Evaluating the Effectiveness of Synthetic Datasets for Dementia Diagnosis Using Deep Learning DOI

Andrew Romitti,

Jiya Shetty,

Praveen Rao

et al.

Published: Sept. 27, 2023

Early and accurate diagnosis of dementia can lead to better treatment the disease improve patients' quality life. Advanced neuroimaging technologies such as magnetic resonance imaging (MRI) deep learning hold promise for early diagnosis. However, there is limited number real-world MRI datasets training deep-learning models classify a patient's degree dementia. Generative adversarial networks (GANs) are learning-based generative that generate synthetic data samples based on real dataset's distribution. They have been successfully used in clinical studies. In this work, we investigate how images generated by GANs performance accurately classifying level (i.e., very mildly demented, moderately no dementia.) We trained state-of-the-art model image classification, namely, Data-Efficient Image Transformer (DeiT) using dataset along with GANs. combined during varying proportion set. evaluated accuracy F1-score DeiT images. Our results showed achieve good even Hence, offer promising solution improving via especially when scarce.

Language: Английский

Predicting early Alzheimer’s with blood biomarkers and clinical features DOI Creative Commons

Muaath Ebrahim AlMansoori,

Sherlyn Jemimah,

Ferial Abuhantash

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: March 13, 2024

Abstract Alzheimer’s disease (AD) is an incurable neurodegenerative disorder that leads to dementia. This study employs explainable machine learning models detect dementia cases using blood gene expression, single nucleotide polymorphisms (SNPs), and clinical data from Disease Neuroimaging Initiative (ADNI). Analyzing 623 ADNI participants, we found the Support Vector Machine classifier with Mutual Information (MI) feature selection, trained on all three modalities, achieved exceptional performance (accuracy = 0.95, AUC 0.94). When expression SNP separately, very good (AUC 0.65, 0.63, respectively). Using SHapley Additive exPlanations (SHAP), identified significant features, potentially serving as AD biomarkers. Notably, genetic-based biomarkers linked axon myelination synaptic vesicle membrane formation could aid early detection. In summary, this biomarker approach, integrating SHAP, shows promise for precise diagnosis, discovery, offers novel insights understanding treating disease. approach addresses challenges of accurate which crucial given complexities associated need non-invasive diagnostic methods.

Language: Английский

Citations

14
Integrating Convolutional Neural Networks with Attention Mechanisms for Magnetic Resonance Imaging-Based Classification of Brain Tumors DOI Creative Commons
Zahid Rasheed, Yongkui Ma, Inam Ullah

et al.

Bioengineering, Journal Year: 2024, Volume and Issue: 11(7), P. 701 - 701

Published: July 10, 2024

The application of magnetic resonance imaging (MRI) in the classification brain tumors is constrained by complex and time-consuming characteristics traditional diagnostics procedures, mainly because need for a thorough assessment across several regions. Nevertheless, advancements deep learning (DL) have facilitated development an automated system that improves identification medical images, effectively addressing these difficulties. Convolutional neural networks (CNNs) emerged as steadfast tools image visual perception. This study introduces innovative approach combines CNNs with hybrid attention mechanism to classify primary tumors, including glioma, meningioma, pituitary, no-tumor cases. proposed algorithm was rigorously tested benchmark data from well-documented sources literature. It evaluated alongside established pre-trained models such Xception, ResNet50V2, Densenet201, ResNet101V2, DenseNet169. performance metrics method were remarkable, demonstrating accuracy 98.33%, precision recall 98.30%, F1-score 98.20%. experimental finding highlights superior new identifying most frequent types tumors. Furthermore, shows excellent generalization capabilities, making it invaluable tool healthcare diagnosing conditions accurately efficiently.

Language: Английский

Citations

9
A deep learning model for Alzheimer’s disease diagnosis based on patient clinical records DOI Creative Commons
José Luis Ávila-Jiménez, Vanesa Cantón-Habas, María Pilar Carrera-González

et al.

Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 169, P. 107814 - 107814

Published: Dec. 9, 2023

Dementia, with Alzheimer's disease (AD) being the most common type of this neurodegenerative disease, is an under-diagnosed health problem in older people. The creation classification models based on AD risk factors using Deep Learning a promising tool to minimize impact under-diagnosis.

Language: Английский

Citations

19
Neural Computation-Based Methods for the Early Diagnosis and Prognosis of Alzheimer’s Disease Not Using Neuroimaging Biomarkers: A Systematic Review DOI Creative Commons
Ylermi Cabrera-León, Patricio García Báez,

Pablo Fernández-López

et al.

Journal of Alzheimer s Disease, Journal Year: 2024, Volume and Issue: 98(3), P. 793 - 823

Published: March 10, 2024

Background: The growing number of older adults in recent decades has led to more prevalent geriatric diseases, such as strokes and dementia. Therefore, Alzheimer’s disease (AD), the most common type dementia, become frequent too. Objective: goals this work are present state-of-the-art studies focused on automatic diagnosis prognosis AD its early stages, mainly mild cognitive impairment, predicting how research topic may change future. Methods: Articles found existing literature needed fulfill several selection criteria. Among others, their classification methods were based artificial neural networks (ANNs), including deep learning, data not from brain signals or neuroimaging techniques used. Considering our criteria, 42 articles published last decade finally selected. Results: medically significant results shown. Similar quantities shallow ANNs found. Recurrent transformers with speech longitudinal studies. Convolutional (CNNs) popular gait combined others modular approaches. Above one third cross-sectional utilized multimodal data. Non-public datasets frequently used studies, whereas opposite ones. databases indicated, which will be helpful for future researchers field. Conclusions: introduction CNNs superb did negatively affect usage other modalities. In fact, new ones emerged.

Language: Английский

Citations

7
Neuropsychological and electrophysiological measurements for diagnosis and prediction of dementia: a review on Machine Learning approach DOI
Claudia Carrarini,

Cristina Nardulli,

Laura Titti

et al.

Ageing Research Reviews, Journal Year: 2024, Volume and Issue: 100, P. 102417 - 102417

Published: July 14, 2024

Language: Английский

Citations

6
Enhancing Health and Public Health through Machine Learning: Decision Support for Smarter Choices DOI Creative Commons
Pedro Rodrigues, João Paulo do Vale Madeiro, João Alexandre Lôbo Marques

et al.

Bioengineering, Journal Year: 2023, Volume and Issue: 10(7), P. 792 - 792

Published: July 2, 2023

In recent years, the integration of Machine Learning (ML) techniques in field healthcare and public health has emerged as a powerful tool for improving decision-making processes [...].

Language: Английский

Citations

12
Systematic review of recent years: machine learning-based interactive therapy for people suffering from dementia DOI Creative Commons
Christian Röhrer,

Souhir Ben Souissi,

Mascha Kurpicz-Briki

et al.

Artificial Intelligence Review, Journal Year: 2025, Volume and Issue: 58(3)

Published: Jan. 6, 2025

Medical advances over the last century have significantly extended life expectancy. Today, world's population is quite old, and will become even older in years to come. Diseases that particularly concern elderly are therefore more frequent, dementia one of them. This condition mainly affects cannot be cured today. However, people suffering from do require care, this entails significant costs for our society. Machine learning could useful a context where it difficult find medical staff cost reduction priority. In recent years, research has been conducted ways treating with machine learning-based therapies which patient can actively participate. paper, systematic literature review these conducted: (a) paper metadata analysed, (b) dataset characteristics examined, (c) therapy types compared, (d) suggested architectures considered, (e) performance reviewed, (f) usability discussed, g) ethical considerations taken into account. Twenty-three papers were selected various use cell phones, computers, robots, or virtual reality. The results tests very positive, both terms cognitive faculties evolution satisfaction.

Language: Английский

Citations

0
Using Artificial Intelligence for Predictive Analysis of Dementia Awareness Among Community Adult Learners and Evaluation of Dementia-Friendliness in Community Environments DOI

Chia-Hui Hou,

Yihui Liu

Computers in Human Behavior, Journal Year: 2025, Volume and Issue: unknown, P. 108604 - 108604

Published: Feb. 1, 2025

Language: Английский

Citations

0
The role of magnetic resonance imaging in the diagnosis and prognosis of dementia DOI Open Access
Milica Živanović, Aleksandra Aracki-Trenkić, Vuk Milošević

et al.

Bosnian Journal of Basic Medical Sciences, Journal Year: 2022, Volume and Issue: unknown

Published: Nov. 30, 2022

Dementia is a syndrome characterized by multidomain acquired chronic cognitive impairment that has profound impact on daily life. Neurogenerative diseases such as Alzheimer's disease or nondegenerative vascular dementia are considered to cause dementia. The need for further diagnostic improvement originates from the prevalence of these conditions, especially in developed countries with predominance elderly population. Today, diagnosis and follow-up all neurodegenerative cannot be performed without radiological imaging, primarily magnetic resonance imaging (MRI). introduction 3T MRI its modern techniques, arterial spin labeling, enabled better visualization morphologic changes For patients dementia, various semiquantitative scales have been designed improve accuracy assessment decrease interobserver variability. Moreover, there growing novel therapies their side effects. To apply findings both already early stages, aim this paper review available literature summarize specific changes.

Language: Английский

Citations

17
Artificial Intelligence Diagnosing of Oral Lichen Planus: A Comparative Study DOI Creative Commons

Shandong Yu,

Wansu Sun,

DaWei Mi

et al.

Bioengineering, Journal Year: 2024, Volume and Issue: 11(11), P. 1159 - 1159

Published: Nov. 18, 2024

Early diagnosis of oral lichen planus (OLP) is challenging, which traditionally dependent on clinical experience and subjective interpretation. Artificial intelligence (AI) technology has been widely applied in objective rapid diagnoses. In this study, we aim to investigate the potential AI OLP evaluate its effectiveness improving diagnostic accuracy accelerating decision making. A total 128 confirmed patients were included, lesion images from various anatomical sites collected. The was performed using platforms, including ChatGPT-4O, ChatGPT (Diagram-Date extension), Claude Opus, for directly identification pre-training identification. After feature training, platforms significantly improved, with overall recognition rates Opus increasing 59%, 68%, 15% 77%, 80%, 50%, respectively. Additionally, buccal mucosa reached 94%, 93%, 56%, However, less effectively when recognizing lesions common complex cases; instance, gums only 60%, 20%, demonstrating significant limitations. study highlights strengths limitations different technologies provides a reference future applications medicine.

Language: Английский

Citations

3