Ludwing Rivera

Background: Alzheimer’s disease (AD) remains the leading cause of dementia worldwide, with anti-amyloid monoclonal antibodies (MABs) marking a significant advance as the first disease-modifying therapies. Their use, however, is limited by amyloid-related imaging abnormalities (ARIA), which appear as vasogenic edema or effusion (ARIA-E) and hemosiderin-related changes (ARIA-H) on MRI. Variability in imaging protocols, subtle early findings, and the lack of standardized risk models challenge detection and management. Methods: This narrative review summarizes current artificial intelligence (AI) applications for ARIA detection and risk prediction. A comprehensive literature search across PubMed, Embase, and Scopus identified studies focusing on MRI-based AI analysis, lesion quantification, and predictive modeling. Results: The evidence is organized into six thematic domains: ARIA definitions, imaging challenges, foundations of AI in neuroimaging, detection tools, predictive frameworks, and future perspectives. Conclusions: AI offers promising avenues to standardize ARIA evaluation, improve lesion quantification, and enable individualized risk prediction. Progress will depend on multicenter datasets, shared frameworks, and prospective validation. Ultimately, AI-driven neuroimaging may transform how treatment-related complications are monitored in the era of anti-amyloid therapy.

Background/Objectives: Hippocampal atrophy indicates Alzheimer’s disease (AD) progression and guides follow-up and trial enrichment. Identifying high-risk patients is crucial for optimizing care, but accessible, interpretable machine-learning models (ML) are limited. We developed an explainable ML model using clinical data and CSF erythrocyte load (CTRED) to classify adults with AD as high- or low-risk based on hippocampal volume decline. Methods: Included ADNI participants with ≥2 MRIs, baseline lumbar puncture, and vital signs within 6 months of MRI (n = 26). The outcome was the Annual Percentage Change (APC) in hippocampal volume, classified as low or high risk. Predictors were standardized; models included SVM, logistic regression, and Ridge Classifier, tuned and tested on a set (n = 6). Thresholds were based on out-of-fold predictions under a 10–90% positive rate. Explainability used PFI and SHAP for per-patient contributions. Results: All models gave identical classifications, but discrimination varied: Ridge AUC = 1.00, logistic = 0.889, and SVM = 0.667. PFI highlighted MAPres and sex as main signals; CTRED contributed, and age had a minor impact. Conclusions: The explainable ML model with clinical data and CTRED can stratify AD patients by hippocampal atrophy risk, aiding follow-up and vascular assessment planning rather than treatment decisions. Validation in larger cohorts is needed. This is the first ML study to use CSF erythrocyte load to predict hippocampal atrophy risk in AD.

Background/Objectives: Alzheimer’s disease (AD) exhibits clinical and biological variability. This study aimed to identify MRI-defined subtypes reflecting distinct biological pathways of neurodegeneration and cognitive decline. Methods: We applied principal component analysis followed by k-means clustering (k = 3) on volumetric MRI data from 924 participants and validated clusters using cerebrospinal fluid (CSF) biomarkers (Aβ₄₂, total tau, p-tau, CTRED, MAPres, glucose, CTWHITE). Results: Three major phenotypes emerged: (1) a tau/vascular limbic subtype with pronounced hippocampal and amygdala atrophy and elevated tau and CTRED levels; (2) a volume-preserved, low-amyloid subtype consistent with early-stage or cognitively resilient AD; and (3) a diffuse-atrophy subtype with high amyloid and tau burden and ventriculomegaly. Comparative analysis revealed progressive biological shifts from amyloid accumulation to tau aggregation and vascular compromise across these clusters. Conclusions: MRI-based clustering validated by CSF biomarkers delineates biologically meaningful AD endophenotypes. The results suggest a gradual cognitive decline driven by amyloid–tau–vascular interactions, supporting multimodal phenotyping as a practical approach for precision staging and intervention.

Background/Objectives: Cerebrospinal fluid erythrocyte load (CTRED) reflects occult red-blood-cell ingress into brain/CSF and consequent heme–iron exposure, a toxic pathway relevant to Alzheimer’s disease (AD). We aimed to develop explainable machine learning (ML) models that classify high vs. low CTRED from routine, largely non-invasive inputs, and to position a blood-first workflow leveraging contemporary plasma amyloid–tau biomarkers. Methods: Twenty-six ADNI participants were analyzed. Inputs were age, sex, mean arterial pressure (MAPres), amyloid (Aβ42), total tau, phosphorylated tau, and hippocampal atrophy rate (APC) derived from longitudinal MRI. APC was computed from normalized hippocampal volumes. CTRED was binarized at the median (0 vs. >0). Data were split into train (n = 20) and held-out test (n = 6). Five classifiers (linear SVM, ridge, logistic regression, random forests, and MLP) were trained in leakage-safe pipelines with stratified five-fold cross-validation. To provide a comprehensive assessment, we presented the contribution AUC, thresholded performance metrics, summarized model performance, and the permutation feature importance (PFI). Results: On the test set, SVM, ridge, logistic regression, and random forests achieved AUC = 1.00, while the MLP achieved AUC = 0.833. Across models, PFI consistently prioritized p-tau/tau, Aβ42, and MAPres; age, sex, and APC contributed secondarily. The attribution profile aligns with mechanisms linking BBB dysfunction and amyloid-related microvascular fragility with tissue vulnerability to heme–iron. Conclusions: In this proof-of-concept study, explainable ML predicted CTRED from routine variables with biologically coherent drivers. Although ADNI measurements were CSF-based and the sample was small, the framework is non-invasive by adding plasma p-tau217/Aβ1–42 for amyloid, tau inputs, and integrating demographics, hemodynamic context, and MRI. External, plasma-based validation in larger cohorts is warranted, alongside extension to MCI and multimodal correlation (QSM, DCE-MRI) to establish clinically actionable CTRED thresholds.

Background: This narrative review summarizes recent progress in artificial intelligence (AI), especially radiomics and deep learning, for non-invasive diagnosis and molecular profiling of gliomas. Methodology: A thorough literature search was conducted on PubMed, Scopus, and Embase for studies published from January 2020 to July 2025, focusing on clinical and technical research. In key areas, these studies examine AI models’ predictive capabilities with multi-parametric Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET). Results: The domains identified in the literature include the advancement of radiomic models for tumor grading and biomarker prediction, such as Isocitrate Dehydrogenase (IDH) mutation, O6-methylguanine-dna methyltransferase (MGMT) promoter methylation, and 1p/19q codeletion. The growing use of convolutional neural networks (CNNs) and generative adversarial networks (GANs) in tumor segmentation, classification, and prognosis was also a significant topic discussed in the literature. Deep learning (DL) methods are evaluated against traditional radiomics regarding feature extraction, scalability, and robustness to imaging protocol differences across institutions. Conclusions: This review analyzes emerging efforts to combine clinical, imaging, and histology data within hybrid or transformer-based AI systems to enhance diagnostic accuracy. Significant findings include the application of DL to predict cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) deletion and chemokine CCL2 expression. These highlight the expanding capabilities of imaging-based genomic inference and the importance of clinical data in multimodal fusion. Challenges such as data harmonization, model interpretability, and external validation still need to be addressed.

Background: Alzheimer’s disease (AD) involves ongoing neurodegeneration, with phosphorylated tau (PTAU) intracellular accumulation closely associated with cortical shrinking. However, not everyone with high PTAU levels shows the same degree of neurodegeneration, implying that other biological stress factors might influence tau’s harmful effects. This research explores whether cerebrospinal fluid erythrocyte burden (CTRED), a marker indicating vascular–CSF barrier disruption and heme toxicity, affects the link between PTAU181 levels and entorhinal cortex atrophy in AD. Methods: We examined 25 observations from 18 patients with AD using a linear mixed effects model. The dependent variable was entorhinal cortex volume, with fixed effects for PTAU, CTRED, and their interaction. Random intercepts accounted for variability within subjects. A cognitively normal (CN) control group was included for comparison. Results: CTRED is significantly associated with reduced entorhinal volume (p = 0.005). A notable interaction between CTRED and PTAU was also found (p = 0.004), suggesting that higher CTRED enhances PTAU’s atrophic effects. PTAU alone was not a significant predictor. No significant effects were observed in the CN group, which supports the specificity of the disease. Conclusions: CTRED alters the neurotoxic impact of PTAU on the entorhinal cortex in AD, supporting a multi-hit model of degeneration that involves tau pathology and erythrocyte-derived stress. These findings emphasize the clinical importance of vascular–CSF biomarkers in predicting neurodegeneration and guiding targeted treatments.