Classification of Alzheimer’s disease using MRI data based on Deep Learning Techniques

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2024, Volume and Issue: 36(2), P. 101940 - 101940

Published: Jan. 24, 2024

Alzheimer's Disease (AD) is a worldwide concern impacting millions of people, with no effective treatment known to date. Unlike cancer, which has seen improvement in preventing its progression, early detection remains critical managing the burden AD. This paper suggests novel AD-DL approach for detecting AD using Deep Learning (DL) Techniques. The dataset consists pictures brain magnetic resonance imaging (MRI) used evaluate and validate suggested model. method includes stages pre-processing, DL model training, evaluation. Five models autonomous feature extraction binary classification are shown. divided into two categories: without Data Augmentation (without-Aug), CNN-without-AUG, (with-Aug), CNNs-with-Aug, CNNs-LSTM-with-Aug, CNNs-SVM-with-Aug, Transfer learning VGG16-SVM-with-Aug. main goal build best accuracy, recall, precision, F1 score, training time, testing time. recommended methodology, showing encouraging results. experimental results show that CNN-LSTM superior, an accuracy percentage 99.92%. outcomes this study lay groundwork future DL-based research identification.

Language: Английский

---

Automatic Classification of Colour Fundus Images for Prediction Eye Disease Types Based on Hybrid Features

Diagnostics, Journal Year: 2023, Volume and Issue: 13(10), P. 1706 - 1706

Published: May 11, 2023

Early detection of eye diseases is the only solution to receive timely treatment and prevent blindness. Colour fundus photography (CFP) an effective examination technique. Because similarity in symptoms early stages difficulty distinguishing between type disease, there a need for computer-assisted automated diagnostic techniques. This study focuses on classifying disease dataset using hybrid techniques based feature extraction with fusion methods. Three strategies were designed classify CFP images diagnosis disease. The first method Artificial Neural Network (ANN) features from MobileNet DenseNet121 models separately after reducing high dimensionality repetitive Principal Component Analysis (PCA). second ANN basis fused before features. third handcrafted Based features, attained AUC 99.23%, accuracy 98.5%, precision 98.45%, specificity 99.4%, sensitivity 98.75%.

Language: Английский

---

Hybrid Techniques for the Diagnosis of Acute Lymphoblastic Leukemia Based on Fusion of CNN Features

Diagnostics, Journal Year: 2023, Volume and Issue: 13(6), P. 1026 - 1026

Published: March 8, 2023

Acute lymphoblastic leukemia (ALL) is one of the deadliest forms due to bone marrow producing many white blood cells (WBC). ALL most common types cancer in children and adults. Doctors determine treatment according its stages spread body. rely on analyzing samples under a microscope. Pathologists face challenges, such as similarity between infected normal WBC early stages. Manual diagnosis prone errors, differences opinion, lack experienced pathologists compared number patients. Thus, computer-assisted systems play an essential role assisting detection ALL. In this study, with high efficiency accuracy were developed analyze images C-NMC 2019 ALL-IDB2 datasets. all proposed systems, micrographs improved then fed active contour method extract WBC-only regions for further analysis by three CNN models (DenseNet121, ResNet50, MobileNet). The first strategy two datasets hybrid technique CNN-RF CNN-XGBoost. DenseNet121, MobileNet deep feature maps. produce features redundant non-significant features. So, maps Principal Component Analysis (PCA) select highly representative sent RF XGBoost classifiers classification using serially fused models. DenseNet121-ResNet50, ResNet50-MobileNet, DenseNet121-MobileNet, DenseNet121-ResNet50-MobileNet merged classified XGBoost. classifier reached AUC 99.1%, 98.8%, sensitivity 98.45%, precision 98.7%, specificity 98.85% dataset. With dataset, achieved 100% results AUC, accuracy, sensitivity, precision, specificity.

Language: Английский

---

Automatic Analysis of MRI Images for Early Prediction of Alzheimer’s Disease Stages Based on Hybrid Features of CNN and Handcrafted Features

Diagnostics, Journal Year: 2023, Volume and Issue: 13(9), P. 1654 - 1654

Published: May 8, 2023

Alzheimer’s disease (AD) is considered one of the challenges facing health care in modern century; until now, there has been no effective treatment to cure it, but are drugs slow its progression. Therefore, early detection vital take needful measures before it develops into brain damage which cannot be treated. Magnetic resonance imaging (MRI) techniques have contributed diagnosis and prediction MRI images require highly experienced doctors radiologists, analysis takes time analyze each slice. Thus, deep learning play a role analyzing huge amount with high accuracy detect predict Because similarities characteristics stages Alzheimer’s, this study aimed extract features several methods integrate extracted from more than method same matrix. This development three methodologies, two systems, all systems at achieving satisfactory for AD predicting The first methodology by Feed Forward Neural Network (FFNN) GoogLeNet DenseNet-121 models separately. second FFNN network combined between Dense-121 after high-dimensionality reduction using Principal Component Analysis (PCA) algorithm. third separately Discrete Wavelet Transform (DWT), Local Binary Pattern (LBP) Gray Level Co-occurrence Matrix (GLCM) called handcrafted features. All yielded super results detecting With handcrafted, achieved an 99.7%, sensitivity 99.64%, AUC 99.56%, precision 99.63%, specificity 99.67%.

Language: Английский

---

Multi-Method Diagnosis of Histopathological Images for Early Detection of Breast Cancer Based on Hybrid and Deep Learning

Mathematics, Journal Year: 2023, Volume and Issue: 11(6), P. 1429 - 1429

Published: March 15, 2023

Breast cancer (BC) is a type of suffered by adult females worldwide. A late diagnosis BC leads to death, so early essential for saving lives. There are many methods diagnosing BC, including surgical open biopsy (SOB), which however constitutes an intense workload pathologists follow SOB and additionally takes long time. Therefore, artificial intelligence systems can help accurately earlier; it tool that assist doctors in making sound diagnostic decisions. In this study, two proposed approaches were applied, each with systems, diagnose dataset magnification factors (MF): 40×, 100×, 200×, 400×. The first method hybrid technology between CNN (AlexNet GoogLeNet) models extracts features classify them using the support vector machine (SVM). Thus, all datasets diagnosed AlexNet + SVM GoogLeNet SVM. second diagnoses ANN based on combining handcrafted extracted fuzzy color histogram (FCH), local binary pattern (LBP), gray level co-occurrence matrix (GLCM), collectively called fusion features. Finally, fed into neural network (ANN) classification. This has proven its superior ability histopathological images (HI) accurately. algorithm achieved results 100% metrics 400× dataset.

Language: Английский

---

Analysis of WSI Images by Hybrid Systems with Fusion Features for Early Diagnosis of Cervical Cancer

Diagnostics, Journal Year: 2023, Volume and Issue: 13(15), P. 2538 - 2538

Published: July 31, 2023

Cervical cancer is one of the most common types malignant tumors in women. In addition, it causes death latter stages. Squamous cell carcinoma and aggressive form cervical must be diagnosed early before progresses to a dangerous stage. Liquid-based cytology (LBC) swabs are best commonly used for screening converted from glass slides whole-slide images (WSIs) computer-assisted analysis. Manual diagnosis by microscopes limited prone manual errors, tracking all cells difficult. Therefore, development computational techniques important as diagnosing many samples can done automatically, quickly, efficiently, which beneficial medical laboratories professionals. This study aims develop automated WSI image analysis models squamous dataset. Several systems have been designed analyze accurately distinguish progression. For proposed systems, were optimized show contrast edges low-contrast cells. Then, analyzed segmented isolated rest using Active Contour Algorithm (ACA). hybrid method between deep learning (ResNet50, VGG19 GoogLeNet), Random Forest (RF), Support Vector Machine (SVM) algorithms based on ACA algorithm. Another RF SVM fused features deep-learning (DL) (ResNet50-VGG19, VGG19-GoogLeNet, ResNet50-GoogLeNet). It concluded systems' performance that DL models' combined help significantly improve networks. The novelty this research combines extracted ResNet50-GoogLeNet) with images. results demonstrate SVM. network ResNet50-VGG19 achieved an AUC 98.75%, sensitivity 97.4%, accuracy 99%, precision 99.6%, specificity 99.2%.

Language: Английский

---

Role of artificial intelligence in early diagnosis and treatment of infectious diseases

Infectious Diseases, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 26

Published: Nov. 14, 2024

Infectious diseases remain a global health challenge, necessitating innovative approaches for their early diagnosis and effective treatment. Artificial Intelligence (AI) has emerged as transformative force in healthcare, offering promising solutions to address this challenge. This review article provides comprehensive overview of the pivotal role AI can play treatment infectious diseases. It explores how AI-driven diagnostic tools, including machine learning algorithms, deep learning, image recognition systems, enhance accuracy efficiency disease detection surveillance. Furthermore, it delves into potential predict outbreaks, optimise strategies, personalise interventions based on individual patient data be used gear up drug discovery development (D3) process.The ethical considerations, challenges, limitations associated with integration management are also examined. By harnessing capabilities AI, healthcare systems significantly improve preparedness, responsiveness, outcomes battle against

Language: Английский

---

An efficient multi-stage ensemble deep learning framework for diagnosing infectious diseases

Decision Analytics Journal, Journal Year: 2024, Volume and Issue: 11, P. 100458 - 100458

Published: April 7, 2024

This study presents an efficient four-stage ensemble deep learning framework for diagnosing infectious diseases. The model is evaluated on three standard datasets. In our proposed transfer learning-based neural architecture (4s-min-FN), the images pass through four stages, each attempting to classify as positive. A negative class confirmed if every stage classifies image negative. (4S-min-FN) ensures minimization of false negatives. When new cases go a changing scenario, same modified (4S-min-FP) minimize positives following but with different transition rule. We use adaptive threshold setting in find proper trade-off between sensitivity, specificity, and good accuracy. well-known pre-trained architectures like Inception, ResNet-50, DenseNet-121, MobileNet experimental evaluation predicted class, which provided better insights about condition. can perform at par existing techniques terms accuracy while reducing negatives depending requirement.

Language: Английский

---

Tuberculosis Detection Comparison by Using Preprocessed and Non-preprocessed Chest X-Ray Images

IFMBE proceedings, Journal Year: 2025, Volume and Issue: unknown, P. 53 - 64

Published: Jan. 1, 2025

Language: Английский

---

Hybrid Models Based on Fusion Features of a CNN and Handcrafted Features for Accurate Histopathological Image Analysis for Diagnosing Malignant Lymphomas

Diagnostics, Journal Year: 2023, Volume and Issue: 13(13), P. 2258 - 2258

Published: July 4, 2023

Malignant lymphoma is one of the most severe types disease that leads to death as a result exposure lymphocytes malignant tumors. The transformation cells from indolent B-cell (DBCL) life-threatening. Biopsies taken patient are gold standard for analysis. Glass slides under microscope converted into whole slide images (WSI) be analyzed by AI techniques through biomedical image processing. Because multiplicity lymphomas, manual diagnosis pathologists difficult, tedious, and subject disagreement among physicians. importance artificial intelligence (AI) in early significant has revolutionized field oncology. use offers numerous benefits, including improved accuracy, faster diagnosis, risk stratification. This study developed several strategies based on hybrid systems analyze histopathological lymphomas. For all proposed models, extraction were optimized gradient vector flow (GVF) algorithm. first strategy diagnosing relied system between three deep learning (DL) networks, XGBoost algorithms, decision tree (DT) algorithms GVF second was fusing features MobileNet-VGG16, VGG16-AlexNet, MobileNet-AlexNet models classifying them DT ant colony optimization (ACO) color, shape, texture features, which called handcrafted extracted four traditional feature algorithms. similarity biological characteristics early-stage fused combined with classified ACO We concluded performance two networks DT, DL handcrafted, achieved best performance. network MobileNet-VGG16 resulted an AUC 99.43%, accuracy 99.8%, precision 99.77%, sensitivity 99.7%, specificity 99.8%. highlights role lymphoma, offering expedited enhanced leveraging processing; analysis biopsies allows systems, combining demonstrated promising results images. Furthermore, fusion classification models.

Language: Английский

---