m+Medelevator.RU
← Поиск публикаций
medRxiv : the preprint server for health sciences · 2026

Subtyping and staging of Alzheimer's disease from routine structural MRI with PHASE-AD.

СтатьяПрепринтНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…

История · было → стало

Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.

Baumeister Hannah, Lüsebrink Falk, Kleineidam Luca, Hansen Niels, Schmid Matthias, Moscoso Alexis, Leuzy Antoine, Mastenbroek Sophie E, Groot Colin, Brosseron Frederic, Ramirez Alfredo, Preis Lukas, Gref Daria, Spruth Eike J, Gemenetzi Maria, Altenstein Slawek, Fliessbach Klaus, Kimmich Okka, Schott Björn H, Rostamzadeh Ayda, Glanz Wenzel, Incesoy Enise I, Butryn Michaela, Janowitz Daniel, Rauchmann Boris-Stephan, Mladinov Mihovil, Grazia Alice, Sodenkamp Sebastian, Stöcker Tony, Hetzer Stefan, Dechent Peter, Stoecklein Sophia, Alzheimer’s Disease Repository Without Borders Investigators, Alzheimer’s Disease Neuroimaging Initiative, DELCODE study group, Oliver Peters, Hellmann-Regen Julian, Priller Josef, Schneider Anja, Wiltfang Jens, Buerger Katharina, Perneczky Robert, Teipel Stefan, Laske Christoph, Spottke Annika, Wagner Michael, Jessen Frank, Düzel Emrah, Berron David

Аннотация

Structural MRI is routinely acquired in the clinical assessment of Alzheimer's disease, yet quantitative morphometric indices derived from these scans remain largely confined to research settings. Here we present PHASE-AD - a framework that translates such indices into clinically interpretable classifications of atrophy subtype and stage that jointly capture atrophy progression while accounting for inter-individual atrophy heterogeneity. PHASE-AD is trained on MRI scans from 8,415 participants and robustly captures limbic-predominant and hippocampal-sparing atrophy subtypes that were identified across seven independent datasets. Two cross-validation schemes revealed high robustness across different field strength and scanner manufacturer configurations. Atrophy classifications were associated with diverging clinical profiles and tau accumulation patterns. In prospective designs mirroring contemporary AD trials, they stratified longitudinal cognitive trajectories and outperformed semi-quantitative visual MRI assessments as a clinically established comparator. These findings support the integration of automated atrophy subtyping and staging into clinical practice and pharmacological trials.