Comparison of Anatomical and Diffusion MRI for detecting Parkinson's Disease using Deep Convolutional Neural Network

Parkinson's disease (PD) is a progressive neurodegenerative disease that affects over 10 million people worldwide. Brain atrophy and microstructural abnormalities tend to be more subtle in PD than in other age-related conditions such as Alzheimer's disease, so there is interest in how well machine learning methods can detect PD in radiological scans. Deep learning models based on convolutional neural networks (CNNs) can automatically distil diagnostically useful features from raw MRI scans, but most CNN-based deep learning models have only been tested on T1-weighted brain MRI. Here we examine the added value of diffusion-weighted MRI (dMRI) - a variant of MRI, sensitive to microstructural tissue properties - as an additional input in CNN-based models for PD classification. Our evaluations used data from 3 separate cohorts - from Chang Gung University, the University of Pennsylvania, and the PPMI dataset. We trained CNNs on various combinations of these cohorts to find the best predictive model. Although tests on more diverse data are warranted, deep-learned models from dMRI show promise for PD classification.Clinical Relevance- This study supports the use of diffusion-weighted images as an alternative to anatomical images for AI-based detection of Parkinson's disease.

Comparison of Anatomical and Diffusion MRI for detecting Parkinson's Disease using Deep Convolutional Neural Network

Parkinson's disease (PD) is a progressive neurodegenerative disease that affects over 10 million people worldwide. Brain atrophy and microstructural abnormalities tend to be more subtle in PD than in other age-related conditions such as Alzheimer's disease, so there is interest in how well machine learning methods can detect PD in radiological scans. Deep learning models based on convolutional neural networks (CNNs) can automatically distil diagnostically useful features from raw MRI scans, but most CNN-based deep learning models have only been tested on T1-weighted brain MRI. Here we examine the added value of diffusion-weighted MRI (dMRI) - a variant of MRI, sensitive to microstructural tissue properties - as an additional input in CNN-based models for PD classification. Our evaluations used data from 3 separate cohorts - from Chang Gung University, the University of Pennsylvania, and the PPMI dataset. We trained CNNs on various combinations of these cohorts to find the best predictive model. Although tests on more diverse data are warranted, deep-learned models from dMRI show promise for PD classification.

Clinical Relevance

This study supports the use of diffusion-weighted images as an alternative to anatomical images for AI-based detection of Parkinson's disease.

Local molecular and global connectomic contributions to cross-disorder cortical abnormalities

Numerous brain disorders demonstrate structural brain abnormalities, which are thought to arise from molecular perturbations or connectome miswiring. The unique and shared contributions of these molecular and connectomic vulnerabilities to brain disorders remain unknown, and has yet to be studied in a single multi-disorder framework. Using MRI morphometry from the ENIGMA consortium, we construct maps of cortical abnormalities for thirteen neurodevelopmental, neurological, and psychiatric disorders from N = 21,000 participants and N = 26,000 controls, collected using a harmonised processing protocol. We systematically compare cortical maps to multiple micro-architectural measures, including gene expression, neurotransmitter density, metabolism, and myelination (molecular vulnerability), as well as global connectomic measures including number of connections, centrality, and connection diversity (connectomic vulnerability). We find a relationship between molecular vulnerability and white-matter architecture that drives cortical disorder profiles. Local attributes, particularly neurotransmitter receptor profiles, constitute the best predictors of both disorder-specific cortical morphology and cross-disorder similarity. Finally, we find that cross-disorder abnormalities are consistently subtended by a small subset of network epicentres in bilateral sensory-motor, inferior temporal lobe, precuneus, and superior parietal cortex. Collectively, our results highlight how local molecular attributes and global connectivity jointly shape cross-disorder cortical abnormalities.

Thalamic Subregions and Obsessive-Compulsive Symptoms in 2,500 Children From the General Population

OBJECTIVE

Pediatric obsessive-compulsive disorder (OCD) and clinically relevant obsessive-compulsive symptoms in the general population are associated with increased thalamic volume. It is unknown whether this enlargement is explained by specific thalamic subregions. The relation between obsessive-compulsive symptoms and volume of thalamic subregions was investigated in a population-based sample of children.

METHOD

Obsessive-compulsive symptoms were measured in children (9-12 years of age) from the Generation R Study using the Short Obsessive-Compulsive Disorder Screener (SOCS). Thalamic nuclei volumes were extracted from structural 3T magnetic resonance imaging scans using the ThalamicNuclei pipeline and regrouped into anterior, ventral, intralaminar/medial, lateral, and pulvinar subregions. Volumes were compared between children with symptoms above clinical cutoff (probable OCD cases, SOCS ≥ 6, n = 156) and matched children without symptoms (n = 156). Linear regression models were fitted to investigate the association between continuous SOCS score and subregional volume in the whole sample (N = 2500).

RESULTS

Children with probable OCD had larger ventral nuclei compared with children without symptoms (d = 0.25, p = .025, false discovery rate adjusted p = .126). SOCS score showed a negative association with pulvinar volume when accounting for overall thalamic volume (β = -0.057, p = .009, false discovery rate adjusted p = .09). However, these associations did not survive multiple testing correction.

CONCLUSION

The results suggest that individual nuclei groups contribute in varying degrees to overall thalamic volume in children with probable OCD, although this did not survive multiple comparisons correction. Understanding the role of thalamic nuclei and their associated circuits in pediatric OCD could lead toward treatment strategies targeting these circuits.

"Eyes Open - Eyes Closed" EEG/fMRI data set including dedicated "Carbon Wire Loop" motion detection channels

This data set contains electroencephalography (EEG) data as well as simultaneous EEG with functional magnetic resonance imaging (EEG/fMRI) data. During EEG/fMRI, the EEG cap was outfitted with a hardware-based add-on consisting of carbon-wire loops (CWL). These yielded six extra׳CWL׳ signals related to Faraday induction of these loops in the main magnetic field “Measurement and reduction of motion and ballistocardiogram artefacts from simultaneous EEG and fMRI recordings” (Masterton et al., 2007) [1].

In this data set, the CWL data make it possible to do a direct regression approach to deal with the BCG and specifically He artifact.

The CWL-EEG/fMRI data in this paper has been recorded on two MRI scanners with different Helium pump systems (4 subjects on a 3 T TIM Trio and 4 subjects on a 3T VERIO). Separate EEG/fMRI data sets have been recorded for the helium pump ON as well as the helium pump OFF conditions. The EEG-only data (same subjects) has been recorded for a motion artifact-free reference EEG signal outside of the scanner.

This paper also links to an EEGlab “EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis” (Delorme and Makeig, 2004) [2] plugin to perform a CWL regression approach to deal with the He pump artifact, as published in the main paper “Carbon-wire loop based artifact correction outperforms post-processing EEG/fMRI corrections-A validation of a real-time simultaneous EEG/fMRI correction method” (van der Meer et al., 2016) [3].

Cold hands, warm feet: sleep deprivation disrupts thermoregulation and its association with vigilance

STUDY OBJECTIVES

Vigilance is affected by induced and spontaneous skin temperature fluctuations. Whereas sleep deprivation strongly affects vigilance, no previous study examined in detail its effect on human skin temperature fluctuations and their association with vigilance.

DESIGN

In a repeated-measures constant routine design, skin temperatures were assessed continuously from 14 locations while performance was assessed using a reaction time task, including eyes-open video monitoring, performed five times a day for 2 days, after a normal sleep or sleep deprivation night.

SETTING

Participants were seated in a dimly lit, temperature-controlled laboratory.

PATIENTS OR PARTICIPANTS

Eight healthy young adults (five males, age 22.0 ± 1.8 yr (mean ± standard deviation)).

INTERVENTION

One night of sleep deprivation.

MEASUREMENTS AND RESULTS

Mixed-effect regression models were used to evaluate the effect of sleep deprivation on skin temperature gradients of the upper (ear-mastoid), middle (hand-arm), and lower (foot-leg) body, and on the association between fluctuations in performance and in temperature gradients. Sleep deprivation induced a marked dissociation of thermoregulatory skin temperature gradients, indicative of attenuated heat loss from the hands co-occurring with enhanced heat loss from the feet. Sleep deprivation moreover attenuated the association between fluctuations in performance and temperature gradients; the association was best preserved for the upper body gradient.

CONCLUSIONS

Sleep deprivation disrupts coordination of fluctuations in thermoregulatory skin temperature gradients. The dissociation of middle and lower body temperature gradients may therefore be evaluated as a marker for sleep debt, and the upper body gradient as a possible aid in vigilance assessment when sleep debt is unknown. Importantly, our findings suggest that sleep deprivation affects the coordination between skin blood flow fluctuations and the baroreceptor-mediated cardiovascular regulation that prevents venous pooling of blood in the lower limbs when there is the orthostatic challenge of an upright posture.