On Extracting Digitized Spiral Dynamics' Representations: A Study on Transfer Learning for Early Alzheimer's Detection.

BackgroundAs the world’s elderly population increases, health monitoring technologies that can automatically detect subtle changes resulting from Alzheimer’s disease (AD) have become increasingly important. In this respect, speech is one of the promising clues due to the expansion of voice‐based interaction systems such as smartphones and tablets. Indeed, previous studies have succeeded in quantifying language dysfunctions and identifying AD and mild cognitive impairment (MCI) from speech data collected during neuropsychological tests conducted by clinicians. Conducting such assessments in an automated fashion by using computer devices would extend opportunities for assessments and help with the early detection of AD. In particular, if we can detect language dysfunctions related to AD from various types of speech data (e.g., question answering and daily conversations), it would extend the scope of application and help improve the current worldwide low‐diagnosis coverage.MethodIn this study, we developed a tablet‐based application and collected spontaneous speech data during interview tasks from 106 Japanese seniors consisting of 48 healthy controls (HC), 33 MCI, and 25 AD.Participants answered nine questions relating to their current condition, yesterday’s dinner, games played as a child, and future travel plans. For comparison, we also collected speech data during neuropsychological tests (e.g., verbal fluency and picture description tasks) using the tablet. We extracted vocal and prosodic features from both speech data and then built binary classification models for differentiating MCI or AD from HC using a support vector machine with a feature selection method. We evaluated the models by leave‐one‐subject‐out cross‐validation.ResultWe found that the models using speech data during neuropsychological tests achieved the accuracy of 86.3% for HC vs. AD and 80.2% for HC vs. MCI. The models using spontaneous speech data from interview tasks achieved comparable accuracies: 87.7% for HC vs. AD and 81.6% for HC vs. MCI.ConclusionWe demonstrated the possibility of using tablet‐based automatic assessments for detecting patients with both MCI and AD. In addition, our results suggest that speech data from not only neuropsychological tasks but also when answering everyday questions might contain useful information to help early detection of AD.

Alzheimer’s disease (AD) is a remarkable challenge for healthcare in the 21st century. Since 2017, deep learning models with transfer learning approaches have been gaining recognition in AD detection, and progression prediction by using neuroimaging biomarkers. This paper presents a systematic review of the current state of early AD detection by using deep learning models with transfer learning and neuroimaging biomarkers. Five databases were used and the results before screening report 215 studies published between 2010 and 2020. After screening, 13 studies met the inclusion criteria. We noted that the maximum accuracy achieved to date for AD classification is 98.20% by using the combination of 3D convolutional networks and local transfer learning, and that for the prognostic prediction of AD is 87.78% by using pre-trained 3D convolutional network-based architectures. The results show that transfer learning helps researchers in developing a more accurate system for the early diagnosis of AD. However, there is a need to consider some points in future research, such as improving the accuracy of the prognostic prediction of AD, exploring additional biomarkers such as tau-PET and amyloid-PET to understand highly discriminative feature representation to separate similar brain patterns, managing the size of the datasets due to the limited availability.

Alzheimer’s disease (AD), the most prevalent neurodegenerative disorder globally, has emerged as a significant health concern, particularly due to the increasing aging population. Recently, it has been revealed that extracellular vesicles (EVs) originating from neurons play a critical role in AD pathogenesis and progression. These neuronal EVs can cross the blood-brain barrier and enter peripheral circulation, offering a less invasive means for assessing blood-based AD biomarkers. In this study, we analyzed plasma EV-derived messenger RNA (mRNA) from 82 subjects, including individuals with AD, mild cognitive impairment (MCI), and healthy controls, using next-generation sequencing (NGS) to profile their gene expression for functional enrichment and pathway analysis. Based on the differentially expressed genes identified in both MCI and AD groups, we established a diagnostic model by implementing a machine learning classifier. The refined model demonstrated an average diagnostic accuracy over 98% and showed a strong correlation with different AD stages, suggesting the potential of plasma EV-derived mRNA as a promising non-invasive biomarker for early detection and ongoing monitoring of AD.

AI-based tool for early detection of Alzheimer's disease

Neuroscience is of emerging importance along with the contributions of Operational Research to the practices of diagnosing neurodegenerative diseases with computer-aided systems based on brain image analysis. Although multiple biomarkers derived from Magnetic Resonance Imaging (MRI) data have proven to be effective in diagnosing Alzheimer’s disease (AD) and mild cognitive impairment (MCI), no specific system has yet been a part of routine clinical practice. This paper aims to introduce a fully-automated voxel-based procedure, Voxel-MARS, for detection of AD and MCI in early stages of progression. Performance was evaluated on a dataset of 508 MRI volumes gathered from the Alzheimer’s Disease Neuroimaging Initiative database. Data were transformed into a high-dimensional space through a feature extraction process. A novel 3-step feature selection procedure was applied. Multivariate Adaptive Regression Splines method was used as a classifier for the first time in the field of brain MRI analysis. The results were compared to those presented in a previous study on 28 voxel-based methods in terms of their ability to separate control normal (CN) subjects from the ones diagnosed with AD and MCI. It was observed that our method outperformed all of the others in sensitivity (83.58% in AD/CN and 78.38% in MCI/CN classification) with acceptable specificity values (over 85% in both cases). Furthermore, the method worked for discriminating MCI patients which converted to AD in 18 months (MCIc) from non-converters (MCInc) with a sensitivity outcome better than 27 of 28 methods. Overall, it was shown that the proposed method is promising in early detection of AD.

The hippocampus is one of the earliest brain regions affected in Alzheimer's disease (AD) and tests of hippocampal function have the potential to detect AD in its earliest stages. Given that the hippocampus is critically involved in allocentric spatial memory, this study applied a short test of spatial memory, the 4 Mountains Test (4MT), to determine whether test performance can differentiate mild cognitive impairment (MCI) patients with and without CSF biomarker evidence of underlying AD and whether the test can distinguish patients with MCI and mild AD dementia when applied in different cultural settings. Healthy controls (HC), patients with MCI, and mild AD dementia were recruited from study sites in UK and Italy. Study numbers were: HC (UK 20, Italy 10), MCI (UK 21, Italy 14), and AD (UK 11, Italy 9). Nineteen UK MCI patients were grouped into CSF biomarker-positive (MCI+, n = 10) and biomarker-negative (MCI-, n = 9) subgroups. Behavioral data were correlated with hippocampal volume and cortical thickness of the precuneus and posterior cingulate gyrus. Spatial memory was impaired in both UK and Italy MCI and AD patients. Test performance additionally differentiated between MCI+ and MCI- subgroups (P = 0.001). A 4MT score of ≤8/15 was associated with 100% sensitivity and 90% specificity for detection of early AD (MCI+ and mild AD dementia) in the UK population, and with 100% sensitivity and 50% specificity for detection of MCI and AD in the Italy sample. 4MT performance correlated with hippocampal volume in the UK population and cortical thickness of the precuneus in both study populations. In conclusion, performance on a hippocampus-sensitive test of spatial memory differentiates MCI due to AD with high diagnostic sensitivity and specificity. The observation that similar diagnostic sensitivity was obtained in two separate study populations, allied to the scalability and usability of the test in community memory clinics, supports future application of the 4MT in the diagnosis of pre-dementia due to AD.

BackgroundThe recent promise of disease-modifying therapies for Alzheimer’s disease (AD) has reinforced the need for accurate biomarkers for early disease detection, diagnosis and treatment monitoring. Advances in the development of novel blood-based biomarkers for AD have revealed that plasma levels of tau phosphorylated at various residues are specific and sensitive to AD dementia. However, the currently available tests have shortcomings in access, throughput, and scalability that limit widespread implementation.MethodsWe evaluated the diagnostic and prognostic performance of a high-throughput and fully-automated Lumipulse plasma p-tau181 assay for the detection of AD. Plasma from older clinically unimpaired individuals (CU, n = 463) and patients with mild cognitive impairment (MCI, n = 107) or AD dementia (n = 78) were obtained from the longitudinal Stanford University Alzheimer’s Disease Research Center (ADRC) and the Stanford Aging and Memory Study (SAMS) cohorts. We evaluated the discriminative accuracy of plasma p-tau181 for clinical AD diagnosis, association with amyloid β peptides and p-tau181 concentrations in CSF, association with amyloid positron emission tomography (PET), and ability to predict longitudinal cognitive and functional change.ResultsThe assay showed robust performance in differentiating AD from control participants (AUC 0.959, CI: 0.912 to 0.990), and was strongly associated with CSF p-tau181, CSF Aβ42/Aβ40 ratio, and amyloid-PET global SUVRs. Associations between plasma p-tau181 with CSF biomarkers were significant when examined separately in Aβ+ and Aβ− groups. Plasma p-tau181 significantly increased over time in CU and AD diagnostic groups. After controlling for clinical diagnosis, age, sex, and education, baseline plasma p-tau181 predicted change in MoCA overall and change in CDR Sum of Boxes in the AD group over follow-up of up to 5 years.ConclusionsThis fully-automated and available blood-based biomarker assay therefore may be useful for early detection, diagnosis, prognosis, and treatment monitoring of AD.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that poses a significant global health challenge. Early and accurate detection is crucial for timely intervention and for the development of new therapies. Machine learning (ML) has emerged as a powerful tool for analyzing the complex, high-dimensional data associated with AD. This systematic review was conducted by searching major academic databases for peer-reviewed literature published between 2020 and 2025. We identified studies that applied ML models to neuroimaging data for the early detection of AD. Information on ML models, datasets, and performance metrics was extracted and synthesized to provide a comprehensive overview of the field. A range of ML models are employed, from traditional supervised learning algorithms like Support Vector Machines (SVM) to more advanced ensemble (e.g., Random Forest) and deep learning methods (e.g., CNNs). Studies consistently show that ensemble and deep learning models achieve high performance (>90% accuracy in many cases), particularly in multiclass classification. However, the field faces persistent challenges, including severe class imbalance in common datasets, issues of data quality and anomalies, and the "black box" nature of complex models, which limits their interpretability and clinical trust. ML models show immense promise for the early and accurate detection of AD. However, for these tools to be successfully translated into clinical practice, future research must focus on developing robust, generalizable, and interpretable models that can effectively address the challenges of data imbalance and pathological heterogeneity.

Imaging of amyloid β in Alzheimer's disease with 18F-BAY94-9172, a novel PET tracer: proof of mechanism

Early detection of Alzheimer’s disease (AD) has become one of the major research topics nowadays. The utilization of the computerized system may help medical experts to better understand and analyse the magnetic resonance imaging (MRI) images of AD patients for early detection. One of the commonly steps taken for the analysis of the image is image denoising using certain filters. However, finding shows that previous researchers use different approaches. This study aims to analyse the effect of different denoising techniques towards detection of Alzheimer’s disease. Data of two different groups (AD patients and Normal Control) were collected from Alzheimer's Disease Neuroimaging Initiative (ADNI). Then brain extraction and skull stripping were performed. Several image denoising techniques were implemented for both groups namely median filter, Wiener filter, histogram equalization filter and Gaussian low pass filter. After that, all images underwent texture feature extraction process and analysis were made to see the effect of those denoising techniques towards the features of Gray-Level Co-occurrence Matrix (GLCM) extracted which are the contract, correlation, energy and homogeneity features. The result shows that the use of mentioned denoising filters do not give effect to the extracted features. All values of contrast, correlation, energy and homogeneity cannot clearly distinguish between AD and NC groups. Without any filters on the other hand, contrast feature gives the best output in distinguishing between AD and NC groups with the normalized value of 0.1. The result from this study may help in thorough investigation of other features or hybrid features that could be used for the purpose of detection and classification of AD.

Alzheimer disease (AD) is a neurological disorder which shrinks the brain and causes dementia. In the past, this disease was more prevalent in American countries. However, it is now common in other countries as well. When compared to youth, older people are more affected by this disease. The number of people affected by this disease is gradually increasing each year, and according to one study, this number may reach around 15 million in the near future. People who are affected will experience symptoms such as memory loss and confusion. Early detection of Alzheimer disease is essential for providing appropriate treatments. Neuroimaging based Machine Learning methods are commonly utilized for the detection and diagnosis of Alzheimer's, but they are time-consuming. The time consumption can be reduced, and the detection accuracy can be increased further with the help of Deep Learning and Transfer Learning algorithms. This proposed work compares 4 different Transfer Learning Models. VGG-16 has the highest accuracy of 97.2 percent out of the four models tested.

Recently, rich computational methods that use deep learning or machine learning have been developed using linguistic biomarkers for the diagnosis of early-stage Alzheimer disease (AD). Moreover, some qualitative and quantitative studies have indicated that certain part-of-speech (PoS) features or tags could be good indicators of AD. However, there has not been a systematic attempt to discover the underlying relationships between PoS features and AD. Moreover, there has not been any attempt to quantify the relative importance of PoS features in detecting AD. Our goal was to disclose the underlying relationship between PoS features and AD, understand whether PoS features are useful in AD diagnosis, and explore which PoS features play a vital role in the diagnosis. The DementiaBank, containing 1049 transcripts from 208 patients with AD and 243 transcripts from 104 older control individuals, was used. A total of 27 PoS features were extracted from each record. Then, the relationship between AD and each of the PoS features was explored. A transformer-based deep learning model for AD prediction using PoS features was trained. Then, a global explainable artificial intelligence method was proposed and used to discover which PoS features were the most important in AD diagnosis using the transformer-based predictor. A global (model-level) feature importance measure was derived as a summary from the local (example-level) feature importance metric, which was obtained using the proposed causally aware counterfactual explanation method. The unique feature of this method is that it considers causal relations among PoS features and can, hence, preclude counterfactuals that are improbable and result in more reliable explanations. The deep learning-based AD predictor achieved an accuracy of 92.2% and an F1-score of 0.955 when distinguishing patients with AD from healthy controls. The proposed explanation method identified 12 PoS features as being important for distinguishing patients with AD from healthy controls. Of these 12 features, 3 (25%) have been identified by other researchers in previous works in psychology and natural language processing. The remaining 75% (9/12) of PoS features have not been previously identified. We believe that this is an interesting finding that can be used in creating tests that might aid in the diagnosis of AD. Note that although our method is focused on PoS features, it should be possible to extend it to more types of features, perhaps even those derived from other biomarkers, such as syntactic features. The high classification accuracy of the proposed deep learner indicates that PoS features are strong clues in AD diagnosis. There are 12 PoS features that are strongly tied to AD, and because language is a noninvasive and potentially cheap method for detecting AD, this work shows some promising directions in this field.

Alzheimer’s disease (AD) is a progressive neurodegenerative condition that affects the elderly population. The early detection and diagnosis of AD is critical for achieving effective treatment, as it can greatly improve the patient experience. AD can be viewed through imaging techniques like MRI, PET, and SPECT, providing valuable information about structural and functional changes. These findings are important in understanding this area. However, each imaging modality offers a different perspective. This information can be better collected from several of the other modalities as well as from some others to improve accuracy and reliability in AD detection. By combining information from different imaging modalities, such as MRI, PET, DTI, and fMRI, automated multimodal medical image frameworks aim to create a fused representation that preserves the relevant features from each modality. Convolutional neural networks (CNNs) and generative adversarial networks (GANs), among other deep learning techniques, have been prevalent in these frameworks for learning discriminative and informative features from multi-modal data. In this paper, The Alzheimer’s Disease Neuroimaging Initiative (ADNI) is used for experimental analysis. The proposed work gives 98.94% of accuracy and 1.06% of error which is greater than the existing approaches.

BackgroundWhile several non‐invasive plasma biomarkers are rapidly developing, they still have lower accuracy than the established CSF biomarkers. This study therefore comprehensively examined 6,905 plasma proteins in over 3,000 individuals to identify novel biomarkers for predicting clinical Alzheimer’s disease (AD) and monitoring disease progression.MethodTo identify and validate plasma biomarkers, we performed difference abundance analysis of 6,905 plasma proteins in a total of 3,366 cases and controls from three different datasets (Knight‐ADRC discovery, Knight‐ADRC replication and Stanford ADRC). Enrichment analysis with Gene Ontology and DisGeNet were performed to understand biological significance of the identified proteins. Machine learning model was used to create prediction models for AD risk, which were replicated on the Knight ADRC replication and Stanford datasets.ResultWe identified and replicated 257 plasma proteins (including NEFL, SMOC1, SPON1, and NEUROG1) associated with clinical AD status. Machine learning model identified 89 proteins with strong prediction power for AD (AUC=0.843 in Knight ADRC; AUC=0.771 in Stanford ADRC), but not for Parkinson's disease (AUC< 0.6 in two independent datasets), frontotemporal dementia (AUC=0.726), and dementia with Lewy bodies (AUC=0.712), indicating AD specificity. This proteomic signature also predicted AD individuals with faster progression (clinical dementia rating sum‐of‐boxes changes per year; P=4.7∗10‐5). Furthermore, this model identified a subset of cognitive normal individuals at 201 times higher risk of developing AD individuals (hazard ratio =201.3; P=1.95∗10‐2). Pathway analysis highlighted proteins (APOE, SPP1, and PLTP) related to lipid transport pathways (FDR=2.7∗10‐2). DisGeNet indicated the relevance to senile cardiac amyloidosis (APOE, CRP, and NEFL; FDR=1.1∗10‐4), highlighting transthyretin, that was found to be potentially protective against amyloid beta deposition.ConclusionThis large‐scale plasma proteomic study identified proteins associated with AD, developed a robust prediction model that accurately predicted AD status, examined differences between predicted groups in progression and conversion, and revealed pathways and diseases relevant to AD. These findings from extensive analysis provides the potential of plasma proteins as biomarkers for routine clinical uses in early detection of AD and guiding AD treatment decisions. They provide valuable insights into the relevance of specific pathways and diseases in the context of AD.

Alzheimer's disease (AD) is a disease of the brain that slowly destroys memory and thinking ability. It is one of the most common forms of dementia in people over the age of 65 worldwide. Early detection of Alzheimer's disease helps medical personnel diagnose this disease, which certainly reduces the risk of death. Therefore, early detection of Alzheimer's disease is an important healthcare issue. In order to achieve accurate and timely diagnosis, and for detection of AD in its early stages, many Artificial Intelligence (AI) and Machine learning approaches have been proposed using data from brain imaging. AI and ML are used for many medical and health purposes, such as disease detection. In this chapter, the main focus is on the diagnosis and detection of Alzheimer's disease using CNN model. A Convolutional Neural Network (CNN) is a deep learning model specialized in detecting and classifying patterns in visual data, commonly used for tasks like image recognition, object detection, and classification.