**5. Related Works**

A literature review was conducted to examine the most recent approaches and techniques for medical data classification in this field.

The related works presented here were selected based on their technological similarity to the proposed solution and their focus on medical data. Furthermore, all papers were chosen based on their publication in high-quality journals. Furthermore, as COVID-19 has attracted the attention of researchers in the healthcare field, most of the papers selected in this review were related to the global COVID-19 pandemic.

In [51], the aim was to evaluate the performance of parallel computing and advanced *k*-means clustering as a pre-processing step for data classification and image detection in medical applications. To achieve this, the researchers utilized a parallel logistic regression algorithm and a mobile neural engine processor. The *k*-means clustering technique was used to pre-process both images and data, resulting in improvements in feature extraction, the removal of noise and outlier pixels, and classification accuracy. The results of this study showed that their proposed approach outperformed traditional methods both in terms of both accuracy and efficiency, making it a promising approach for medical data analysis and processing.

In 2021, the researchers in [52] proposed a new method for optimizing the performance of the *k*-means clustering algorithm on parallel and distributed computing systems. The study employed a hybrid approach that combined the traditional Lloyd's algorithm with a new partitioning technique. The proposed approach was evaluated using various datasets, and the results showed that the hybrid approach outperformed both the traditional Lloyd's algorithm and other state-of-the-art parallel *k*-means algorithms, in terms of both accuracy and efficiency. The study concluded that the proposed approach was a promising solution for large-scale clustering tasks on parallel and distributed computing systems. In [53], the authors proposed a new framework for automating the diagnosis of Alzheimer's disease (AD) using a machine-learning approach. The proposed framework utilized a combination of several machine-learning algorithms, including principal component analysis (PCA), support vector machine (SVM), and *k*-nearest neighbors (KNN) to classify brain images as normal or AD. The study used two different datasets, and the results showed that the proposed framework achieved high accuracy and specificity when

classifying brain images as AD. The study concluded that the proposed framework could be a valuable tool for the early diagnosis and monitoring of AD.

In 2022, [54] investigated the potential use of deep learning algorithms for the detection of COVID-19 in chest X-ray images. The study proposed a deep-learning model based on convolutional neural networks (CNNs) that had been trained on a large dataset of chest X-ray images. The model was tested on a separate dataset of chest X-ray images, and the results showed that the proposed model achieved high accuracy, sensitivity, and specificity, in detecting COVID-19. The study concluded that the proposed deep-learning model could be a valuable tool for the rapid and accurate detection of COVID-19 in chest X-ray images, especially in regions with limited access to COVID-19 testing facilities.

A literature review was conducted in order to review the most recent approaches and techniques for medical image detection.

The authors of [55] developed a machine-learning algorithm that could accurately classify patients with severe COVID-19 and predict their risk of in-hospital mortality. The study collected data from electronic health records of patients with severe COVID-19, including demographics, vital signs, laboratory values, and comorbidities. A machine-learning algorithm based on a gradient-boosting machine (GBM) was developed and trained on the collected data. The results showed that the proposed GBM model achieved high accuracy in classifying patients with severe COVID-19 and predicting their risk of in-hospital mortality. The study concluded that the proposed machine-learning algorithm could be a valuable tool for clinicians to make more informed decisions about the management of patients with severe COVID-19.

In 2021, the authors of [56] proposed a classification solution using transfer learning to assess the suitability of 3 pre-trained CNN models (EfficientNetB0, VGG16, and InceptionV3) for mobile applications. These models were selected for their accuracy and efficiency with a relatively small number of parameters. The study used a dataset compiled from various publicly available sources and evaluated the models using performance measurements and deep-learning approaches, such as accuracy, recall, specificity, precision, and F1-scores. The results demonstrated that the proposed method produced a high-quality model with a COVID-19 sensitivity of 94.79% and an overall accuracy of 92.93%. The study suggested that computer-vision techniques could be utilized to improve the efficiency of detection and screening processes.

In 2021, the authors of [57] employed convolutional neural networks (ConvNets) to accurately identify COVID-19 in computed tomography (CT) images, enabling the early classification of chest CT images of COVID-19 by hospital staff. ConvNets automatically learned and extracted features from medical image datasets, including the COVID-CT dataset used in this study. The objective was to train the GoogleNet ConvNet architecture using 425 CT-coronavirus images from the COVID-CT dataset. The experimental results indicated that GoogleNet achieved a validation accuracy of 82.14% on the dataset in 74 min and 37 s. This study demonstrated the potential of ConvNets in improving the accuracy and efficiency of COVID-19 detection in medical imaging.

In 2022, the authors of [58] proposed a new method for improving the quality of CT scans using contrast limited histogram equalization (CLAHE) and developed a convolutional neural network (CNN) model to extract important features from a dataset of 2482 CT-scan images. These features were then used as input for machine-learning methods such as support vector machine (SVM), Gaussian naive Bayes (GNB), logistic regression (LR), random forest (RF), and decision tree (DT). The researchers recommended an ensemble method for classifying COVID-19 CT images and compared the performance of their model with other state-of-the-art methods. The proposed model outperformed existing models with an accuracy of 99.73%, a precision of 99.46%, and a recall of 100%.

In 2022, the authors of [59] described an approach that used a generative adversarial network (GAN) to improve the accuracy of a deep-learning model for classifying COVID-19 infections in chest X-ray images. To generate additional training data, the COVID-19 positive chest X-ray images were fed into a styleGAN2 model, which produced new images

for training the deep-learning model. The resulting dataset was used to train a CNN binary classifier model that achieved a classification accuracy of 99.78%. This method could aid in the rapid and accurate diagnosis of COVID-19 infections from chest X-ray images.
