Neural Decoding of EEG Signals with Machine Learning: A Systematic Review

Brain Sciences, Journal Year: 2021, Volume and Issue: 11(11), P. 1525 - 1525

Published: Nov. 18, 2021

Electroencephalography (EEG) is a non-invasive technique used to record the brain's evoked and induced electrical activity from scalp. Artificial intelligence, particularly machine learning (ML) deep (DL) algorithms, are increasingly being applied EEG data for pattern analysis, group membership classification, brain-computer interface purposes. This study aimed systematically review recent advances in ML DL supervised models decoding classifying signals. Moreover, this article provides comprehensive of state-of-the-art techniques signal preprocessing feature extraction. To end, several academic databases were searched explore relevant studies year 2000 present. Our results showed that application both mental workload motor imagery tasks has received substantial attention years. A total 75% convolutional neural networks with various 36% achieved competitive accuracy by using support vector algorithm. Wavelet transform was found be most common extraction method all types tasks. We further examined specific methods end classifier recommendations discovered systematic review.

Language: Английский

---

Exploiting pretrained CNN models for the development of an EEG-based robust BCI framework

Computers in Biology and Medicine, Journal Year: 2022, Volume and Issue: 143, P. 105242 - 105242

Published: Jan. 24, 2022

Language: Английский

---

Mind the gap: State-of-the-art technologies and applications for EEG-based brain–computer interfaces

APL Bioengineering, Journal Year: 2021, Volume and Issue: 5(3)

Published: July 20, 2021

Brain–computer interfaces (BCIs) provide bidirectional communication between the brain and output devices that translate user intent into function. Among different imaging techniques used to operate BCIs, electroencephalography (EEG) constitutes preferred method of choice, owing its relative low cost, ease use, high temporal resolution, noninvasiveness. In recent years, significant progress in wearable technologies computational intelligence has greatly enhanced performance capabilities EEG-based BCIs (eBCIs) propelled their migration out laboratory real-world environments. This rapid translation a paradigm shift human–machine interaction will deeply transform industries near future, including healthcare wellbeing, entertainment, security, education, marketing. this contribution, state-of-the-art biosensing is reviewed, focusing on development novel electrode for long term noninvasive EEG monitoring. Commercially available platforms are surveyed, comparative analysis presented based benefits limitations they eBCI development. Emerging applications neuroscientific research future trends related widespread implementation eBCIs medical nonmedical uses discussed. Finally, commentary ethical, social, legal concerns associated with increasingly ubiquitous technology provided, as well general recommendations address key issues mainstream consumer adoption.

Language: Английский

---

A Framework for Lung and Colon Cancer Diagnosis via Lightweight Deep Learning Models and Transformation Methods

Diagnostics, Journal Year: 2022, Volume and Issue: 12(12), P. 2926 - 2926

Published: Nov. 23, 2022

Among the leading causes of mortality and morbidity in people are lung colon cancers. They may develop concurrently organs negatively impact human life. If cancer is not diagnosed its early stages, there a great likelihood that it will spread to two organs. The histopathological detection such malignancies one most crucial components effective treatment. Although process lengthy complex, deep learning (DL) techniques have made feasible complete more quickly accurately, enabling researchers study lot patients short time period for less cost. Earlier studies relied on DL models require computational ability resources. Most them depended individual extract features high dimension or perform diagnoses. However, this study, framework based multiple lightweight proposed utilizes several transformation methods feature reduction provide better representation data. In context, histopathology scans fed into ShuffleNet, MobileNet, SqueezeNet models. number acquired from these subsequently reduced using principal component analysis (PCA) fast Walsh-Hadamard transform (FHWT) techniques. Following that, discrete wavelet (DWT) used fuse FWHT's obtained three Additionally, models' PCA concatenated. Finally, diminished as result FHWT-DWT fusion processes four distinct machine algorithms, reaching highest accuracy 99.6%. results show can distinguish variants with lower complexity compared existing methods. also prove utilizing reduce offer superior interpretation data, thus improving diagnosis procedure.

Language: Английский

---

RADIC:A tool for diagnosing COVID-19 from chest CT and X-ray scans using deep learning and quad-radiomics

Chemometrics and Intelligent Laboratory Systems, Journal Year: 2023, Volume and Issue: 233, P. 104750 - 104750

Published: Jan. 2, 2023

Language: Английский

---

An End-to-End Lightweight Multi-Scale CNN for the Classification of Lung and Colon Cancer with XAI Integration

Technologies, Journal Year: 2024, Volume and Issue: 12(4), P. 56 - 56

Published: April 21, 2024

To effectively treat lung and colon cancer save lives, early accurate identification is essential. Conventional diagnosis takes a long time requires the manual expertise of radiologists. The rising number new cases makes it challenging to process massive volumes data quickly. Different machine learning approaches classification detection have been proposed by multiple research studies. However, when comes self-learning tasks, deep (DL) excels. This paper suggests novel DL convolutional neural network (CNN) model for detecting cancer. lightweight multi-scale since uses only 1.1 million parameters, making appropriate real-time applications as provides an end-to-end solution. By incorporating features extracted at scales, can capture both local global patterns within input data. explainability tools such gradient-weighted class activation mapping Shapley additive explanation identify potential problems highlighting specific areas that impact on model’s choice. experimental findings demonstrate detection, was outperformed competition accuracy rates 99.20% achieved multi-class (containing five classes) predictions.

Language: Английский

---

MB-AI-His: Histopathological Diagnosis of Pediatric Medulloblastoma and its Subtypes via AI

Diagnostics, Journal Year: 2021, Volume and Issue: 11(2), P. 359 - 359

Published: Feb. 20, 2021

Medulloblastoma (MB) is a dangerous malignant pediatric brain tumor that could lead to death. It considered the most common cancerous tumor. Precise and timely diagnosis of MB its four subtypes (defined by World Health Organization (WHO)) essential decide appropriate follow-up plan suitable treatments prevent progression reduce mortality rates. Histopathology gold standard modality for subtypes, but manual via pathologist very complicated, needs excessive time, subjective pathologists' expertise skills, which may variability in or misdiagnosis. The main purpose paper propose time-efficient reliable computer-aided (CADx), namely MB-AI-His, automatic from histopathological images. challenge this work lack datasets available limited related work. Related studies are based on either textural analysis deep learning (DL) feature extraction methods. These used individual features perform classification task. However, MB-AI-His combines benefits DL techniques methods through cascaded manner. First, it uses three convolutional neural networks (CNNs), including DenseNet-201, MobileNet, ResNet-50 CNNs extract spatial features. Next, extracts time-frequency discrete wavelet transform (DWT), method. Finally, fuses spatial-time-frequency generated DWT using cosine (DCT) principal component (PCA) produce CADx system. merges privileges different CNN architectures. has binary level classifying among normal abnormal images, multi-classification classify MB. results show accurate both multi-class levels. also system as PCA DCT have efficiently reduced training execution time. performance compared with systems, comparison verified powerfulness outperforming results. Therefore, can support pathologists time cost procedure will correspondingly lower death

Language: Английский

---

DIAROP: Automated Deep Learning-Based Diagnostic Tool for Retinopathy of Prematurity

Diagnostics, Journal Year: 2021, Volume and Issue: 11(11), P. 2034 - 2034

Published: Nov. 3, 2021

Retinopathy of Prematurity (ROP) affects preterm neonates and could cause blindness. Deep Learning (DL) can assist ophthalmologists in the diagnosis ROP. This paper proposes an automated reliable diagnostic tool based on DL techniques called DIAROP to support ophthalmologic It extracts significant features by first obtaining spatial from four Convolution Neural Networks (CNNs) using transfer learning then applying Fast Walsh Hadamard Transform (FWHT) integrate these features. Moreover, explores best-integrated extracted CNNs that influence its capability. The results indicate achieved accuracy 93.2% area under receiving operating characteristic curve (AUC) 0.98. Furthermore, performance is compared with recent ROP tools. Its promising shows may

Language: Английский

---

Shallow Convolutional Network Excel for Classifying Motor Imagery EEG in BCI Applications

IEEE Access, Journal Year: 2021, Volume and Issue: 9, P. 98275 - 98286

Published: Jan. 1, 2021

Many studies applying Brain-Computer Interfaces (BCIs) based on Motor Imagery (MI) tasks for rehabilitation have demonstrated the important role of detecting Event-Related Desynchronization (ERD) to recognize user's motor intention. Nowadays, development MI-based BCI approaches without or with very few calibration stages session-by-session different days weeks is still an open and emergent scope. In this work, a new scheme proposed by Convolutional Neural Networks (CNN) MI classification, using end-to-end Shallow architecture that contains two convolutional layers temporal spatial feature extraction. We hypothesize designed capturing event-related desynchronization/synchronization (ERD/ERS) at CNN input, adequate network design, may enhance classification fewer stages. The system same was tested three public datasets through multiple experiments, including both subject-specific non-subject-specific training. Comparable also superior results respect state-of-the-art were obtained. On subjects whose EEG data never used in training process, our achieved promising existing BCIs, which shows greater progress facilitating clinical applications.

Language: Английский

---

Histo-CADx: duo cascaded fusion stages for breast cancer diagnosis from histopathological images

PeerJ Computer Science, Journal Year: 2021, Volume and Issue: 7, P. e493 - e493

Published: April 27, 2021

Breast cancer (BC) is one of the most common types that affects females worldwide. It may lead to irreversible complications and even death due late diagnosis treatment. The pathological analysis considered gold standard for BC detection, but it a challenging task. Automatic could reduce rates, by creating computer aided (CADx) system capable accurately identifying at an early stage decreasing time consumed pathologists during examinations. This paper proposes novel CADx named Histo-CADx automatic BC. Most related studies were based on individual deep learning methods. Also, did not examine influence fusing features from multiple CNNs handcrafted features. In addition, investigate best combination fused performance CADx. Therefore, two stages fusion. first fusion involves investigation impact several (DL) techniques with feature extraction methods using auto-encoder DL method. also examines searches suitable set improve Histo-CADx. second constructs classifier (MCS) outputs three classifiers, further accuracy proposed evaluated public datasets; specifically, BreakHis ICIAR 2018 datasets. results both datasets verified successfully improved compared constructed Furthermore, process has reduced computation cost system. Moreover, after confirmed reliable capacity classifying more other latest studies. Consequently, can be used help them in accurate decrease effort needed medical experts examination.

Language: Английский

---