Decoding Pathological Neuro-Trajectories - Artificial Intelligence Frameworks for Neurodegenerative Disorders

Computational neuroscience has rapidly evolved through the integration of artificial intelligence (AI) technologies, enabling new insights into neurodegenerative disorders, including Alzheimer’s disease and related dementias (ADRD), demyelinating diseases, Parkinsonian disorders, motor neuron diseases, Prion diseases, Huntington’s disease as well as other neurological conditions with neurodegenerative components, not only in chronic or severe phase of the disease but also in their prodromal stage. Despite these advances, a comprehensive understanding of the complex neural mechanisms underlying these disorders and their clinical progression remains elusive. Recent research highlights the value of AI approaches, including machine learning, deep learning, and generative models, in extracting meaningful patterns from multimodal neuroimaging and genomic data. However, the translation of these computational advances into clinically actionable insights remains in its infancy, with persistent challenges related to reproducibility, model generalizability, and interpretability. This Research Topic aims to accelerate insight into neurodegenerative trajectories by fostering AI-driven frameworks that identify novel neuroimaging and digital biomarkers, enhance early detection, and predict individualized disease progression. We invite original research and comprehensive reviews that prioritize bridging the gap between "black-box" prediction and clinical actionability, specifically addressing the complexities of the interaction between the physiological aging and pathological neurodegeneration, also considering potential environmental (i.e cognitive reserve) as well as genetic (i.e. APOE) modulators. Submissions should investigate how AI can refine patient stratification and monitor disorder evolution while overcoming persistent limitations in model generalizability, transparency, and interpretability.We welcome articles addressing, but not limited to, the following themes:- AI-Driven Biomarker Discovery: Identifying robust, high-dimensional neuroimaging signatures and digital biomarkers predictive of pathological aging and neurodegeneration.- Precision Prognostics: Developing predictive models for early-stage detection by identifying transition states (e.g., healthy aging to MCI, or MCI to AD, etc.) and forecasting individualized trajectories, with an emphasis on continuous-time modeling and clinical validation of transition markers.- Neuro-Subtyping & Stratification: Identifying latent biotypes and clinical phenotypes through unsupervised and semi-supervised learning and linking these strata to distinct longitudinal outcomes.- Multimodal Fusion & Integration: Advanced frameworks for integrating diverse data streams, including structural MRI, fMRI, PET, dMRI, EEG, MEG, SPECT and digital phenotyping to create a holistic view of the aging brain.- Imaging Genomics: Investigating the intersection of radiomics and genomics to reveal how genetic predispositions manifest in brain structure and function, emphasizing biologically constrained AI models.- Generative Neuro-Architectures: Utilizing synthetic data and generative models (e.g., GANs, Diffusion Models) to simulate disease pathways, treatment effects, or hypothetical aging trajectories, or create counterfactual “healthy” brain states for comparative clinical analysis. Topic editors: Anees Abrol, Enrico Premi Source: https://www.frontiersin.org/research-topics/78185/decoding-pathological-neuro-trajectories---artificial-intelligence-frameworks-for-neurodegenerative-disorders