Magnetic resonance imaging-based hippocampus volume for prediction of dementia in mild cognitive impairment: Why does the measurement method matter so little?

Identification of Parkinson's disease and multiple system atrophy using multimodal PET/MRI radiomics

OBJECTIVES

To construct a machine learning model for differentiating Parkinson's disease (PD) and multiple system atrophy (MSA) by using multimodal PET/MRI radiomics and clinical characteristics.

METHODS

One hundred and nineteen patients (81 with PD and 38 with MSA) underwent brain PET/CT and MRI to obtain metabolic images ([18F]FDG, [11C]CFT PET) and structural MRI (T1WI, T2WI, and T2-FLAIR). Image analysis included automatic segmentation on MRI, co-registration of PET images onto the corresponding MRI. Radiomics features were then extracted from the putamina and caudate nuclei and selected to construct predictive models. Moreover, based on PET/MRI radiomics and clinical characteristics, we developed a nomogram. Receiver operating characteristic (ROC) curves were performed to evaluate the performance of the models. Decision curve analysis (DCA) was employed to access the clinical usefulness of the models.

RESULTS

The combined PET/MRI radiomics model of five sequences outperformed monomodal radiomics models alone. Further, PET/MRI radiomics-clinical combined model could perfectly distinguish PD from MSA (AUC = 0.993), which outperformed the clinical model (AUC = 0.923, p = 0.028) in training set, with no significant difference in test set (AUC = 0.860 vs 0.917, p = 0.390). However, no significant difference was found between PET/MRI radiomics-clinical model and PET/MRI radiomics model in training (AUC = 0.988, p = 0.276) and test sets (AUC = 0.860 vs 0.845, p = 0.632). DCA demonstrated the highest clinical benefit of PET/MRI radiomics-clinical model.

CONCLUSIONS

Our study indicates that multimodal PET/MRI radiomics could achieve promising performance to differentiate between PD and MSA in clinics.

CLINICAL RELEVANCE STATEMENT

This study developed an optimal radiomics signature and construct model to distinguish PD from MSA by multimodal PET/MRI imaging methods in clinics for parkinsonian syndromes, which achieved an excellent performance.

KEY POINTS

•Multimodal PET/MRI radiomics from putamina and caudate nuclei increase the diagnostic efficiency for distinguishing PD from MSA. •The radiomics-based nomogram was developed to differentiate between PD and MSA. •Combining PET/MRI radiomics-clinical model achieved promising performance to identify PD and MSA.

Heritability of Subcortical Grey Matter Structures

Background and Objectives: Subcortical grey matter structures play essential roles in cognitive, affective, social, and motoric functions in humans. Their volume changes with age, and decreased volumes have been linked with many neuropsychiatric disorders. The aim of our study was to examine the heritability of six subcortical brain volumes (the amygdala, caudate nucleus, pallidum, putamen, thalamus, and nucleus accumbens) and four general brain volumes (the total intra-cranial volume and the grey matter, white matter, and cerebrospinal fluid (CSF) volume) in twins. Materials and Methods: A total of 118 healthy adult twins from the Hungarian Twin Registry (86 monozygotic and 32 dizygotic; median age 50 ± 27 years) underwent brain magnetic resonance imaging. Two automated volumetry pipelines, Computational Anatomy Toolbox 12 (CAT12) and volBrain, were used to calculate the subcortical and general brain volumes from three-dimensional T1-weighted images. Age- and sex-adjusted monozygotic and dizygotic intra-pair correlations were calculated, and the univariate ACE model was applied. Pearson's correlation test was used to compare the results obtained by the two pipelines. Results: The age- and sex-adjusted heritability estimates, using CAT12 for the amygdala, caudate nucleus, pallidum, putamen, and nucleus accumbens, were between 0.75 and 0.95. The thalamus volume was more strongly influenced by common environmental factors (C = 0.45-0.73). The heritability estimates, using volBrain, were between 0.69 and 0.92 for the nucleus accumbens, pallidum, putamen, right amygdala, and caudate nucleus. The left amygdala and thalamus were more strongly influenced by common environmental factors (C = 0.72-0.85). A strong correlation between CAT12 and volBrain (r = 0.74-0.94) was obtained for all volumes. Conclusions: The majority of examined subcortical volumes appeared to be strongly heritable. The thalamus was more strongly influenced by common environmental factors when investigated with both segmentation methods. Our results underline the importance of identifying the relevant genes responsible for variations in the subcortical structure volume and associated diseases.

Comparing SPM12 and CAT12 segmentation pipelines: a brain tissue volume-based age and Alzheimer's disease study

BACKGROUND

Brain volumes have been used as research biomarkers both in health and in Alzheimer's disease(AD). In order to improve the comparability between studies and aid future analytical software platform choice in the research setting, here we compare two segmentation pipelines of structural brain magnetic resonance imaging(sMRI): the SPM12 toolbox, and a SPM12 add-on, the CAT12 toolbox.

METHODS

We segmented 1.5T and 3T T1-weighted sMRI images (from the OASIS-brain database) using both pipelines and compared them in terms of their impact on: 1)the effect of age on the total grey matter(GM) and white matter(WM), and on the hippocampi GM volumes in a healthy sample(n = 238); 2)the effect of AD diagnosis on the same volume measures; and 3)the accuracy of each volume measure detecting diagnosis (100 patients with AD and 78 age- and gender-matched healthy subjects).

RESULTS AND COMPARISON BETWEEN METHODS

Our results demonstrated that: 1)volume estimates from SPM12 were highly correlated with the ones from CAT12, albeit absolute differences between pipelines were tissue specific; 2)the choice of pipeline modulated the effect of age on all volume measures and of diagnosis on hippocampi GM volumes computed from 3 T data; and 3)pipeline had no impact on the accuracy of any brain volume measure detecting AD diagnosis.

CONCLUSIONS

Our findings indicate that other studies should take these pipeline effects on age and AD diagnosis, into account, for improved comparability in previous literature. Additionally, we encourage future studies to use CAT12 as this is a more advanced and computationally efficient brain segmentation tool.