Medical Images Encryption Based on Adaptive-Robust Multi-Mode Synchronization of Chen Hyper-Chaotic Systems

Sensors, Journal Year: 2021, Volume and Issue: 21(11), P. 3925 - 3925

Published: June 7, 2021

In this paper, a novel medical image encryption method based on multi-mode synchronization of hyper-chaotic systems is presented. The great significance in secure communication tasks such as images. Multi-mode and highly complex issue, especially if there uncertainty disturbance. work, an adaptive-robust controller designed for multimode synchronized chaotic with variable unknown parameters, despite the bounded disturbance known function two modes. first case, it main system some response systems, second circular synchronization. Using theorems proved that methods are equivalent. Our results show that, we able to obtain convergence error parameter estimation zero using Lyapunov’s method. new laws update time-varying estimating bounds proposed stability guaranteed. To assess performance method, various statistical analyzes were carried out encrypted images standard benchmark effective technique telemedicine application.

Language: Английский

---

Application of artificial intelligence techniques for automated detection of myocardial infarction: a review

Physiological Measurement, Journal Year: 2022, Volume and Issue: 43(8), P. 08TR01 - 08TR01

Published: July 8, 2022

Objective.Myocardial infarction (MI) results in heart muscle injury due to receiving insufficient blood flow. MI is the most common cause of mortality middle-aged and elderly individuals worldwide. To diagnose MI, clinicians need interpret electrocardiography (ECG) signals, which requires expertise subject observer bias. Artificial intelligence-based methods can be utilized screen for or automatically using ECG signals.Approach.In this work, we conducted a comprehensive assessment artificial approaches detection based on some other biophysical including machine learning (ML) deep (DL) models. The performance traditional ML relies handcrafted features manual selection whereas DL models automate these tasks.Main results.The review observed that convolutional neural networks (DCNNs) yielded excellent classification diagnosis, explains why they have become prevalent recent years.Significance.To our knowledge, first survey intelligence techniques employed diagnosis signals.

Language: Английский

---

Breast Cancer Dataset, Classification and Detection Using Deep Learning

Healthcare, Journal Year: 2022, Volume and Issue: 10(12), P. 2395 - 2395

Published: Nov. 29, 2022

Incorporating scientific research into clinical practice via informatics, which includes genomics, proteomics, bioinformatics, and biostatistics, improves patients' treatment. Computational pathology is a growing subspecialty with the potential to integrate whole slide images, multi-omics data, health informatics. Pathology laboratory medicine are critical diagnosing cancer. This work will review existing computational digital methods for breast cancer diagnosis special focus on deep learning. The paper starts by reviewing public datasets related diagnosis. Additionally, learning reviewed. publicly available code repositories introduced as well. closed highlighting challenges future works learning-based

Language: Английский

---

The role of artificial intelligence in cardiovascular magnetic resonance imaging

Progress in Cardiovascular Diseases, Journal Year: 2024, Volume and Issue: 86, P. 13 - 25

Published: June 24, 2024

Language: Английский

---

Automated assessment of cardiac pathologies on cardiac MRI using T1-mapping and late gadolinium phase sensitive inversion recovery sequences with deep learning

BMC Medical Imaging, Journal Year: 2024, Volume and Issue: 24(1)

Published: Feb. 13, 2024

Abstract Background A deep learning (DL) model that automatically detects cardiac pathologies on MRI may help streamline the diagnostic workflow. To develop a DL to detect T1-mapping and late gadolinium phase sensitive inversion recovery (PSIR) sequences were used. Methods Subjects in this study either diagnosed with pathology ( n = 137) including acute chronic myocardial infarction, myocarditis, dilated cardiomyopathy, hypertrophic cardiomyopathy or classified as normal 63). Cardiac MR imaging included PSIR sequences. split 65/15/20% for training, validation, hold-out testing. The models based an ImageNet pretrained DenseNet-161 implemented using PyTorch fastai. Data augmentation random rotation mixup was applied. Categorical cross entropy used loss function cyclic rate (1e-3). both developed separately similar training parameters. final chosen its performance validation set. Gradient-weighted class activation maps (Grad-CAMs) visualized decision-making process of model. Results achieved sensitivity, specificity, accuracy 100%, 38%, 88% images 78%, 54%, 70% images. Grad-CAMs demonstrated focused attention myocardium when evaluating Conclusions able reliably T1 mapping alone is particularly note since it does not require contrast agent can be acquired quickly.

Language: Английский

---

Hybrid HCNN-KNN Model Enhances Age Estimation Accuracy in Orthopantomography

Frontiers in Public Health, Journal Year: 2022, Volume and Issue: 10

Published: May 30, 2022

Age estimation in dental radiographs Orthopantomography (OPG) is a medical imaging technique that physicians and pathologists utilize for disease identification legal matters. For example, estimating post-mortem interval, detecting child abuse, drug trafficking, identifying an unknown body. Recent development automated image processing models improved the age estimation's limited precision to approximate range of +/- 1 year. While this often accepted as accurate measurement, should be precise possible most serious matters, such homicide. Current techniques are highly dependent on manual time-consuming processing. time-sensitive matter which time vital. Machine learning-based data methods has decreased processing; however, accuracy these remains further improved. We proposed ensemble method classifiers enhance using OPGs from year couple months (1-3-6). This hybrid model based convolutional neural networks (CNN) K nearest neighbors (KNN). The (HCNN-KNN) was used investigate 1,922 panoramic patients aged 15 23. These were obtained various teaching institutes private clinics Malaysia. To minimize chance overfitting our model, we principal component analysis (PCA) algorithm eliminated features with high correlation. performance performed systematic pre-processing. applied series classifications train model. have successfully demonstrated combining innovative approaches classification segmentation thus age-estimation outcome Our findings suggest first time, best knowledge, estimated classified studies old, 6 months, 3 1-month-old cases accuracies 99.98, 99.96, 99.87, 98.78 respectively.

Language: Английский

---

Enhancing Myocarditis Diagnosis Through Deep Learning and Data Augmentation: A Novel Framework Addressing Imbalance and Initialization Sensitivity

Web Intelligence, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 12, 2025

Myocarditis poses a serious public health risk, with the potential to cause heart failure and sudden death. Traditionally, diagnosing myocarditis relies on non-invasive imaging, particularly cardiac magnetic resonance imaging (MRI), though MRI results can be vulnerable operator bias. Our research addresses this by introducing an innovative deep-learning framework tackle challenges frequently overlooked in past studies, including class imbalance, sensitivity initial weight settings, generalizability. model leverages convolutional neural networks (CNNs) extract detailed feature vectors for highly precise classifying of myocarditis. Since imbalance problem is frequent many training datasets, we will adopt reinforcement learning (RL) strategy shift more emphasis underrepresented classes balanced learning. Additionally, our involves mutual learning-based artificial bee colony (ML-ABC) algorithm efficient pretraining weights. Improve data diversity volume further using online augmentation improved version generative adversarial network (GAN). We enhance performance generator considering information provided features produced discriminator which base its output making it realistic, hence increasing accuracy generator. model, when applied Z-Alizadeh Sani dataset, reaches 90.8%, outperforming previously reported techniques reiterating feasibility clinical purposes. These significantly advance early detection open new avenues enhanced treatment strategies.

Language: Английский

---

Enhancing MRI diagnosis of myocarditis using deep learning and generative adversarial networks

Applied Mathematics and Nonlinear Sciences, Journal Year: 2025, Volume and Issue: 10(1)

Published: Jan. 1, 2025

Abstract In this paper, in order to enhance the MRI diagnosis of myocarditis, a generative adversarial network (GAN)-based diagnostic model for myocarditis is constructed paper. The images provided by hospital were used as data source study, and image format was transformed into NII file saving using Python tool, which uniformly cropped 480×768 pixels, stored form datasets, divided dataset A (the MRI-weighted dataset) B myocarditis). ResNet-34 U-Net generator discriminator, respectively, address problem difficulty training GAN networks, BN layer added between convolutional activation function construction finally completed. Determine loss function, select quantitative evaluation indexes (MAE, RMSE, PSNR, SSIM PCC), set control (CNN, RNN, LSTM, GRU), validate analyze discriminator after 400 iterations training, value both almost 0. paper’s genus pig are higher than other four models. summary, has facilitating effect on myocarditis.

Language: Английский

---

Review of the Current State of Artificial Intelligence in Pediatric Cardiovascular Magnetic Resonance Imaging

Children, Journal Year: 2025, Volume and Issue: 12(4), P. 416 - 416

Published: March 26, 2025

Cardiovascular magnetic resonance (CMR) imaging is essential for the management of congenital heart disease (CHD), due to ability perform anatomic and physiologic assessments patients. However, CMR scans can be time-consuming analyze, creating roadblocks broader use in CHD. Recent publications have shown artificial intelligence (AI) has potential increase efficiency, improve image quality, reduce errors. This review examines AI techniques CHD, by focusing on deep learning applied acquisition reconstruction, processing reporting, clinical cases, future directions.

Language: Английский

---

Integrative analysis of mutated genes and mutational processes reveals novel mutational biomarkers in colorectal cancer

BMC Bioinformatics, Journal Year: 2022, Volume and Issue: 23(1)

Published: April 19, 2022

Colorectal cancer (CRC) is one of the leading causes cancer-related deaths worldwide. Recent studies have observed causative mutations in susceptible genes related to colorectal 10 15% patients. This highlights importance identifying for early detection this more effective treatments among high risk individuals. Mutation considered as key point research. Many performed subtyping based on type frequently mutated genes, or proportion mutational processes. However, best our knowledge, combination these features has never been used together task. potential introduce better and inclusive subtype classification approaches using wider range enable biomarker discovery thus inform drug development CRC.In study, we develop a new pipeline novel concept called 'gene-motif', which merges gene information with tri-nucleotide motif sites, identification. We apply International Cancer Genome Consortium (ICGC) CRC samples identify, first time, 3131 gene-motif combinations that are significantly 536 ICGC samples. Using features, identify seven subtypes distinguishable phenotypes biomarkers, including unique signaling pathways, most them targeted treatment options currently available. Interestingly, also several multiple but sequence contexts.Our results highlight considering both mutation identification biomarkers. The presented study demonstrates distinguished phenotypic properties can be effectively treatments. By knowing associated subtypes, personalized plan developed considers specific their genomic lesion.

Language: Английский

---