Predicting brain age gap with radiomics and automl: A Promising approach for age-Related brain degeneration biomarkers

Journal of Neuroradiology, Год журнала: 2023, Номер 51(3), С. 265 - 273

Опубликована: Сен. 16, 2023

The Brain Age Gap (BAG), which refers to the difference between chronological age and predicted neuroimaging age, is proposed as a potential biomarker for age-related brain degeneration. However, existing prediction models usually rely on single marker can not discover meaningful hidden information in radiographic images. This study focuses application of radiomics, an advanced imaging analysis technique, combined with automated machine learning predict BAG. Our methods achieve promising result mean absolute error 1.509 using Alzheimer's Disease Neuroimaging Initiative dataset. Furthermore, we find that hippocampus parahippocampal gyrus play significant role predicting interpretable method called SHapley Additive exPlanations. Additionally, our investigation discrepancies patients disease (AD) those mild cognitive impairment (MCI) reveals notable correlation clinical assessment scale scores. suggests BAG has serve support diagnosis AD MCI. Overall, this presents valuable insights into neurodegenerative diseases.

Язык: Английский

---

Toward MR protocol-agnostic, unbiased brain age predicted from clinical-grade MRIs

Scientific Reports, Год журнала: 2023, Номер 13(1)

Опубликована: Ноя. 10, 2023

The difference between the estimated brain age and chronological ('brain-PAD') could become a clinical biomarker. However, most models were developed for research-grade high-resolution T1-weighted MRIs, limiting their applicability to clinical-grade MRIs from various protocols. We adopted dual-transfer learning strategy develop model agnostic modality, resolution, or slice orientation. retrained convolutional neural network (CNN) using 6281 1559 patients, among 7 modalities 8 scanner models. CNN was trained estimate synthetic magnetization-prepared rapid gradient-echo (MPRAGEs) generated by 'super-resolution' method. failed with T2-weighted Gradient-Echo MRIs. mean absolute error (MAE) 5.86-8.59 years across other modalities, still higher than but comparable actual MPRAGEs some modalities. modeled "regression bias" in age, its correction is crucial providing unbiased summary statistics of personalized age-based biomarkers. bias generalizable as eliminated any correlation brain-PAD new samples. Brain-PAD reliable demonstrate feasibility predictions arbitrary thereby contributing medicine.

Язык: Английский

---

Advanced structural brain aging in preclinical autosomal dominant Alzheimer disease

Molecular Neurodegeneration, Год журнала: 2023, Номер 18(1)

Опубликована: Дек. 19, 2023

Abstract Background “Brain-predicted age” estimates biological age from complex, nonlinear features in neuroimaging scans. The brain gap (BAG) between predicted and chronological is elevated sporadic Alzheimer disease (AD), but underexplored autosomal dominant AD (ADAD), which progression highly predictable with minimal confounding age-related co-pathology. Methods We modeled BAG 257 deeply-phenotyped ADAD mutation-carriers 179 non-carriers the Dominantly Inherited Network using minimally-processed structural MRI then tested whether differed as a function of mutation cognitive status, or estimated years until symptom onset, it was associated established markers amyloid (PiB PET, CSF amyloid-β-42/40), phosphorylated tau (CSF plasma pTau-181), neurodegeneration neurofilament-light-chain [NfL]), cognition (global neuropsychological composite CDR-sum boxes). compared to other measures, examined heterogeneity variants, APOE ε4 carrier sex, education. Results Advanced aging observed approximately 7 before expected line indicators atrophy. moderately PET strongly pTau-181, NfL, mutation-carriers. Mutation education contributed variability BAG. Conclusions extend prior work ADAD, noting consistent results. associates well pTau, neurodegeneration, cognition, lesser extent, amyloid, ADAD. may capture similar signal measures. However, offers unique benefits simplicity data processing interpretation. Thus, results this cohort few confounds suggest that attributable neuropathology can be accurately quantified MRI.

Язык: Английский

---

Predicting brain age using partition modeling strategy and atlas-based attentional enhancement in the Chinese population

Cerebral Cortex, Год журнала: 2024, Номер 34(2)

Опубликована: Янв. 31, 2024

As a biomarker of human brain health during development, age is estimated based on subtle differences in structure from those under typical developmental. Magnetic resonance imaging (MRI) routine diagnostic method neuroimaging. Brain prediction MRI has been widely studied. However, few studies Chinese population have reported. This study aimed to construct predictive model for the across its lifespan. We developed partition transfer learning and atlas attention enhancement. The participants were separated into four groups, deep was trained each group identify regions most critical prediction. Atlas attention-enhancement also used help models focus only regions. proposed validated using 354 domestic datasets. For performance testing sets, mean absolute error 2.218 ± 1.801 years, Pearson correlation coefficient (r) 0.969, exceeding previous results wide-range In conclusion, could provide estimation assist assessing status health.

Язык: Английский

---

Slower speed of blood pressure recovery after standing is associated with accelerated brain ageing: Evidence from The Irish Longitudinal Study on Ageing (TILDA)

Cerebral Circulation - Cognition and Behavior, Год журнала: 2024, Номер 6, С. 100212 - 100212

Опубликована: Янв. 1, 2024

Impaired recovery of blood pressure (BP) in response to standing up is a prevalent condition older individuals. We evaluated the relationship between early hemodynamic responses and brain health adults over 50.

Язык: Английский

---