Mpox-XDE: an ensemble model utilizing deep CNN and explainable AI for monkeypox detection and classification DOI Creative Commons
Dip Kumar Saha,

Sana Rafi,

M. F. Mridha

и другие.

BMC Infectious Diseases, Год журнала: 2025, Номер 25(1)

Опубликована: Март 25, 2025

The daily surge in cases many nations has made the growing number of human monkeypox (Mpox) an important global concern. Therefore, it is imperative to identify Mpox early prevent its spread. majority studies on identification have utilized deep learning (DL) models. However, research developing a reliable method for accurately detecting stages still lacking. This study proposes ensemble model composed three improved DL models more classify phases. We used widely recognized Skin Images Dataset (MSID), which includes 770 images. enhanced Swin Transformer (SwinViT), proposed Mpox-XDE, and modified models-Xception, DenseNet201, EfficientNetB7-were used. To generate model, were combined via Softmax layer, dense flattened 65% dropout. Four neurons final layer dataset into four categories: chickenpox, measles, normal, Mpox. Lastly, average pooling implemented actual class. Mpox-XDE performed exceptionally well, achieving testing accuracy, precision, recall, F1-score 98.70%, 98.90%, 98.80%, respectively. Finally, popular explainable artificial intelligence (XAI) technique, Gradient-weighted Class Activation Mapping (Grad-CAM), was applied convolutional overlaid areas that effectively highlight each illness class dataset. methodology will aid professionals diagnosing patient's condition.

Язык: Английский

Grad-CAM: Understanding AI Models DOI Open Access
Shuihua Wang,

Yudong Zhang

Computers, materials & continua/Computers, materials & continua (Print), Год журнала: 2023, Номер 76(2), С. 1321 - 1324

Опубликована: Янв. 1, 2023

Artificial intelligence; Grad-CAM; deep learning; convolutional neural networks; classification; location; explainable

Процитировано

14
Multiscale adaptive and attention-dilated convolutional neural network for efficient leukemia detection model with multiscale trans-res-Unet3+ -based segmentation network DOI

K Gokulkannan,

T A Mohanaprakash,

J. DafniRose

и другие.

Biomedical Signal Processing and Control, Год журнала: 2024, Номер 90, С. 105847 - 105847

Опубликована: Янв. 4, 2024

Язык: Английский

Процитировано

6
Hematologic cancer diagnosis and classification using machine and deep learning: State-of-the-art techniques and emerging research directives DOI
Hema J Patel,

Himal Shah,

Gayatri Patel

и другие.

Artificial Intelligence in Medicine, Год журнала: 2024, Номер 152, С. 102883 - 102883

Опубликована: Апрель 21, 2024

Язык: Английский

Процитировано

6
An Explainable Artificial Intelligence Integrated System for Automatic Detection of Dengue From Images of Blood Smears Using Transfer Learning DOI Creative Commons
Hilda Mayrose, Niranjana Sampathila, G. Muralidhar Bairy

и другие.

IEEE Access, Год журнала: 2024, Номер 12, С. 41750 - 41762

Опубликована: Янв. 1, 2024

Dengue fever is a rapidly increasing mosquito-borne ailment spread by the virus DENV in tropics and subtropics worldwide. It significant public health problem accounts for many deaths globally. Implementing more effective methods that can accurately detect dengue cases challenging. The theme of this digital pathology-associated research automatic detection from peripheral blood smears (PBS) employing deep learning (DL) techniques. In recent years, DL has been significantly employed automated computer-assisted diagnosis various diseases medical images. This paper explores pre-trained convolution neural networks (CNNs) detection. Transfer (TL) executed on three state-of-the-art CNNs – ResNet50, MobileNetV3Small, MobileNetV3Large, to customize models differentiating dengue-infected healthy ones. dataset used design test contains 100x magnified control microscopic PBS are validated with 5-fold cross-validation framework tested unseen data. An explainable artificial intelligence (XAI) approach, Gradient-weighted Class Activation Mapping (GradCAM), eventually applied allow visualization precise regions most instrumental making predictions. While all transferred CNN performed well (above 98% overall classification accuracy), MobileNetV3Small recommended model due its less computationally demanding characteristics. Transferred based yielded Accuracy, Recall, Specificity, Precision, F1 Score, Area Under ROC Curve (AUC) 0.982 ± 0.011, 0.973 0.027, 0.99 0.013, 0.989 0.015, 0.981 0.012 respectively, averaged over five folds dataset. Promising results show developed have potential provide high-quality support haematologists expertly performing tedious, repetitive, time-consuming tasks hospitals remote/low-resource settings.

Язык: Английский

Процитировано

5
Explainable artificial intelligence for medical imaging: Review and experiments with infrared breast images DOI
Kaushik Raghavan, S. Balasubramanian,

V. Kamakoti

и другие.

Computational Intelligence, Год журнала: 2024, Номер 40(3)

Опубликована: Июнь 1, 2024

Abstract There is a growing trend of using artificial intelligence, particularly deep learning algorithms, in medical diagnostics, revolutionizing healthcare by improving efficiency, accuracy, and patient outcomes. However, the use intelligence diagnostics comes with critical need to explain reasoning behind intelligence‐based predictions ensure transparency decision‐making. Explainable has emerged as crucial research area address for interpretability diagnostics. techniques aim provide insights into decision‐making process systems, enabling clinicians understand factors algorithms consider reaching their predictions. This paper presents detailed review saliency‐based (visual) methods, such class activation which have gained popularity imaging they visual explanations highlighting regions an image most influential intelligence's decision. We also present literature on non‐visual but focus will be methods. existing experiment infrared breast images detecting cancer. Towards end this paper, we propose “attention guided Grad‐CAM” that enhances visualizations explainable intelligence. The shows are not explored context opens up wide range opportunities further make clinical thermography assistive technology community.

Язык: Английский

Процитировано

5
Automatic Detection of Acute Leukemia (ALL and AML) Utilizing Customized Deep Graph Convolutional Neural Networks DOI Creative Commons

Lida Zare,

Mahsan Rahmani,

Nastaran Khaleghi

и другие.

Bioengineering, Год журнала: 2024, Номер 11(7), С. 644 - 644

Опубликована: Июнь 24, 2024

Leukemia is a malignant disease that impacts explicitly the blood cells, leading to life-threatening infections and premature mortality. State-of-the-art machine-enabled technologies sophisticated deep learning algorithms can assist clinicians in early-stage diagnosis. This study introduces an advanced end-to-end approach for automated diagnosis of acute leukemia classes lymphocytic (ALL) myeloid (AML). gathered complete database 44 patients, comprising 670 ALL AML images. The proposed model’s architecture consisted fusion graph theory convolutional neural network (CNN), with six Conv layers Softmax layer. model achieved classification accuracy 99% kappa coefficient 0.85 classes. suggested was assessed noisy conditions demonstrated strong resilience. Specifically, remained above 90%, even at signal-to-noise ratio (SNR) 0 dB. evaluated against contemporary methodologies research, demonstrating encouraging outcomes. According this, serve as tool identify specific forms leukemia.

Язык: Английский

Процитировано

4
Leukemia detection and classification using computer-aided diagnosis system with falcon optimization algorithm and deep learning DOI Creative Commons
Turky Omar Asar, Mahmoud Ragab

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Сен. 18, 2024

Язык: Английский

Процитировано

4
Revolutionizing chronic lymphocytic leukemia diagnosis: A deep dive into the diverse applications of machine learning DOI Creative Commons
Mohamed Elhadary, Amgad M. Elshoeibi, Ahmed Badr

и другие.

Blood Reviews, Год журнала: 2023, Номер 62, С. 101134 - 101134

Опубликована: Сен. 22, 2023

Chronic lymphocytic leukemia (CLL) is a B cell neoplasm characterized by the accumulation of aberrant monoclonal lymphocytes. CLL predominant type in Western countries, accounting for 25% cases. Although many patients remain asymptomatic, subset may exhibit typical lymphoma symptoms, acquired immunodeficiency disorders, or autoimmune complications. Diagnosis involves blood tests showing increased lymphocytes and further examination using peripheral smear flow cytometry to confirm disease. With significant advancements machine learning (ML) artificial intelligence (AI) recent years, numerous models algorithms have been proposed support diagnosis classification CLL. In this review, we discuss benefits drawbacks applications ML evaluation diagnosed with

Язык: Английский

Процитировано

9
Transformative Advances in AI for Precise Cancer Detection: A Comprehensive Review of Non-Invasive Techniques DOI
Hari Mohan, Joon Yoo, Serhii Dashkevych

и другие.

Archives of Computational Methods in Engineering, Год журнала: 2025, Номер unknown

Опубликована: Янв. 11, 2025

Язык: Английский

Процитировано

0
Artificial intelligence for the detection of acute myeloid leukemia from microscopic blood images; a systematic review and meta-analysis DOI Creative Commons
Feras Al‐Obeidat, Wael Hafez, Asrar Rashid

и другие.

Frontiers in Big Data, Год журнала: 2025, Номер 7

Опубликована: Янв. 17, 2025

Leukemia is the 11th most prevalent type of cancer worldwide, with acute myeloid leukemia (AML) being frequent malignant blood malignancy in adults. Microscopic tests are common methods for identifying subtypes. An automated optical image-processing system using artificial intelligence (AI) has recently been applied to facilitate clinical decision-making. To evaluate performance all AI-based approaches detection and diagnosis (AML). Medical databases including PubMed, Web Science, Scopus were searched until December 2023. We used "metafor" "metagen" libraries R analyze different models studies. Accuracy sensitivity primary outcome measures. Ten studies included our review meta-analysis, conducted between 2016 Most deep-learning have utilized, convolutional neural networks (CNNs). The common- random-effects had accuracies 1.0000 [0.9999; 1.0001] 0.9557 [0.9312, 0.9802], respectively. random effects high values 0.8581, respectively, indicating that machine learning this study can accurately detect true-positive cases. Studies shown substantial variations accuracy sensitivity, as by Q I2 statistics. Our systematic meta-analysis found an overall AI correctly AML Future research should focus on unifying reporting assessment metrics diagnostics. https://www.crd.york.ac.uk/prospero/#recordDetails, CRD42024501980.

Язык: Английский

Процитировано

0