{"doi":"10.1002/nbm.70018","title":"Diffusion MRI in 2‐Month‐Old Mouse Brain Predicts Alzheimer's Pathology Genotype","abstract":"ABSTRACT Diffusion MRI (dMRI) is widely used as a non‐invasive means of detecting changes in brain tissue microstructure. In our previous studies, we demonstrated the sensitivity of dMRI to capture brain microstructural alterations in the triple transgenic (3xTg‐ AD ) mice, particularly brain morphological abnormalities in 2‐month‐old mice, where dMRI was sensitive to myelin abnormalities, to microglia proliferation/activation, and to the larger number of basal forebrain cholinergic neurons previously described in this model at this young age. In this study, we extend our prior work by establishing the dMRI profile of several brain regions relevant to AD pathology in 2‐month‐old 3xTg‐ AD and age‐matched controls (NC) and by investigating the effectiveness of these dMRI metrics in predicting group genotype using elastic net (EN) logistic regression modeling. EN has been shown to be a high‐performance and stable machine learning model for neuroimaging data. Our results demonstrated significant group differences in several ROIs, particularly in the corpus callosum (CC) where fractional anisotropy (FA) ( p < 0.0001; d = −1.87), radial diffusivity (D ┴ ) ( p < 0.0001; d = −1.33), and radial kurtosis (K ┴ ) ( p < 0.0001; d = −1.34) were statistically significant and the most sensitive dMRI metrics to differentiate between the two groups, with large effect sizes (Cohen's d ) values. Moreover, FA in the ventral hippocampus (VH) ( p < 0.0001; d = 1.13) and fimbria (Fi) ( p < 0.0001; d = −1.04) as well as mean diffusivity (MD) ( p < 0.0001; d = 1.10) and D ┴ in the subiculum (Sub) ( p < 0.0001; d = 1.12) were also statistically significant and able to clearly distinguish the two groups. Additionally, our results from the trained EN model indicate that FA in the VH, CC, and cingulate cortex (Ctx‐Cg) were the three best dMRI metrics to classify the 3xTg‐ AD mice with an accuracy of 0.95. Sensitivity and specificity were also calculated to assess the goodness of prediction, resulting in 0.96 and 0.94, respectively.","journal":"NMR in Biomedicine","year":2025,"id":532054,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":3,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9591,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2025-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1410811,"name":"Bryan Granger","orcid":"0009-0008-6663-3755","position":1,"is_corresponding":false},{"id":702963,"name":"Joshua Voltin","orcid":null,"position":2,"is_corresponding":false},{"id":406400,"name":"Paul J. Nietert","orcid":"0000-0002-3933-4986","position":3,"is_corresponding":false},{"id":311328,"name":"Stefano Berto","orcid":"0000-0001-9028-9458","position":4,"is_corresponding":false},{"id":453573,"name":"Jens H. Jensen","orcid":"0000-0003-3219-4287","position":5,"is_corresponding":false},{"id":702286,"name":"Maria F. Falangola","orcid":"0000-0002-6349-7734","position":0,"is_corresponding":true}],"reference_count":76,"raw_metadata":null,"created_at":"2026-07-19T02:51:18.928280Z","pmid":"40037790","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}