GAN-Based Approaches for Generating Structured Data in the Medical Domain

Abedi, Masoud; Hempel, Lars; Sadeghi, Sina; Kirsten, Toralf

doi:10.3390/app12147075

Cited by 23 publications

(14 citation statements)

References 65 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given that the most common augmentation methods used to increase the dataset do not fit with the type of dataset that is being used in this research, we choose GANs as the base model. A variety of versions of the model have been developed, each with a particular purpose [ 28 ].…”

Section: Proposed Methods and Materialsmentioning

confidence: 99%

COVID-19 Hierarchical Classification Using a Deep Learning Multi-Modal

Althenayan,

AlSalamah,

Aly

et al. 2024

Sensors

View full text Add to dashboard Cite

Coronavirus disease 2019 (COVID-19), originating in China, has rapidly spread worldwide. Physicians must examine infected patients and make timely decisions to isolate them. However, completing these processes is difficult due to limited time and availability of expert radiologists, as well as limitations of the reverse-transcription polymerase chain reaction (RT-PCR) method. Deep learning, a sophisticated machine learning technique, leverages radiological imaging modalities for disease diagnosis and image classification tasks. Previous research on COVID-19 classification has encountered several limitations, including binary classification methods, single-feature modalities, small public datasets, and reliance on CT diagnostic processes. Additionally, studies have often utilized a flat structure, disregarding the hierarchical structure of pneumonia classification. This study aims to overcome these limitations by identifying pneumonia caused by COVID-19, distinguishing it from other types of pneumonia and healthy lungs using chest X-ray (CXR) images and related tabular medical data, and demonstrate the value of incorporating tabular medical data in achieving more accurate diagnoses. Resnet-based and VGG-based pre-trained convolutional neural network (CNN) models were employed to extract features, which were then combined using early fusion for the classification of eight distinct classes. We leveraged the hierarchal structure of pneumonia classification within our approach to achieve improved classification outcomes. Since an imbalanced dataset is common in this field, a variety of versions of generative adversarial networks (GANs) were used to generate synthetic data. The proposed approach tested in our private datasets of 4523 patients achieved a macro-avg F1-score of 95.9% and an F1-score of 87.5% for COVID-19 identification using a Resnet-based structure. In conclusion, in this study, we were able to create an accurate deep learning multi-modal to diagnose COVID-19 and differentiate it from other kinds of pneumonia and normal lungs, which will enhance the radiological diagnostic process.

show abstract

Section: Proposed Methods and Materialsmentioning

confidence: 99%

COVID-19 Hierarchical Classification Using a Deep Learning Multi-Modal

Althenayan,

AlSalamah,

Aly

et al. 2024

Sensors

View full text Add to dashboard Cite

show abstract

“…Huang et al [22] leveraged Generative Adversarial Networks (GANs) to propose an end-to-end pipeline covering the entire lifecycle of a machine learning project, along with alternatives and best practices in both the academic and practical realms. Abedi et al [23] developed and validated an evaluation framework wherein binary classifiers are trained on an extended dataset comprising real and synthetic data. The findings suggest that classifiers trained with data generated from advanced GAN models can achieve improved accuracy, even when the original data are limited.…”

Section: Related Workmentioning

confidence: 99%

FLTGAN: A Novel Framework for Enhanced Diabetes Classification in Imbalanced Datasets

Yang,

Liu,

Wang

et al. 2024

Preprint

View full text Add to dashboard Cite

Class imbalances in diabetes datasets are common in datasets in the medical research field, and class imbalances may cause the training effects of machine learning models to be biased toward a larger number of classes, thus affecting the model's ability to predict a small number of classes. In order to solve this problem, this study designed and implemented a diabetes classification model based on Focal Loss Tabular Generative Adversarial Network (FLTGAN) to effectively solve the problem of sample imbalance in uneven datasets. The model, integrating advanced machine learning algorithms, has been validated using two real-world datasets. Comparative results demonstrated that our model significantly outperformed traditional TGAN and SMOTE methods across multiple evaluation metrics, it achieved an F1 score of 0.78, an AUC of 0.84, an accuracy of 0.80, a precision of 0.83, and a recall of 0.75, indicating its superior classification performance. This study confirmed the robust capability of the proposed intelligent diabetes classification method in accurately diagnosing diabetes and addressing class imbalance, offering a new effective tool for the intelligent diagnosis of diabetes.

show abstract

“…In GAN [17] the discriminator is initially updated using the SGD in the first k steps during training. During this phase, the parameters and weights of the generator network remain fixed.…”

Section: Ganmentioning

confidence: 99%

A Novel Mixed Precision Distributed TPU GAN for Accelerated Learning Curve

Ravikumar¹,

Sriraman²

2023

Computer Systems Science and Engineering

View full text Add to dashboard Cite

Deep neural networks are gaining importance and popularity in applications and services. Due to the enormous number of learnable parameters and datasets, the training of neural networks is computationally costly. Parallel and distributed computation-based strategies are used to accelerate this training process. Generative Adversarial Networks (GAN) are a recent technological achievement in deep learning. These generative models are computationally expensive because a GAN consists of two neural networks and trains on enormous datasets. Typically, a GAN is trained on a single server. Conventional deep learning accelerator designs are challenged by the unique properties of GAN, like the enormous computation stages with non-traditional convolution layers. This work addresses the issue of distributing GANs so that they can train on datasets distributed over many TPUs (Tensor Processing Unit). Distributed learning training accelerates the learning process and decreases computation time. In this paper, the Generative Adversarial Network is accelerated using the distributed multi-core TPU in distributed data-parallel synchronous model. For adequate acceleration of the GAN network, the data parallel SGD (Stochastic Gradient Descent) model is implemented in multi-core TPU using distributed TensorFlow with mixed precision, bfloat16, and XLA (Accelerated Linear Algebra). The study was conducted on the MNIST dataset for varying batch sizes from 64 to 512 for 30 epochs in distributed SGD in TPU v3 with 128 × 128 systolic array. An extensive batch technique is implemented in bfloat16 to decrease the storage cost and speed up floating-point computations. The accelerated learning curve for the generator and discriminator network is obtained. The training time was reduced by 79% by varying the batch size from 64 to 512 in multi-core TPU.

show abstract

GAN-Based Approaches for Generating Structured Data in the Medical Domain

Cited by 23 publications

References 65 publications

COVID-19 Hierarchical Classification Using a Deep Learning Multi-Modal

COVID-19 Hierarchical Classification Using a Deep Learning Multi-Modal

FLTGAN: A Novel Framework for Enhanced Diabetes Classification in Imbalanced Datasets

A Novel Mixed Precision Distributed TPU GAN for Accelerated Learning Curve

Contact Info

Product

Resources

About