Advancing neural computation: experimental validation and optimization of dendritic learning in feedforward tree networks DOI
Seyed‐Ali Sadegh‐Zadeh,

Pooya Hazegh

American Journal of Neurodegenerative Disease, Journal Year: 2024, Volume and Issue: 13(5), P. 49 - 69

Published: Jan. 1, 2024

This study aims to explore the capabilities of dendritic learning within feedforward tree networks (FFTN) in comparison traditional synaptic plasticity models, particularly context digit recognition tasks using MNIST dataset. We employed FFTNs with nonlinear segment amplification and Hebbian rules enhance computational efficiency. The dataset, consisting 70,000 images handwritten digits, was used for training testing. Key performance metrics, including accuracy, precision, recall, F1-score, were analysed. models significantly outperformed plasticity-based across all metrics. Specifically, framework achieved a test accuracy 91%, compared 88% demonstrating superior classification. Dendritic offers more powerful by closely mimicking biological neural processes, providing enhanced efficiency scalability. These findings have important implications advancing both artificial intelligence systems neuroscience.

Language: Английский

Artificial Intelligence in Audiology: A Scoping Review of Current Applications and Future Directions DOI Creative Commons
Andrea Frosolini, Leonardo Franz, Valeria Caragli

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(22), P. 7126 - 7126

Published: Nov. 6, 2024

The integration of artificial intelligence (AI) into medical disciplines is rapidly transforming healthcare delivery, with audiology being no exception. By synthesizing the existing literature, this review seeks to inform clinicians, researchers, and policymakers about potential challenges integrating AI audiological practice. PubMed, Cochrane, Google Scholar databases were searched for articles published in English from 1990 2024 following query: "(audiology) AND ("artificial intelligence" OR "machine learning" "deep learning")". PRISMA extension scoping reviews (PRISMA-ScR) was followed. database research yielded 1359 results, selection process led inclusion 104 manuscripts. has evolved significantly over succeeding decades, 87.5% manuscripts last 4 years. Most types consistently used specific purposes, such as logistic regression other statistical machine learning tools (e.g., support vector machine, multilayer perceptron, random forest, deep belief network, decision tree, k-nearest neighbor, or LASSO) automated audiometry clinical predictions; convolutional neural networks radiological image analysis; large language models automatic generation diagnostic reports. Despite advances technologies, different ethical professional are still present, underscoring need larger, more diverse data collection bioethics studies field audiology.

Language: Английский

Citations

6
Machine Learning Models Can Predict Tinnitus and Noise-Induced Hearing Loss DOI
Zahra Jafari,

Ryan Harari,

Glenn Hole

et al.

Ear and Hearing, Journal Year: 2025, Volume and Issue: unknown

Published: May 6, 2025

Objectives: Despite the extensive use of machine learning (ML) models in health sciences for outcome prediction and condition classification, their application differentiating various types auditory disorders remains limited. This study aimed to address this gap by evaluating efficacy five ML distinguishing (a) individuals with tinnitus from those without (b) noise-induced hearing loss (NIHL) age-related (ARHL). Design: We used data a cross-sectional Canadian population, which included audiologic demographic information 928 adults aged 30 100 years, diagnosed either ARHL or NIHL due long-term occupational noise exposure. The applied were artificial neural networks (ANNs), K-nearest neighbors, logistic regression, random forest (RF), support vector machines. Results: revealed that prevalence was over twice as high group compared group, frequency 27.85% versus 8.85% constant 18.55% 10.86% intermittent tinnitus. In pattern recognition, significantly greater found at medium- high-band frequencies ARHL. both ARHL, showed better pure-tone sensitivity than Among models, ANN achieved highest overall accuracy (70%), precision (60%), F1-score (87%) predicting tinnitus, an area under curve 0.71. RF outperformed other (79% NIHL, 85% ARHL), recall (85% NIHL), (81% (0.90). Conclusions: Our findings highlight particularly RF, advancing diagnostic potentially providing framework integrating techniques into clinical audiology improved precision. Future research is suggested expand datasets include diverse populations integrate longitudinal data.

Language: Английский

Citations

0
Immunoglobulin G4: Cross Talk in Hearing Loss Manifestation DOI
Arnavaz Hajizadeh Barfejani,

Abolfazl Ghobadi

Indian Journal of Otology, Journal Year: 2025, Volume and Issue: 31(1), P. 10 - 16

Published: Jan. 1, 2025

Abstract Immunoglobulin G4-related disease (IgG4-RD) is a systemic immune-mediated condition characterized by tissue infiltration with IgG4-positive plasma cells and elevated serum IgG4 levels. While IgG4-RD can affect multiple organs, its involvement in the auditory system, leading to hearing loss, less frequent but clinically significant manifestation. This review comprehensively examines underlying pathophysiology, diagnostic techniques, management options for loss related IgG4-RD. The pathogenesis involves complex dysregulation of B- T-cell responses, resulting chronic inflammation fibrosis affected tissues. Diagnosis typically requires combination clinical presentation, levels, imaging studies, histopathological findings. Treatment primarily consists corticosteroids, immunosuppressive agents like rituximab considered refractory cases. highlights importance early diagnosis appropriate prevent long-term complications improve patient outcomes. By increasing clinicians’ awareness IgG4-related otological diseases, this aims enhance understanding facilitate better care patients.

Language: Английский

Citations

0
Comparative analysis of dimensionality reduction techniques for EEG-based emotional state classification DOI
Seyed‐Ali Sadegh‐Zadeh, Nasrin Sadeghzadeh,

Ommolbanin Soleimani

et al.

American Journal of Neurodegenerative Disease, Journal Year: 2024, Volume and Issue: 13(4), P. 23 - 33

Published: Jan. 1, 2024

The aim of this study is to evaluate the impact various dimensionality reduction methods, including principal component analysis (PCA), Laplacian score, and Chi-square feature selection, on classification performance an electroencephalogram (EEG) dataset.

Language: Английский

Citations

2
Neural reshaping: the plasticity of human brain and artificial intelligence in the learning process DOI
Seyed‐Ali Sadegh‐Zadeh, Mahboobe Bahrami,

Ommolbanin Soleimani

et al.

American Journal of Neurodegenerative Disease, Journal Year: 2024, Volume and Issue: 13(5), P. 34 - 48

Published: Jan. 1, 2024

This study explores the concept of neural reshaping and mechanisms through which both human artificial intelligence adapt learn. To investigate parallels distinctions between brain plasticity network plasticity, with a focus on their learning processes. A comparative analysis was conducted using literature reviews machine experiments, specifically employing multi-layer perceptron to examine regression classification problems. Experimental findings demonstrate that models, similar neuroplasticity, enhance performance iterative optimization, drawing in strengthening adjusting connections. Understanding shared principles limitations can drive advancements AI design cognitive neuroscience, paving way for future interdisciplinary innovations.

Language: Английский

Citations

2
ChatGPT-4 extraction of heart failure symptoms and signs from electronic health records DOI
T. Elizabeth Workman, Ali M. Ahmed,

Helen Sheriff

et al.

Progress in Cardiovascular Diseases, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 1, 2024

Language: Английский

Citations

0
Advancing neural computation: experimental validation and optimization of dendritic learning in feedforward tree networks DOI
Seyed‐Ali Sadegh‐Zadeh,

Pooya Hazegh

American Journal of Neurodegenerative Disease, Journal Year: 2024, Volume and Issue: 13(5), P. 49 - 69

Published: Jan. 1, 2024

This study aims to explore the capabilities of dendritic learning within feedforward tree networks (FFTN) in comparison traditional synaptic plasticity models, particularly context digit recognition tasks using MNIST dataset. We employed FFTNs with nonlinear segment amplification and Hebbian rules enhance computational efficiency. The dataset, consisting 70,000 images handwritten digits, was used for training testing. Key performance metrics, including accuracy, precision, recall, F1-score, were analysed. models significantly outperformed plasticity-based across all metrics. Specifically, framework achieved a test accuracy 91%, compared 88% demonstrating superior classification. Dendritic offers more powerful by closely mimicking biological neural processes, providing enhanced efficiency scalability. These findings have important implications advancing both artificial intelligence systems neuroscience.

Language: Английский

Citations

0