Neuroimage-Clinical最新文献

Background

In Huntington’s disease clinical trials, recruitment and stratification approaches primarily rely on genetic load, cognitive and motor assessment scores. They focus less on in vivo brain imaging markers, which reflect neuropathology well before clinical diagnosis. Machine learning methods offer a degree of sophistication which could significantly improve prognosis and stratification by leveraging multimodal biomarkers from large datasets. Such models specifically tailored to HD gene expansion carriers could further enhance the efficacy of the stratification process.

Objectives

To improve stratification of Huntington’s disease individuals for clinical trials.

Methods

We used data from 451 gene positive individuals with Huntington’s disease (both premanifest and diagnosed) from previously published cohorts (PREDICT, TRACK, TrackON, and IMAGE). We applied whole-brain parcellation to longitudinal brain scans and measured the rate of lateral ventricular enlargement, over 3 years, which was used as the target variable for our prognostic random forest regression models. The models were trained on various combinations of features at baseline, including genetic load, cognitive and motor assessment score biomarkers, as well as brain imaging-derived features. Furthermore, a simplified stratification model was developed to classify individuals into two homogenous groups (low risk and high risk) based on their anticipated rate of ventricular enlargement.

Results

The predictive accuracy of the prognostic models substantially improved by integrating brain imaging features alongside genetic load, cognitive and motor biomarkers: a 24 % reduction in the cross-validated mean absolute error, yielding an error of 530 mm³/year. The stratification model had a cross-validated accuracy of 81 % in differentiating between moderate and fast progressors (precision = 83 %, recall = 80 %).

Conclusions

This study validated the effectiveness of machine learning in differentiating between low- and high-risk individuals based on the rate of ventricular enlargement. The models were exclusively trained using features from HD individuals, which offers a more disease-specific, simplified, and accurate approach for prognostic enrichment compared to relying on features extracted from healthy control groups, as done in previous studies. The proposed method has the potential to enhance clinical utility by: i) enabling more targeted recruitment of individuals for clinical trials, ii) improving post-hoc evaluation of individuals, and iii) ultimately leading to better outcomes for individuals through personalized treatment selection.

Background

Functional neurological disorder (FND) is a common neuropsychiatric condition with established diagnostic criteria and effective treatments but for which the underlying neuropathophysiological mechanisms remain incompletely understood. Recent neuroimaging studies have revealed FND as a multi-network brain disorder, unveiling alterations across limbic, self-agency, attentional/salience, and sensorimotor networks. However, the relationship between identified brain alterations and disease progression or improvement is less explored.

Methods

This study included resting-state functional magnetic resonance imaging (fMRI) data from 79 patients with FND and 74 age and sex-matched healthy controls (HC). First, voxel-wise BOLD signal variability was computed for each participant and the group-wise difference was calculated. Second, we investigated the potential of BOLD signal variability to serve as a prognostic biomarker for clinical outcome in 47 patients who attended a follow-up measurement after eight months.

Results

The results demonstrated higher BOLD signal variability in key networks, including the somatomotor, salience, limbic, and dorsal attention networks, in patients compared to controls. Longitudinal analysis revealed an increase in BOLD signal variability in the supplementary motor area (SMA) in FND patients who had an improved clinical outcome, suggesting SMA variability as a potential state biomarker. Additionally, higher BOLD signal variability in the left insula at baseline predicted a worse clinical outcome.

Conclusion

This study contributes to the understanding of FND pathophysiology, emphasizing the dynamic nature of neural activity and highlighting the potential of BOLD signal variability as a valuable research tool. The insula and SMA emerge as promising regions for further investigation as prognostic and state markers.

Background

The effects of low-moderate prenatal alcohol exposure (PAE) on brain development have been infrequently studied.

Aim

To compare cortical and white matter structure between children aged 6 to 8 years with low-moderate PAE in trimester 1 only, low-moderate PAE throughout gestation, or no PAE.

Methods

Women reported quantity and frequency of alcohol consumption before and during pregnancy. Magnetic resonance imaging was undertaken for 143 children aged 6 to 8 years with PAE during trimester 1 only (n = 44), PAE throughout gestation (n = 58), and no PAE (n = 41). T₁-weighted images were processed using FreeSurfer, obtaining brain volume, area, and thickness of 34 cortical regions per hemisphere. Fibre density (FD), fibre cross-section (FC) and fibre density and cross-section (FDC) metrics were computed for diffusion images. Brain measures were compared between PAE groups adjusted for age and sex, then additionally for intracranial volume.

Results

After adjustments, the right caudal anterior cingulate cortex volume (p_FDR = 0.045) and area (p_FDR = 0.008), and right cingulum tract cross-sectional area (p_FWE < 0.05) were smaller in children exposed to alcohol throughout gestation compared with no PAE.

Conclusion

This study reports a relationship between low-moderate PAE throughout gestation and cingulate cortex and cingulum tract alterations, suggesting a teratogenic vulnerability. Further investigation is warranted.

Right hemisphere stroke patients frequently present with a combination of lateralised and non-lateralised attentional deficits characteristic of the neglect syndrome. Attentional deficits are associated with poor functional outcome and are challenging to treat, with non-lateralised deficits often persisting into the chronic stage and representing a common complaint among patients and families.

In this study, we investigated the effects of non-invasive brain stimulation on non-lateralised attentional deficits in right-hemispheric stroke. In a randomised double-blind sham-controlled crossover study, twenty-two patients received real and sham transcranial Direct Current Stimulation (tDCS) whilst performing a non-lateralised attentional task. A high definition tDCS montage guided by stimulation modelling was employed to maximise current delivery over the right dorsolateral prefrontal cortex, a key node in the vigilance network. In a parallel study, we examined brain network response to this tDCS montage by carrying out concurrent fMRI during stimulation in healthy participants and patients.

At the group level, stimulation improved target detection in patients, reducing overall error rate when compared with sham stimulation. TDCS boosted performance throughout the duration of the task, with its effects briefly outlasting stimulation cessation. Exploratory lesion analysis indicated that response to stimulation was related to lesion location rather than volume. In particular, reduced stimulation response was associated with damage to the thalamus and postcentral gyrus. Concurrent stimulation-fMRI revealed that tDCS did not affect local connectivity but influenced functional connectivity within large-scale networks in the contralesional hemisphere.

This combined behavioural and functional imaging approach shows that brain stimulation targeted to surviving tissue in the ipsilesional hemisphere improves non-lateralised attentional deficits following stroke. This effect may be exerted via contralesional network effects.

Alzheimer’s disease (AD) and its related age at onset (AAO) are highly heterogeneous, due to the inherent complexity of the disease. They are affected by multiple factors, such as neuroimaging and genetic predisposition. Multimodal integration of various data types is necessary; however, it has been nontrivial due to the high dimensionality of each modality. We aimed to identify multimodal biomarkers of AAO in AD using an extended version of sparse canonical correlation analysis, in which we integrated two imaging modalities, functional magnetic resonance imaging (fMRI) and positron emission tomography (PET), and genetic data in the form of single-nucleotide polymorphisms (SNPs) obtained from the Alzheimer’s disease neuroimaging initiative database. These three modalities cover low-to-high-level complementary information and offer multiscale insights into the AAO. We identified multivariate markers of AAO in AD using fMRI, PET, and SNP. Furthermore, the markers identified were largely consistent with those reported in the existing literature. In particular, our serial mediation analysis suggests that genetic variants influence the AAO in AD by indirectly affecting brain connectivity by mediation of amyloid-beta protein accumulation, supporting a plausible path in existing research. Our approach provides comprehensive biomarkers related to AAO in AD and offers novel multimodal insights into AD.

The VASARI MRI feature set is a quantitative system designed to standardise glioma imaging descriptions. Though effective, deriving VASARI is time-consuming and seldom used clinically. We sought to resolve this problem with software automation and machine learning. Using glioma data from 1172 patients, we developed VASARI-auto, an automated labelling software applied to open-source lesion masks and an openly available tumour segmentation model. Consultant neuroradiologists independently quantified VASARI features in 100 held-out glioblastoma cases. We quantified 1) agreement across neuroradiologists and VASARI-auto, 2) software equity, 3) an economic workforce analysis, and 4) fidelity in predicting survival. Tumour segmentation was compatible with the current state of the art and equally performant regardless of age or sex. A modest inter-rater variability between in-house neuroradiologists was comparable to between neuroradiologists and VASARI-auto, with far higher agreement between VASARI-auto methods. The time for neuroradiologists to derive VASARI was substantially higher than VASARI-auto (mean time per case 317 vs. 3 s). A UK hospital workforce analysis forecast that three years of VASARI featurisation would demand 29,777 consultant neuroradiologist workforce hours and >£1.5 ($1.9) million, reducible to 332 hours of computing time (and £146 of power) with VASARI-auto. The best-performing survival model utilised VASARI-auto features instead of those derived by neuroradiologists. VASARI-auto is a highly efficient and equitable automated labelling system, a favourable economic profile if used as a decision support tool, and non-inferior survival prediction. Future work should iterate upon and integrate such tools to enhance patient care.

Research Purpose

Principal Results

Major Conclusions

Background

Methods

Results

Conclusions