temporal Fusion Transformer, NBEATS and Echo State Network.

Human Activity Recognition Algorithm with Physiological and Inertial Signals Fusion: Photoplethysmography, Electrodermal Activity, and Accelerometry.

Inertial signals are the most widely used signals in human activity recognition (HAR) applications, and extensive research has been performed on developing HAR classifiers using accelerometer and gyroscope data. This study aimed to investigate the potential enhancement of HAR models through the fusion of biological signals with inertial signals. The classification of eight common low-, medium-, and high-intensity activities was assessed using machine learning (ML) algorithms, trained on accelerometer (ACC), blood volume pulse (BVP), and electrodermal activity (EDA) data obtained from a wrist-worn sensor. Two types of ML algorithms were employed: a random forest (RF) trained on features; and a pre-trained deep learning (DL) network (ResNet-18) trained on spectrogram images. Evaluation was conducted on both individual activities and more generalized activity groups, based on similar intensity. Results indicated that RF classifiers outperformed corresponding DL classifiers at both individual and grouped levels. However, the fusion of EDA and BVP signals with ACC data improved DL classifier performance compared to a baseline DL model with ACC-only data. The best performance was achieved by a classifier trained on a combination of ACC, EDA, and BVP images, yielding F1-scores of 69 and 87 for individual and grouped activity classifications, respectively. For DL models trained with additional biological signals, almost all individual activity classifications showed improvement (p-value < 0.05). In grouped activity classifications, DL model performance was enhanced for low- and medium-intensity activities. Exploring the classification of two specific activities, ascending/descending stairs and cycling, revealed significantly improved results using a DL model trained on combined ACC, BVP, and EDA spectrogram images (p-value < 0.05).

Less-supervised learning with knowledge distillation for sperm morphology analysis

ABSTRACT Sperm Morphology Analysis (SMA) is pivotal in diagnosing male infertility. However, manual analysis is subjective and time-intensive. Artificial intelligence presents automated alternatives, but hurdles like limited data and image quality constraints hinder its efficacy. These challenges impede Deep Learning (DL) models from grasping crucial sperm features. A solution enabling DL models to learn sample nuances, even with limited data, would be invaluable. This study proposes a Knowledge Distillation (KD) method to distinguish normal from abnormal sperm cells, leveraging the Modified Human Sperm Morphology Analysis dataset. Despite low-resolution, blurry images, our method yields relevant results. We exclusively utilize normal samples to train the model for anomaly detection, crucial in scenarios lacking abnormal data – a common issue in medical tasks. Our aim is to train an Anomaly Detection model using a dataset comprising unclear images and limited samples, without direct exposure to abnormal data. Our method achieves Receiver ROC/AUC scores of 70.4%, 87.6%, and 71.1% for head, vacuole, and acrosome, respectively, our method matches traditional DL model performance with less than 70% of the data. This less-supervised approach shows promise in advancing SMA despite data scarcity. Furthermore, KD enables model adaptability to edge devices in fertility clinics, requiring less processing power.

A comparative study of pre-processing methods to improve glioma segmentation performance in brain MRI using deep learning

Background: Glioma is the most common brain tumor in adult patients and requires accurate treatment. The delineation of tumor boundaries must be accurate and precise, which is crucial for treatment planning. Currently, delineating boundaries for tumors is a tedious, time-consuming task and may be prone to human error among oncologists. Therefore, artificial intelligence plays a vital role in reducing these problems. Objective: This study aims to find a relationship between improving image enhancement and evaluating the performance of deep learning models for segmenting glioma image data on brain MRI images. Materials and methods: The BraTs2023 dataset was used in this study. The image dataset was converted from three dimensions to two dimensions and then subjected to pre-processing via four image enhancement techniques, including contrast-limited adaptive histogram equalization (CLAHE), gamma correction (GC), non-local mean filter (NLMF), and median and Wiener filter (MWF). Subsequently, it was evaluated for structural similarity index (SSIM) and mean squared error. The deep learning segmentation model was created using the U-Net architecture and assessed for dice similarity coefficient (DSC), accuracy, precision, recall, F1-score, and Jaccard index for tumor segmentation. Results: The performance of enhanced image results for CLAHE, GC, NLMF, and MWF techniques shows SSIM values of 0.912, 0.905, 0.999, and 0.911, respectively. The dice similarity coefficient (DSC) for segmentation without image enhancement was 0.817. The DSC of segmentation for CLAHE, GC, NLMF, and MWF techniques were 0.818, 0.812, 0.820, and 0.797, respectively. Conclusion: The enhanced image technique could affect the performance of tumor segmentation. by the enhanced image for use in a trained model may increase or decrease performance depending on the chosen image enhancement technique and the parameters determined by each method.

Deep learning based fault detection of automobile dry clutch system using spectrogram plots

Abstract In this research study, a novel approach that leverages the power of transfer learning (a famous deep learning technique) is proposed to diagnose multiple types of clutch faults including, worn release fingers, fractured pressure plates, deteriorated pressure plates, loss of friction material and distorted tangential strips using spectrogram plots. To train and validate the diagnostic system, vibration readings were taken from a specially designed test rig with the help of piezoelectric accelerometer while the clutch system was operated under different load conditions of 0 (no load), 5 and 10 kg This procedure of data collection was then repeated to acquire the vibration data for all of the fault conditions by replacing the good with fault components individually. These vibration signals were further processed and transformed into spectrogram plots that serves as the input data for the deep learning models considered. Fine-tuning techniques were applied on pretrained networks to maximise the prediction accuracy of the models to effectively determine and diagnose faults in the clutch system. For this study 12 pre-trained networks were chosen namely, Xception, InceptionResNet, DenseNet, AlexNet, VGG16, GoogLeNet, VGG19, ResNet101, ResNet50, InceptionV3, MobileNetV2 and ShuffleNet. To optimize the performance of deep learning models, a systematic adjustment of hyperparameters such as the train-test split ratio, learning rate, optimizer and batch size for each network model was carried out. Through careful experimentation and analysis, significant improvements in fault classification accuracy were achieved thereby enhancing the reliability and effectiveness of the diagnostic system. From the results it was noted that 100% classification accuracy was displayed by AlexNet (for the no load condition and the 10kg load condition) and GoogLeNet (for 5 kg load condition) with extremely low computation times.

JPSSL: SAR Terrain Classification Based on Jigsaw Puzzles and FC-CRF

Effective features play an important role in synthetic aperture radar (SAR) image interpretation. However, since SAR images contain a variety of terrain types, it is not easy to extract effective features of different terrains from SAR images. Deep learning methods require a large amount of labeled data, but the difficulty of SAR image annotation limits the performance of deep learning models. SAR images have inevitable geometric distortion and coherence speckle noise, which makes it difficult to extract effective features from SAR images. If effective semantic context features cannot be learned for SAR images, the extracted features struggle to distinguish different terrain categories. Some existing terrain classification methods are very limited and can only be applied to some specified SAR images. To solve these problems, a jigsaw puzzle self-supervised learning (JPSSL) framework is proposed. The framework comprises a jigsaw puzzle pretext task and a terrain classification downstream task. In the pretext task, the information in the SAR image is learned by completing the SAR image jigsaw puzzle to extract effective features. The terrain classification downstream task is trained using only a small number of labeled data. Finally, fully connected conditional random field processing is performed to eliminate noise points and obtain a high-quality terrain classification result. Experimental results on three large-scene high-resolution SAR images confirm the effectiveness and generalization of our method. Compared with the supervised methods, the features learned in JPSSL are highly discriminative, and the JPSSL achieves good classification accuracy when using only a small amount of labeled data.

Forecasting short-term methane based on corrected numerical weather prediction outputs

Methane (CH4) represents a significant greenhouse gas, and the control of its emissions is crucial in impacting global climate change. Accurate forecasting of CH4 emissions is important for identifying future sources of high CH4 emissions, rational design of experiments, and information gathering. This study presented a method to improve numerical weather prediction (NWP) outputs using the Extreme Gradient Boosting (XGB) algorithm. In contrast to previous conventional correction methods, the XGB model greatly enhances the accuracy and stability of the NWP outputs. Short-term (1-3 d) CH4 forecasting models were built for 53 stations worldwide using three machine learning algorithms: Convolutional Neural Network (CNN), Random Forest (RF), and XGB. The models were validated with flux data from the stations. The results showed that the CNN model performed best in forecasting CH4 for all lead times (mean NRMSE=0.34), followed by the RF model (mean NRMSE=0.68) and the XGB model performed worst (mean NRMSE=0.74). For land cover types (e.g., savanna (WSA) and forest (EBF)), the accuracy of the model may be greater for 2 d lead times than for 1 d lead times, which is closely related to the regional climate and terrain conditions. For lead times, the CNN model achieves high CH4 forecast accuracy at 1-2 d lead times. In addition, the RF and XGB models with 1 d lead times exhibit significant overestimation at most stations in North America (PBIAS>0.5). This work quantified the performance of machine learning models for correcting NWP outputs, providing a technical approach to the correction of NWP outputs; and assessed the performance of deep learning models for forecasting short-term CH4, providing a reference for model selection for CH4 forecasting.

Automatic segmentation of femoral tumors by nnU-net

BackgroundMetastatic femoral tumors may lead to pathological fractures during daily activities. A CT-based finite element analysis of a patient's femurs was shown to assist orthopedic surgeons in making informed decisions about the risk of fracture and the need for a prophylactic fixation. Improving the accuracy of such analyses ruires an automatic and accurate segmentation of the tumors and their automatic inclusion in the finite element model. We present herein a deep learning algorithm (nnU-Net) to automatically segment lytic tumors within the femur. MethodA dataset consisting of fifty CT scans of patients with manually annotated femoral tumors was created. Forty of them, chosen randomly, were used for training the nnU-Net, while the remaining ten CT scans were used for testing. The deep learning model's performance was compared to two experienced radiologists. FindingsThe proposed algorithm outperformed the current state-of-the-art solutions, achieving dice similarity scores of 0.67 and 0.68 on the test data when compared to two experienced radiologists, while the dice similarity score for inter-individual variability between the radiologists was 0.73. InterpretationThe automatic algorithm may segment lytic femoral tumors in CT scans as accurately as experienced radiologists with similar dice similarity scores. The influence of the realistic tumors inclusion in an autonomous finite element algorithm is presented in (Rachmil et al., 2024 [14]).

Comparison of Deep Learning Models and Optimization Algorithms in the Detection of Scoliosis and Spondylolisthesis from X-Ray Images

The spine is composed of pieces of bone called vertebrae that lie between the skull and the tailbone. Various medical conditions can affect the spine. In this study, two types of degenerative diseases, scoliosis, and spondylolisthesis, were studied. Deep AI architectures have recently enabled further disease diagnosis innovation using medical images. Various traditional and deep learning studies use medical images for disease diagnosis in the literature. This study aims to classify spine X-ray images according to three possible conditions (Normal, Scoliosis, and Spondylolisthesis) and to exploit the potential of these X-ray images to detect possible diseases occurring in the spine. The performance of deep learning models and optimization algorithms used in this process was evaluated. The study uses a data set created and/or analyzed during an existing study. This data set consists of images that belong to three different classes: scoliosis, spondylolisthesis, or x-ray images of normal (i.e. healthy) individuals. A total of 338 spine X-ray images, 188 scoliosis images, 79 spondylolisthesis images, and 71 normal images. Six different deep-learning architectures have been used in the study. These architectures are Alexnet, GoogLeNet, ResNet-18, ResNet-50, ResNet-101, and EfficientNet-bo. While working on these deep architectures, each model has been evaluated using different optimization algorithms. These optimization algorithms are RmsProp, SGDM, and Adam. According to the classification processes, the deep learning model with the highest accuracy value was Alexnet, and the optimization algorithm used with it, Sgdm (99.01%), and the training time lasted 38 seconds. According to the classification processes, the deep learning model with the fastest completion time (30 seconds) was Alexnet and the optimization algorithm used with it was RmsProp. An accuracy rate of 98.02% has been obtained in the training of this model.

GAMNet: A deep learning approach for precise gesture identification

Deep learning has gained popularity across several industries, including object recognition and classification. In the case of Convolutional Neural Networks (CNN), the first layers extract the most noticeable elements, such as shape and margin. As the model progresses, it learns to extract more complex features such as texture and color; conversely, skeleton features encompass significant locations (joints) that do not naturally align with the grid-like architecture intended for these networks. This study emphasizes the importance of structural features in enhancing the performance of deep learning models. It introduces the Gesture Analysis Module Network (GAMNet), which computes abstract structural values within the architecture for feature extraction, prioritization, and classification. These values go through a rigorous evaluation process along with the cutting-edge deep learning model, CNN, and result in intermediate representations, leading to better performance in gesture analysis. An automated dance gesture identification system can address the challenges of recognizing hand movements in unpredictable lighting, varied backgrounds, noise, and changing camera angles. Despite these challenges, GAMNet performed remarkably well, surpassing renowned models like VGGNet, ResNet, EfficientNet, and CNN, achieving a classification accuracy of 96.80%, even in challenging image circumstances. This paper highlights how GAMNet can revolutionize the world of classical Indian dance, opening up new opportunities for research and development in this field.

Distributed Denial of Service Attack Detection in IoT Networks using Deep Learning and Feature Fusion: A Review

The explosive growth of Internet of Things (IoT) devices has led to escalating threats from distributed denial of service (DDoS) attacks. Moreover, the scale and heterogeneity of IoT environments pose unique security challenges, and intelligent solutions tailored for the IoT are needed to defend critical infrastructure. The deep learning technique shows great promise because automatic feature learning capabilities are well suited for the complex and high-dimensional data of IoT systems. Additionally, feature fusion approaches have gained traction in enhancing the performance of deep learning models by combining complementary feature sets extracted from multiple data sources. This paper aims to provide a comprehensive literature review focused specifically on deep learning techniques and feature fusion for DDoS attack detection in IoT networks. Studies employing diverse deep learning models and feature fusion techniques are analysed, highlighting key trends and developments in this crucial domain. This review provides several significant contributions, including an overview of various types of DDoS attacks, a comparison of existing surveys, and a thorough examination of recent applications of deep learning and feature fusion for detecting DDoS attacks in IoT networks. Importantly, it highlights the current challenges and limitations of these deep learning techniques based on the literature surveyed. This review concludes by suggesting promising areas for further research to enhance deep learning security solutions, which are specifically tailored to safeguarding the fast-growing IoT infrastructure against DDoS attacks.

The effect of time normalization and biomechanical signal processing techniques of ground reaction force curves on deep-learning model performance

Time-series data are common in biomechanical studies. These data often undergo pre-processing steps such as time normalization or filtering prior to use in further analyses, including deep-learning classification. In this context, it remains unclear how these preprocessing steps affect deep-learning model performance. Thus, the aim of this study is to assess the effect of time-normalization and filtering on the performance of deep-learning classification models. We also investigated the effect of amplitude scaling. Using a public dataset (Gutenburg Gait Database, a ground reaction force database of level overground walking at self-selected walking speed involving 350 healthy individuals), we trained convolutional neural network (CNN) and long short-term memory (LSTM) models to predict binary sex (male, female) using three-dimensional ground-reaction forces to which we applied different processing approaches: zero padding, interpolation to 100% of signal, filtering, and scaling (min–max, body mass). The results show that transformations resulted in differences in model performances. Highest performance was obtained using unfiltered data, zero-padding, and min–max amplitude scaling (F1-score of 91 and 87% for CNN and LSTM, respectively). Not filtering data and using min–max scaling generally improve performance for both model architectures. For interpolation, results are not consistent across model architectures. This study suggests that processing steps must be considered in applications where deep-learning classification performance is relevant.

A Deep Learning Approach for Malnutrition Detection

The timely detection of malnutrition in children is of paramount importance, as it allows for early intervention and treatment. This proactive approach not only prevents further health deterioration but also fosters proper growth, minimizing the long-term consequences of malnutrition, such as stunted growth, impaired cognitive development, and increased vulnerability to diseases. Our work encompasses the creation of a new dataset comprising images of children in Healthy, Undernourished, Stunting, and Wasting categories. The core objective is to assess the deep learning model performance in classifying these children images. The experimentation is carried out by varying epochs, batch size, optimizers AdamW, Adamax, and RMSprop; and different values of the learning rate 0.1, 0.01, 0.001, and 0.0001 during model training. The model is trained on image dataset constructed by cleaning images generated by the stable diffusion model. The model is tested on randomly selected child images from websites. The model successfully classified two classes with 95% accuracy, 97.6% F1 score, precision 97.6%, and 97.6% recall with Adam optimizers, 0.0001 learning rate, and Batch size 4. Additionally, for the four-class categorization scenario, the study broadens the classification. The model achieved 88.87% accuracy, 90.3% recall, 90.2% precision, and an F1 score of 90% for four-class categorization with AdamW optimization, 0.0001 learning rate, and batch size 6. These results are satisfactory for prediction of malnutrition category in children.

Utilizing Polarization Diversity in GBSAR Data-Based Object Classification.

In recent years, the development of intelligent sensor systems has experienced remarkable growth, particularly in the domain of microwave and millimeter wave sensing, thanks to the increased availability of affordable hardware components. With the development of smart Ground-Based Synthetic Aperture Radar (GBSAR) system called GBSAR-Pi, we previously explored object classification applications based on raw radar data. Building upon this foundation, in this study, we analyze the potential of utilizing polarization information to improve the performance of deep learning models based on raw GBSAR data. The data are obtained with a GBSAR operating at 24 GHz with both vertical (VV) and horizontal (HH) polarization, resulting in two matrices (VV and HH) per observed scene. We present several approaches demonstrating the integration of such data into classification models based on a modified ResNet18 architecture. We also introduce a novel Siamese architecture tailored to accommodate the dual input radar data. The results indicate that a simple concatenation method is the most promising approach and underscore the importance of considering antenna polarization and merging strategies in deep learning applications based on radar data.

Synthesizing 3D Multi-Contrast Brain Tumor MRIs Using Tumor Mask Conditioning.

Data scarcity and data imbalance are two major challenges in training deep learning models on medical images, such as brain tumor MRI data. The recent advancements in generative artificial intelligence have opened new possibilities for synthetically generating MRI data, including brain tumor MRI scans. This approach can be a potential solution to mitigate the data scarcity problem and enhance training data availability. This work focused on adapting the 2D latent diffusion models to generate 3D multi-contrast brain tumor MRI data with a tumor mask as the condition. The framework comprises two components: a 3D autoencoder model for perceptual compression and a conditional 3D Diffusion Probabilistic Model (DPM) for generating high-quality and diverse multi-contrast brain tumor MRI samples, guided by a conditional tumor mask. Unlike existing works that focused on generating either 2D multi-contrast or 3D single-contrast MRI samples, our models generate multi-contrast 3D MRI samples. We also integrated a conditional module within the UNet backbone of the DPM to capture the semantic class-dependent data distribution driven by the provided tumor mask to generate MRI brain tumor samples based on a specific brain tumor mask. We trained our models using two brain tumor datasets: The Cancer Genome Atlas (TCGA) public dataset and an internal dataset from the University of Texas Southwestern Medical Center (UTSW). The models were able to generate high-quality 3D multi-contrast brain tumor MRI samples with the tumor location aligned by the input condition mask. The quality of the generated images was evaluated using the Fréchet Inception Distance (FID) score. This work has the potential to mitigate the scarcity of brain tumor data and improve the performance of deep learning models involving brain tumor MRI data.

Improving accuracy of vascular access quality classification in hemodialysis patients using deep learning with K highest score feature selection.

To develop and evaluate a novel feature selection technique, using photoplethysmography (PPG) sensors, for enhancing the performance of deep learning models in classifying vascular access quality in hemodialysis patients. This cross-sectional study involved creating a novel feature selection method based on SelectKBest principles, specifically designed to optimize deep learning models for PPG sensor data, in hemodialysis patients. The method effectiveness was assessed by comparing the performance of multiple deep learning models using the feature selection approach versus complete feature set. The model with the highest accuracy was then trained and tested using a 70:30 approach, respectively, with the full dataset and the SelectKBest dataset. Performance results were compared using Student's paired t-test. Data from 398 hemodialysis patients were included. The 1-dimensional convolutional neural network (CNN1D) displayed the highest accuracy among different models. Implementation of the SelectKBest-based feature selection technique resulted in a statistically significant improvement in the CNN1D model's performance, achieving an accuracy of 92.05% (with feature selection) versus 90.79% (with full feature set). These findings suggest that the newly developed feature selection approach might aid in accurately predicting vascular access quality in hemodialysis patients. This advancement may contribute to the development of reliable diagnostic tools for identifying vascular complications, such as stenosis, potentially improving patient outcomes and their quality of life.