Search

It is commonly asserted that MRI-derived lesion masks outperform CT-derived lesion masks in lesion-mapping analysis. However, no quantitative analysis has been conducted to support or refute this claim. This study reports an objective comparison of lesion-mapping analyses based on CT- and MRI-derived lesion masks to clarify how input imaging type may ultimately impact analysis results.Routine CT and MRI data were collected from 85 acute stroke survivors. These data were employed to create binarized lesion masks and conduct lesion-mapping analyses on simulated behavioral data. Following standard lesion-mapping analysis methodology, each voxel or region of interest (ROI) were considered as the underlying “target” within CT and MRI data independently. The resulting thresholded z-maps were compared between matched CT- and MRI-based analyses. Paired MRI- and CT-derived lesion masks were found to exhibit good agreement in location, overlap, and size. In single-voxel simulations, MRI-based lesion mapping was able to include more voxels than CT-based analyses, but CT-based analysis results were closer to the underlying target voxel. Simulated lesion symptom mapping results yielded by paired CT and MRI lesion symptom mapping analyses demonstrated good agreement in terms of Dice coefficient when systematic differences in cluster size and lesion overlay are considered. In ROI-level simulations, both CT and MRI-derived analyses yielded low Dice similarity coefficients but no significant differences in performance were identified between modalities.Overall, these results suggest that CT- and MRI-based lesion mapping analyses produce comparable results. This finding is critically important as it suggests that future studies can and should employ CT-derived lesion masks if these scans are available within the appropriate time-window.

BackgroundMonoclonal antibodies against tumour necrosis factor (TNF) markedly reduce inflammation and disease activity in rheumatoid arthritis; however, the mechanisms through which they affect pain are not fully understood.AimsThe aim of this study was to investigate how monoclonal antibodies against TNF alter pain processing and to determine whether neuroimaging can be used as a marker of treatment efficacy and a predictor of treatment response.MethodsFunctional magnetic resonance imaging was used to study the neural correlates of clinically-relevant pain evoked by pressing the most painful joint of the right hand and experimental pain evoked by a thermal stimulus applied to the right forearm. A flashing checkerboard was used as a control stimulus. Patients with severe rheumatoid arthritis, qualifying for the anti-TNF treatment, were scanned before the beginning of the therapy and then approximately one and six months after the first injection.ResultsTNF inhibition was associated with a marked reduction in pain ratings, inflammation, disease activity as well as depression and catastrophising scores. Effective treatment was linked with less pressure-evoked brain activation in the regions involved in the processing of the sensory aspect of pain and in the limbic structures. Baseline pressure-evoked activation in the thalamus predicted future response to treatment. There was no reduction in heat-evoked brain activation; on the contrary, there was an increase in the activation in the precuneus, which is involved in interoception. There were no differences in response to the visual stimulus.ConclusionsTNF inhibition strongly reduces brain activation in response to clinically relevant pressure pain but not experimental heat pain and these changes reflect the decrease of nociceptive input from the periphery due to the reduction of inflammation as well as central changes in pain modulation. Neuroimaging methods have the potential to explain and predict treatment effects in inflammatory pain conditions.

Modelling population reference curves or normative modelling is increasingly used with the advent of large neuroimaging studies. In this paper we assess the performance of fitting methods from the perspective of clinical applications and investigate the influence of the sample size. Further, we evaluate linear and nonlinear models for percentile curve estimation and highlight how the bias-variance trade-off manifests in typical neuroimaging data.We created plausible ground truth distributions of hippocampal volumes in the age range of 45 to 80 years, as an example application. Based on these distributions we repeatedly simulated samples for sizes between 50 and 50,000 data points, and for each simulated sample we fitted a range of normative models. We compared the fitted models and their variability across repetitions to the ground truth, with specific focus on the outer percentiles (1th, 5th, 10th) as these are the most clinically relevant.Our results quantify the expected decreasing trend in variance of the volume estimates with increasing sample size. However, bias in the volume estimates only decreases a modest amount, without much improvement at large sample sizes. The uncertainty of model performance is substantial for what would often be considered large samples in a neuroimaging context and rises dramatically at the ends of the age range, where fewer data points exist. Flexible models perform better across sample sizes, especially for nonlinear ground truth.Surprisingly large samples of several thousand data points are needed to accurately capture outlying percentiles across the age range for applications in research and clinical settings. Performance evaluation methods should assess both, bias and variance. Furthermore, extreme caution is needed when attempting to extrapolate beyond the age range included in the source dataset. To help with such evaluations of normative models we have made our code available to guide researchers developing or utilising normative models.

Many applications in image-guided surgery and therapy require fast and reliable non-linear, multi-modal image registration. Recently proposed unsupervised deep learning-based registration methods have demonstrated superior performance compared to iterative methods in just a fraction of the time. Most of the learning-based methods have focused on mono-modal image registration. The extension to multi-modal registration depends on the use of an appropriate similarity function, such as the mutual information (MI). We propose guiding the training of a deep learning-based registration method with MI estimation between an image-pair in an end-to-end trainable network. Our results show that a small, 2-layer network produces competitive results in both mono- and multi-modal registration, with sub-second run-times. Comparisons to both iterative and deep learning-based methods show that our MI-based method produces topologically and qualitatively superior results with an extremely low rate of non-diffeomorphic transformations. Real-time clinical application will benefit from a better visual matching of anatomical structures and less registration failures/outliers., 4 pages, 4 figures, 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), oral presentation

BackgroundAccurate registration between microscopy and MRI data is necessary for validating imaging biomarkers against neuropathology, and to disentangle complex signal dependencies in microstructural MRI. Existing registration methods often rely on serial histological sampling or significant manual input, providing limited scope to work with a large number of stand-alone histology sections. Here we present a customisable pipeline to automate the registration of stand-alone histology sections to whole-brain MRI data.MethodsOur pipeline registers stained histology sections to whole-brain post-mortem MRI in 4 stages, with the help of two photographic intermediaries: a block face image (to undistort histology sections) and coronal brain slice photographs (to insert them into MRI space). Each registration stage is implemented as a configurable stand-alone Python script using our novel platform, Tensor Image Registration Library (TIRL), which provides flexibility for wider adaptation. We report our experience of registering 87 PLP-stained histology sections from 14 subjects and perform various experiments to assess the accuracy and robustness of each stage of the pipeline.ResultsAll 87 histology sections were successfully registered to MRI. Histology-to-block registration (Stage 1) achieved 0.2-0.4 mm accuracy, better than commonly used existing methods. Block-to-slice matching (Stage 2) showed great robustness in automatically identifying and inserting small tissue blocks into whole brain slices with 0.2 mm accuracy. Simulations demonstrated sub-voxel level accuracy (0.13 mm) of the slice-to-volume registration (Stage 3) algorithm, which was observed in over 200 actual brain slice registrations, compensating 3D slice deformations up to 6.5 mm. Stage 4 combined the previous stages and generated refined pixelwise aligned multi-modal histology-MRI stacks.ConclusionsOur open-source pipeline provides robust automation tools for registering stand-alone histology sections to MRI data with sub-voxel level precision, and the underlying framework makes it readily adaptable to a diverse range of microscopy-MRI studies.HighlightsNew software framework for prototyping bespoke image registration pipelinesAutomated pipeline to register stand-alone histology sections to whole-brain MRINovel deformable slice-to-volume registration algorithmNo strict necessity for serial histological sectioning for MRI-histology registration

Accurate registration between microscopy and MRI data is necessary for validating imaging biomarkers against neuropathology, and to disentangle complex signal dependencies in microstructural MRI. Existing registration methods often rely on serial histological sampling or significant manual input, providing limited scope to work with a large number of stand-alone histology sections. Here we present a customisable pipeline to assist the registration of stand-alone histology sections to whole-brain MRI data.Our pipeline registers stained histology sections to whole-brain post-mortem MRI in 4 stages, with the help of two photographic intermediaries: a block face image (to undistort histology sections) and coronal brain slab photographs (to insert them into MRI space). Each registration stage is implemented as a configurable stand-alone Python script using our novel platform, Tensor Image Registration Library (TIRL), which provides flexibility for wider adaptation. We report our experience of registering 87 PLP-stained histology sections from 14 subjects and perform various experiments to assess the accuracy and robustness of each stage of the pipeline.All 87 histology sections were successfully registered to MRI. Histology-to-block registration (Stage 1) achieved 0.2-0.4 mm accuracy, better than commonly used existing methods. Block-to-slice matching (Stage 2) showed great robustness in automatically identifying and inserting small tissue blocks into whole brain slices with 0.2 mm accuracy. Simulations demonstrated sub-voxel level accuracy (0.13 mm) of the slice-to-volume registration (Stage 3) algorithm, which was observed in over 200 actual brain slice registrations, compensating 3D slice deformations up to 6.5 mm. Stage 4 combined the previous stages and generated refined pixelwise aligned multi-modal histology-MRI stacks.Our open-source pipeline provides robust automation tools for registering stand-alone histology sections to MRI data with sub-voxel level precision, and the underlying framework makes it readily adaptable to a diverse range of microscopy-MRI studies.

BackgroundSpinal cord pathology is a major determinant of irreversible disability in progressive multiple sclerosis. The demyelinated lesion is a cardinal feature. Given the well-characterized anatomy of the spinal cord, the study of lesion topography and its extent of inflammatory activity may provide important clues about disease pathogenesis.MethodsWe studied a large cohort of cervical, thoracic, and lumbar spinal cord tissue derived from 119 pathologically confirmed multiple sclerosis cases. Immunohistochemistry was used to detect demyelination (proteolipid protein) and classify its inflammatory activity (CD68). Prevalence and distribution of demyelination, staged by lesion activity, was determined and topographical maps created to identify patterns of pathology using mixed models and permutation-based cluster analysis.ResultsSpinal cord lesions were observed in 76.5% of cases. Cases were more likely to harbour lesions at the cervical level compared to the lumbar region. Inflammatory activity (active or mixed active/inactive) was observed in 87.9% of cases. The distribution of lesions throughout the spinal cord mirrored the vascular network, spared the subpial surface, and consistently affected the dorsal columns, lateral corticospinal tracts, and the central canal. However, when present, subpial demyelination only affected a limited circumference of the spinal cord (ConclusionsWe demonstrate that spinal cord demyelination is common and highly inflammatory, has a predilection for the cervical level, and relates to clinical measures of disability even at late disease stages. Lesions relatively spare the subpial surface and instead preferentially affect the posterior and lateral columns and central canal, which points to a primary role of the vasculature in lesion pathogenesis. These findings fundamentally challenge the notion that demyelinating pathology in end-stage progressive multiple sclerosis is not inflammatory and that spinal cord lesions exhibit an outside-in topographical gradient. Taken together, this study highlights the importance of early intervention to target ongoing inflammatory demyelination and nominates vascular dysfunction as an important potential target to study further.

We previously reported that neuropathic pain was associated with smaller posterior cingulate cortical (PCC) volumes, suggesting that a smaller/dysfunctional PCC may contribute to development of pain via impaired mind wandering. A gap in our previous report was lack of evidence for a mechanism for the genesis of PCC atrophy in HIV peripheral neuropathy. Here we investigate if volumetric differences in the subcortex for those with neuropathic paresthesia may contribute to smaller PCC volumes, potentially through deafferentation of ascending white matter tracts resulting from peripheral nerve damage in HIV neuropathy. Since neuropathic pain and paresthesia are highly correlated, statistical decomposition was used to separate pain and paresthesia symptoms to determine which regions of brain atrophy are associated with both pain and paresthesia and which are associated separately with pain or paresthesia. HIV+ individuals (N = 233) with and without paresthesia in a multisite study underwent structural brain magnetic resonance imaging. Voxel-based morphometry and a segmentation/registration tool were used to investigate regional brain volume changes associated with paresthesia. Analysis of decomposed variables found that smaller midbrain and thalamus volumes were associated with paresthesia rather than pain. However, atrophy in the PCC was related to both pain and paresthesia. Peak thalamic atrophy (p = 0.004; MNI x = − 14, y = − 24, z = − 2) for more severe paresthesia was in a region with reciprocal connections with the PCC. This provides initial evidence that smaller PCC volumes in HIV peripheral neuropathy are related to ascending white matter deafferentation caused by small fiber damage observed in HIV peripheral neuropathy.

Anorexia Nervosa (AN) is an eating disorder with high mortality and morbidity rates, partly due to treatment resistance and high relapse rates. Treatment adherence and recovery has been found to be hindered by insight deficits, a lack of appreciation of one’s illness, or its consequences, most frequent in restrictive AN. However, to date, insight disturbances in AN have mainly been studied in relation to treatment outcomes rather than explanatory mechanisms. One possibility is that interoception (the sensing, awareness and interpretation of physiological signals) and particularly its metacognitive aspects such as prospective (self-efficacy) and retrospective (insight) beliefs about one’s interoceptive abilities may be affected in AN. To our knowledge however such aspects of global metacognition, and their relation to key interoceptive and body perception impairments, have not been assessed in AN. Here in two experiments (nAN=51 and 28, nAN-WR=47 and 21, nHC=63 and 34, respectively), we tested, (a) how women with and weight-restored from AN (AN-WR), in comparison to healthy controls (HCs), formulate explicit interoceptive self-efficacy beliefs (i.e., estimates of performance in a cardiac perception task) prospectively and then update them following performance and then following explicit feedback and (b) how they formulate prospectively and then update following feedback two types of body-size beliefs (estimates about the envisioned body, ‘How thin it looks' vs the emotional body, ‘How thin it feels’). Results of Experiment 1 confirmed our hypotheses that the AN (but not the AN-WR) group formulated more pessimistic interoceptive self-efficacy beliefs in comparison to HCs both before and after otherwise comparable performance. In Experiment 2 we found that the AN group envisioned and felt (also the AN-WR group) their body size to be bigger than it really is in comparison to controls. Post-feedback, the AN but not AN-WR group significantly overestimated both their envisioned and emotional body and they also updated their emotional body size estimates at a slower rate than the HCs. These observed group differences in belief updating about interoceptive self-efficacy and body size estimates warrant further studies in interoceptive metacognition and belief updating in AN, and their relation with insight deficits, particularly at the acute stages of the disease.

Transient ischaemic attack (TIA) may be a warning sign of stroke and difficult to differentiate from minor stroke and TIA-mimics. Urgent evaluation and diagnosis is important as treating TIA early can prevent subsequent strokes. Recent improvements in mass spectrometer technology allow quantification of hundreds of plasma proteins and lipids, yielding large datasets that would benefit from different approaches including machine learning. Using plasma protein, lipid and radiological biomarkers, our study will develop predictive algorithms to distinguish TIA from minor stroke (positive control) and TIA-mimics (negative control). Analysis including machine learning employs more sophisticated modelling, allowing non-linear interactions, adapting to datasets and enabling development of multiple specialised test-panels for identification and differentiation.Patients attending the Emergency Department, Stroke Ward or TIA Clinic at the Royal Adelaide Hospital with TIA, minor stroke or TIA-like symptoms will be recruited consecutively by staff-alert for this prospective cohort study. Advanced neuroimaging will be performed for each participant, with images assessed independently by up to three expert neurologists. Venous blood samples will be collected within 48 hours of symptom onset. Plasma proteomic and lipid analysis will use advanced mass spectrometry (MS) techniques. Principal component analysis and hierarchical cluster analysis will be performed using MS software. Output files will be analysed for relative biomarker quantitative differences between the three groups. Differences will be assessed by linear regression, one-way analysis of variance, Kruskal-Wallis H-test, χPatients will provide written informed consent to participate in this grant-funded study. The Central Adelaide Local Health Network Human Research Ethics Committee approved this study (HREC/18/CALHN/384; R20180618). Findings will be disseminated through peer-reviewed publication and conferences; data will be managed according to our Data Management Plan (DMP2020-00062).

The therapeutic effects of touch have been long reported. However, in the field of psychotherapy touch is the exception and talking therapies are the norm. Critically, evidence on clients’ experiences and perspectives of touch during psychotherapy is scarce and reliant on small samples. Moreover, despite converging evidence on the associations between touch and attachment, research on attachment traits and perceptions of psychotherapeutic touch is lacking. Here, we utilised the largest-to-date, UK survey on touch (N = 39254), identifying 6878 individuals reporting having received psychological therapy in the last 10 years, to explore 1) the perceived quantity and affective quality of therapeutic touch experiences, and 2) clients’ desire to be touched. We hypothesised that the above experiences and desires are moderated by therapeutic modality, adult attachment style dimensions and general touch attitudes. We found that 30% of the responders reported physical contact with their therapist, 70% of the sample reported that touch communicated support by the therapist, while 4% of clients reported it was inappropriate. In addition, higher scores in attachment-avoidance were negatively associated with affective quality of touch experiences. In relation to touch desire, 40% of our sample wanted their therapist to touch them, with individuals scoring higher on attachment-anxiety style being more likely to show desire for touch, whereas attachment avoidance reduced desire. Having cognitive-behavioural therapy reduced the desire for touch, whereas having body-oriented therapy increased it. This unprecedented, large scale data warrant further investigation on the potential usefulness of touch interventions in certain clients and in given modalities.

Research on segmentation of the hippocampus in magnetic resonance images through deep learning convolutional neural networks (CNNs) shows promising results, suggesting that these methods can identify small structural abnormalities of the hippocampus, which are among the earliest and most frequent brain changes associated with Alzheimer disease (AD). However, CNNs typically achieve the highest accuracy on datasets acquired from the same domain as the training dataset. Transfer learning allows domain adaptation through further training on a limited dataset. In this study, we applied transfer learning on a network called spatial warping network segmentation (SWANS), developed and trained in a previous study. We used MR images of patients with clinical diagnoses of mild cognitive impairment (MCI) and AD, segmented by two different raters. By using transfer learning techniques, we developed four new models, using different training methods. Testing was performed using 26% of the original dataset, which was excluded from training as a hold-out test set. In addition, 10% of the overall training dataset was used as a hold-out validation set. Results showed that all the new models achieved better hippocampal segmentation quality than the baseline SWANS model (ps

Acquisition of high quality manual annotations is vital for the development of segmentation algorithms. However, to create them we require a substantial amount of expert time and knowledge. Large numbers of labels are required to train convolutional neural networks due to the vast number of parameters that must be learned in the optimisation process. Here, we develop the STAMP algorithm to allow the simultaneous training and pruning of a UNet architecture for medical image segmentation with targeted channelwise dropout to make the network robust to the pruning. We demonstrate the technique across segmentation tasks and imaging modalities. It is then shown that, through online pruning, we are able to train networks to have much higher performance than the equivalent standard UNet models while reducing their size by more than 85% in terms of parameters. This has the potential to allow networks to be directly trained on datasets where very low numbers of labels are available.

Cerebral microbleeds (CMBs) are associated with white matter damage, various neu-rodegenerative and cerebrovascular diseases. CMBs occur as small, circular hypointense lesions on T2*-weighted gradient recalled echo (GRE) and susceptibility weighted imaging (SWI) images, and hyperintense on quantitative susceptibility mapping (QSM) images due to their paramagnetic nature. Accurate detection of CMBs would help to determine the CMB lesion count and distribution, which would be further useful to understand the clinical impact of CMBs and to obtain quantitative imaging biomarkers. In this work, we propose a fully automated, deep learning-based, 2-step algorithm, using structural and anatomical properties of CMBs from any single input image modality (e.g. GRE / SWI / QSM) for their accurate detection. Our method consists of an initial candidate detection step, that detects CMBs with high sensitivity and a candidate discrimination step using a knowledge distillation framework to classify CMB and non-CMB instances, followed by a morphological clean-up step. We used 4 datasets consisting of different modalities specified above, acquired using various protocols and with a variety of pathological and demographic characteristics. On cross-validation within datasets, our method achieved a cluster-wise true positive rate (TPR) over 90% with an average of less than 2 false positives per subject. Our method is flexible in terms of the input modality and provides comparable cluster-wise TPR and a better cluster-wise precision compared to existing state-of-the-art methods. When evaluated across different datasets, our method showed good generalisability with a cluster-wise TPR greater than 80% with different modalities.

The quantification of subcortical volume development from 3D fetal ultrasound can provide important diagnostic information during pregnancy monitoring. However, manual segmentation of subcortical structures in ultra-sound volumes is time-consuming and challenging due to low soft tissue contrast, speckle and shadowing artifacts. For this reason, we developed a convolutional neural network (CNN) for the automated segmentation of the choroid plexus (CP), lateral posterior ventricle horns (LPVH), cavum septum pellucidum et vergae (CSPV), and cerebellum (CB) from 3D ultrasound. As ground-truth labels are scarce and expensive to obtain, we applied few-shot learning, in which only a small number of manual annotations (n = 9) are used to train a CNN. We compared training a CNN with only a few individually annotated volumes versus many weakly labelled volumes obtained from atlas-based segmentations. This showed that segmentation performance close to intra-observer variability can be obtained with only a handful of manual annotations. Finally, the trained models were applied to a large number (n = 278) of ultrasound image volumes of a diverse, healthy population, obtaining novel US-specific growth curves of the respective structures during the second trimester of gestation.

Cerebral microbleeds (CMBs) appear as small, circular, well defined hypointense lesions of a few mm in size on T2*-weighted gradient recalled echo (T2*-GRE) images and appear enhanced on susceptibility weighted images (SWI). Due to their small size, contrast variations and other mimics (e.g. blood vessels), CMBs are highly challenging to detect automatically. In large datasets (e.g. the UK Biobank dataset), exhaustively labelling CMBs manually is difficult and time consuming. Hence it would be useful to preselect candidate CMB subjects in order to focus on those for manual labelling, which is essential for training and testing automated CMB detection tools on these datasets. In this work, we aim to detect CMB candidate subjects from a larger dataset, UK Biobank, using a machine learning-based, computationally light pipeline. For our evaluation, we used 3 different datasets, with different intensity characteristics, acquired with different scanners. They include the UK Biobank dataset and two clinical datasets with different pathological conditions. We developed and evaluated our pipelines on different types of images, consisting of SWI or GRE images. We also used the UK Biobank dataset to compare our approach with alternative CMB preselection methods using non-imaging factors and/or imaging data. Finally, we evaluated the pipeline’s generalisability across datasets. Our method provided subject-level detection accuracy > 80% on all the datasets (withindataset results), and showed good generalisability across datasets, providing a consistent accuracy of over 80%, even when evaluated across different modalities.

Cranial cavity extraction is often the first step in quantitative neuroimaging analyses. However, few automated, validated extraction tools have been developed for non-contrast enhanced CT scans (NECT). The purpose of this study was to compare and contrast freely available tools in an unseen dataset of real-world clinical NECT head scans in order to assess the performance and generalisability of these tools. This study included data from a demographically representative sample of 428 patients who had completed NECT scans following hospitalisation for stroke. In a subset of the scans (n = 20), the intracranial spaces were segmented using automated tools and compared to the gold standard of manual delineation to calculate accuracy, precision, recall, and dice similarity coefficient (DSC) values. Further, three readers independently performed regional visual comparisons of the quality of the results in a larger dataset (n = 428). Three tools were found; one of these had unreliable performance so subsequent evaluation was discontinued. The remaining tools included one that was adapted from the FMRIB software library (fBET) and a convolutional neural network- based tool (rBET). Quantitative comparison showed comparable accuracy, precision, recall and DSC values (fBET: 0.984 ± 0.002; rBET: 0.984 ± 0.003; p = 0.99) between the tools; however, intracranial volume was overestimated. Visual comparisons identified characteristic regional differences in the resulting cranial cavity segmentations. Overall fBET had highest visual quality ratings and was preferred by the readers in the majority of subject results (84%). However, both tools produced high quality extractions of the intracranial space and our findings should improve confidence in these automated CT tools. Pre- and post-processing techniques may further improve these results.

In recent decades, the research traditions of (first-person) embodied cognition and of (third-person) social cognition have approached the study of self-awareness with relative independence. However, neurological disorders of self-awareness offer a unifying perspective to empirically investigate the contribution of embodiment and social cognition to self-awareness. This study focused on a neuropsychological disorder of bodily self-awareness following right-hemisphere damage, namely anosognosia for hemiplegia (AHP). A previous neuropsychological study has shown AHP patients, relative to neurological controls, to have a specific deficit in third-person, allocentric inferences in a story-based, mentalisation task. However, no study has tested directly whether verbal awareness of motor deficits is influenced by either perspective-taking or centrism, and if these deficits in social cognition are correlated with damage to anatomical areas previously linked to mentalising, including the supramarginal and superior temporal gyri and related limbic white matter connections. Accordingly, two novel experiments were conducted with right-hemisphere stroke patients with (n = 17) and without AHP (n = 17) that targeted either their own (egocentric, experiment 1) or another stooge patient’s (experiment 2) motor abilities from a first-or-third person (allocentric in Experiment 2) perspective. In both experiments, neurological controls showed no significant difference between perspectives, suggesting that perspective-taking deficits are not a general consequence of right-hemisphere damage. More specifically, experiment 1 found AHP patients were more aware of their own motor paralysis when asked from a third compared to a first-person perspective, using both group level and individual level analysis. In experiment 2, AHP patients were less accurate than controls in making allocentric, third-person perspective judgements about the stooge patient, but with only a trend towards significance and with no within-group, difference between perspectives. Deficits in egocentric and allocentric third-person perspective taking were associated with lesions in the middle frontal gyrus, superior temporal and supramarginal gyri, with white matter disconnections more predominate in deficits in allocentricity. This study confirms previous clinical and empirical investigations on the selectivity of first-person motor awareness deficits in anosognosia for hemiplegia and experimentally demonstrates for the first time that verbal egocentric 3PP-taking can positively influence 1PP body awareness.

usc Research into machine learning (ML) for clinical vascular analysis, such as those useful for stroke and coronary artery disease, varies greatly between imaging modalities and vascular regions. Limited accessibility to large diverse patient imaging datasets, as well as a lack of transparency in specific methods, are obstacles to further development. This paper reviews the current status of quantitative vascular ML, identifying advantages and disadvantages common to all imaging modalities. Literature from the past 8 years was systematically collected from MEDLINE® and Scopus database searches in January 2021. Papers satisfying all search criteria, including a minimum of 50 patients, were further analysed and extracted of relevant data, for a total of 47 publications. Current ML image segmentation, disease risk prediction, and pathology quantitation methods have shown sensitivities and specificities over 70%, compared to expert manual analysis or invasive quantitation. Despite this, inconsistencies in methodology and the reporting of results have prevented inter-model comparison, impeding the identification of approaches with the greatest potential. The clinical potential of this technology has been well demonstrated in Computed Tomography of coronary artery disease, but remains practically limited in other modalities and body regions, particularly due to a lack of routine invasive reference measurements and patient datasets. Refereed/Peer-reviewed

Cerebral microbleeds (CMBs) appear as small, circular, well defined hypointense lesions of a few mm in size on T2*-weighted gradient recalled echo (T2*-GRE) images and appear enhanced on susceptibility weighted images (SWI). Due to their small size, contrast variations and other mimics (e.g., blood vessels), CMBs are highly challenging to detect automatically. In large datasets (e.g., the UK Biobank dataset), exhaustively labelling CMBs manually is difficult and time consuming. Hence it would be useful to preselect candidate CMB subjects in order to focus on those for manual labelling, which is essential for training and testing automated CMB detection tools on these datasets. In this work, we aim to detect CMB candidate subjects from a larger dataset, UK Biobank, using a machine learning-based, computationally light pipeline. For our evaluation, we used 3 different datasets, with different intensity characteristics, acquired with different scanners. They include the UK Biobank dataset and two clinical datasets with different pathological conditions. We developed and evaluated our pipelines on different types of images, consisting of SWI or GRE images. We also used the UK Biobank dataset to compare our approach with alternative CMB preselection methods using non-imaging factors and/or imaging data. Finally, we evaluated the pipeline's generalisability across datasets. Our method provided subject-level detection accuracy > 80% on all the datasets (within-dataset results), and showed good generalisability across datasets, providing a consistent accuracy of over 80%, even when evaluated across different modalities.

Gliomas appear with wide variation in their characteristics both in terms of their appearance and location on brain MR images, which makes robust tumour segmentation highly challenging, and leads to high inter-rater variability even in manual segmentations. In this work, we propose a triplanar ensemble network, with an independent tumour core prediction module, for accurate segmentation of these tumours and their sub-regions. On evaluating our method on the MICCAI Brain Tumor Segmentation (BraTS) challenge validation dataset, for tumour sub-regions, we achieved a Dice similarity coefficient of 0.77 for both enhancing tumour (ET) and tumour core (TC). In the case of the whole tumour (WT) region, we achieved a Dice value of 0.89, which is on par with the top-ranking methods from BraTS'17-19. Our method achieved an evaluation score that was the equal 5th highest value (with our method ranking in 10th place) in the BraTS'20 challenge, with mean Dice values of 0.81, 0.89 and 0.84 on ET, WT and TC regions respectively on the BraTS'20 unseen test dataset.

Longitudinal imaging forms an essential component in the management and follow-up of many medical conditions. The presence of lesion changes on serial imaging can have significant impact on clinical decision making, highlighting the important role for automated change detection. Lesion changes can represent anomalies in serial imaging, which implies a limited availability of annotations and a wide variety of possible changes that need to be considered. Hence, we introduce a new unsupervised anomaly detection and localisation method trained exclusively with serial images that do not contain any lesion changes. Our training automatically synthesises lesion changes in serial images, introducing detection and localisation pseudo-labels that are used to self-supervise the training of our model. Given the rarity of these lesion changes in the synthesised images, we train the model with the imbalance robust focal Tversky loss. When compared to supervised models trained on different datasets, our method shows competitive performance in the detection and localisation of new demyelinating lesions on longitudinal magnetic resonance imaging in multiple sclerosis patients. Code for the models will be made available at https://github.com/toson87/MSChangeDetection.

The combination of deep learning image analysis methods and large-scale imaging datasets offers many opportunities to imaging neuroscience and epidemiology. However, despite the success of deep learning when applied to many neuroimaging tasks, there remain barriers to the clinical translation of large-scale datasets and processing tools. Here, we explore the main challenges and the approaches that have been explored to overcome them. We focus on issues relating to data availability, interpretability, evaluation and logistical challenges, and discuss the challenges we believe are still to be overcome to enable the full success of big data deep learning approaches to be experienced outside of the research field.

Highlights • We harmonised measures of WMHs across two studies on healthy ageing. • Specific pre-processing strategies can increase comparability across studies. • Modelling of biological differences is crucial to provide calibrated measures., Large scale neuroimaging datasets present the possibility of providing normative distributions for a wide variety of neuroimaging markers, which would vastly improve the clinical utility of these measures. However, a major challenge is our current poor ability to integrate measures across different large-scale datasets, due to inconsistencies in imaging and non-imaging measures across the different protocols and populations. Here we explore the harmonisation of white matter hyperintensity (WMH) measures across two major studies of healthy elderly populations, the Whitehall II imaging sub-study and the UK Biobank. We identify pre-processing strategies that maximise the consistency across datasets and utilise multivariate regression to characterise study sample differences contributing to differences in WMH variations across studies. We also present a parser to harmonise WMH-relevant non-imaging variables across the two datasets. We show that we can provide highly calibrated WMH measures from these datasets with: (1) the inclusion of a number of specific standardised processing steps; and (2) appropriate modelling of sample differences through the alignment of demographic, cognitive and physiological variables. These results open up a wide range of applications for the study of WMHs and other neuroimaging markers across extensive databases of clinical data.

BackgroundThe medium-term effects of Coronavirus disease (COVID-19) on multiple organ health, exercise capacity, cognition, quality of life and mental health are poorly understood.MethodsFifty-eight COVID-19 patients post-hospital discharge and 30 comorbidity-matched controls were prospectively enrolled for multiorgan (brain, lungs, heart, liver and kidneys) magnetic resonance imaging (MRI), spirometry, six-minute walk test, cardiopulmonary exercise test (CPET), quality of life, cognitive and mental health assessments.FindingsAt 2-3 months from disease-onset, 64% of patients experienced persistent breathlessness and 55% complained of significant fatigue. On MRI, tissue signal abnormalities were seen in the lungs (60%), heart (26%), liver (10%) and kidneys (29%) of patients. COVID-19 patients also exhibited tissue changes in the thalamus, posterior thalamic radiations and sagittal stratum on brain MRI and demonstrated impaired cognitive performance, specifically in the executive and visuospatial domain relative to controls. Exercise tolerance (maximal oxygen consumption and ventilatory efficiency on CPET) and six-minute walk distance (405±118m vs 517±106m in controls, pInterpretationA significant proportion of COVID-19 patients discharged from hospital experience ongoing symptoms of breathlessness, fatigue, anxiety, depression and exercise limitation at 2-3 months from disease-onset. Persistent lung and extra-pulmonary organ MRI findings are common. In COVID-19 survivors, chronic inflammation may underlie multiorgan abnormalities and contribute to impaired quality of life.FundingNIHR Oxford and Oxford Health Biomedical Research Centres, British Heart Foundation Centre for Research Excellence, UKRI, Wellcome Trust, British Heart Foundation.

Neuromyelitis optica associated with aquaporin-4-antibodies (NMOSD-AQP4) and myelin oligodentrocyte-glycoprotein antibody-associated disorder (MOGAD) have been recently recognised as different from multiple sclerosis.Although conventional MRI may help distinguish multiple sclerosis from antibody-mediated diseases, the use of quantitative and non-conventional imaging may give more pathological information and explain the clinical differences.We compared, using non-conventional imaging, brain MRI findings in 75 subjects in remission with NMOSD-AQP4, MOGAD, multiple sclerosis or healthy controls (HC). Volumetrics, white matter and cortical lesions, and tissue integrity measures using diffusion imaging, were analysed in the four groups along with their association with disability (expanded disability status scale [EDSS] and visual acuity).The volumetric analysis showed that, deep grey matter volumes were significantly lower in multiple sclerosis (p=0.0001) and MOGAD (p=0.02), compared to HC. Relapsing MOGAD had lower white matter, pallidus and hippocampus volumes than in monophasic (pOptic chiasm volume was reduced only in NMOSD-AQP4 who had at least one episode of optic neuritis (ON) (NMOSD-AQP4-ON vs NMOSD-AQP4 pLesion volume was greatest in multiple sclerosis followed by MOGAD and in these two diseases, the lesion volume correlated with disease duration (multiple sclerosis R=0.46, p=0.05, MOGAD R=0.81, pLesional-fractional anisotropy (FA) was reduced and mean diffusivity increased in all patients, but overall, FA was only reduced in the non-lesional tissue in multiple sclerosis (p=0.01), although focal reductions were noted in NMOSD-AQP4, reflecting mainly optic nerve and corticospinal tract pathways.Cortical/juxtacortical lesions were seen in a minority of MOGAD, while cortical/juxtacortical and purely cortical lesions were identified in the majority of multiple sclerosis and in none of the NMOSD-AQP4.Non-lesional FA in NMOSD-AQP4, lower white-matter volume and female sex in multiple sclerosis, and lower brainstem volume in MOGAD were the best predictors of EDSS disability (accounting for 46%, 49% and 19% respectively). Worse visual acuity associated with lower optic chiasm volume in NMOSD-AQP4 and lower thalamus volume in MOGAD (accounting for 58% and 35% respectively).Although MOGAD patients had good outcomes, deep grey matter atrophy was present. In contrast, NMOSD-AQP4 patients showed a relative sparing of deep grey matter volumes, despite greater residual disability as compared with MOGAD patients. NMOSD-AQP4 but not MOGAD patients showed reduced FA in non-lesional tissue.

Large scale neuroimaging datasets present the possibility of providing normative distributions for a wide variety of neuroimaging markers, which would vastly improve the clinical utility of these measures. However, a major challenge is our current poor ability to integrate measures across different large-scale datasets, due to inconsistencies in imaging and non-imaging measures across the different protocols and populations. Here we explore the harmonisation of white matter hyperintensity (WMH) measures across two major studies of healthy elderly populations, the Whitehall II imaging sub-study and the UK Biobank. We identify pre-processing strategies that maximise the consistency across datasets and utilise multivariate regression to characterise sample differences contributing to study-level differences in WMH variations. We also present a parser to harmonise WMH-relevant non-imaging variables across the two datasets. We show that we can provide highly calibrated WMH measures from these datasets with: (1) the inclusion of a number of specific standardised processing steps; and (2) appropriate modelling of sample differences through the alignment of demographic, cognitive and physiological variables. These results open up a wide range of applications for the study of WMHs and other neuroimaging markers across extensive databases of clinical data.HIGHLIGHTSWe harmonised measures of WMHs across two studies on healthy ageingSpecific pre-processing strategies can increase comparability across studiesModelling of biological differences is crucial to provide calibrated measures

White matter hyperintensities (WMHs) have been associated with various cerebrovascular and neurodegenerative diseases. Reliable quantification of WMHs is essential for understanding their clinical impact in normal and pathological populations. Automated segmentation of WMHs is highly challenging due to heterogeneity in WMH characteristics between deep and periventricular white matter, presence of artefacts and differences in the pathology and demographics of populations. In this work, we propose an ensemble triplanar network that combines the predictions from three different planes of brain MR images to provide an accurate WMH segmentation. Also, the network uses anatomical information regarding WMH spatial distribution in loss functions for improving the efficiency of segmentation and to overcome the contrast variations between deep and periventricular WMHs. We evaluated our method on 5 datasets, of which 3 are part of a publicly available dataset (training data for MICCAI WMH Segmentation Challenge 2017 - MWSC 2017) consisting of subjects from three different cohorts. On evaluating our method separately in deep and periventricular regions, we observed robust and comparable performance in both regions. Our method performed better than most of the existing methods, including FSL BIANCA, and on par with the top ranking deep learning method of MWSC 2017.

Artificial intelligence is one of the most exciting methodological shifts in our era. It holds the potential to transform healthcare as we know it, to a system where humans and machines work together to provide better treatment for our patients. It is now clear that cutting edge artificial intelligence models in conjunction with high-quality clinical data will lead to improved prognostic and diagnostic models in neurological disease, facilitating expert-level clinical decision tools across healthcare settings. Despite the clinical promise of artificial intelligence, machine and deep-learning algorithms are not a one-size-fits-all solution for all types of clinical data and questions. In this article, we provide an overview of the core concepts of artificial intelligence, particularly contemporary deep-learning methods, to give clinician and neuroscience researchers an appreciation of how artificial intelligence can be harnessed to support clinical decisions. We clarify and emphasize the data quality and the human expertise needed to build robust clinical artificial intelligence models in neurology. As artificial intelligence is a rapidly evolving field, we take the opportunity to iterate important ethical principles to guide the field of medicine is it moves into an artificial intelligence enhanced future., Artificial intelligence can transform healthcare as we know it, to a system where humans and machines work together to facilitating expert-level clinical decision tools across healthcare settings. We review contemporary deep-learning methods and ethical principles in artificial intelligence, as we are moving into the future of precision medicine., Graphical Abstract Graphical Abstract

The brain is a highly complex system that is inaccessible to biopsy, which puts human brain imaging at the heart of our attempts to understand psychiatric disorders. Imaging has the potential to uncover the pathophysiology, provide biomarkers for use in the development and monitoring of treatments, and stratify patients for studies and trials. This chapter introduces the three main brain imaging technologies that are used to assay brain structure and function: magnetic resonance imaging (MRI), molecular imaging positron emission tomography (PET), and single-photon emission computed tomography (SPECT); electrophysiology [electroencephoaography (EEG)]; and magnetoencephalograpy (MEG). The chapter outlines the principles behind their use and the nature of the information that can be extracted. Together, these brain imaging methods can provide complementary windows into the living brain as an increasingly essential suite of tools for experimental medicine in psychiatry.

IntroductionTransient ischaemic attack (TIA) may be a warning sign of stroke and difficult to differentiate from minor stroke and TIA-mimics. Urgent evaluation and diagnosis is important as treating TIA early can prevent subsequent strokes. Recent improvements in mass spectrometer technology allow quantification of hundreds of plasma proteins and lipids, yielding large datasets that would benefit from different approaches including machine learning. Using plasma protein, lipid and radiological biomarkers, our study will develop predictive algorithms to distinguish TIA from minor stroke (positive control) and TIA-mimics (negative control). Analysis including machine learning employs more sophisticated modelling, allowing non-linear interactions, adapting to datasets and enabling development of multiple specialised test-panels for identification and differentiation.Methods and analysisPatients attending the Emergency Department, Stroke Ward or TIA Clinic at the Royal Adelaide Hospital with TIA, minor stroke or TIA-like symptoms will be recruited consecutively by staff-alert for this prospective cohort study. Advanced neuroimaging will be performed for each participant, with images assessed independently by up to three expert neurologists. Venous blood samples will be collected within 48 hours of symptom onset. Plasma proteomic and lipid analysis will use advanced mass spectrometry (MS) techniques. Principal component analysis and hierarchical cluster analysis will be performed using MS software. Output files will be analysed for relative biomarker quantitative differences between the three groups. Differences will be assessed by linear regression, one-way analysis of variance, Kruskal-Wallis H-test, χ2test or Fisher’s exact test. Machine learning methods will also be applied including deep learning using neural networks.Ethics and disseminationPatients will provide written informed consent to participate in this grant-funded study. The Central Adelaide Local Health Network Human Research Ethics Committee approved this study (HREC/18/CALHN/384; R20180618). Findings will be disseminated through peer-reviewed publication and conferences; data will be managed according to our Data Management Plan (DMP2020-00062).

The Developing Human Connectome Project (dHCP) seeks to create the first 4-dimensional connectome of early life. Understanding this connectome in detail may provide insights into normal as well as abnormal patterns of brain development. Following established best practices adopted by the WU-MINN Human Connectome Project (HCP), and pioneered by FreeSurfer, the project utilises cortical surface-based processing pipelines. In this paper, we propose a fully automated processing pipeline for the structural Magnetic Resonance Imaging (MRI) of the developing neonatal brain. This proposed pipeline consists of a refined framework for cortical and sub-cortical volume segmentation, cortical surface extraction, and cortical surface inflation, which has been specifically designed to address considerable differences between adult and neonatal brains, as imaged using MRI. Using the proposed pipeline our results demonstrate that images collected from 465 subjects ranging from 28 to 45 weeks post-menstrual age (PMA) can be processed fully automatically; generating cortical surface models that are topologically correct, and correspond well with manual evaluations of tissue boundaries in 85% of cases. Results improve on state-of-the-art neonatal tissue segmentation models and significant errors were found in only 2% of cases, where these corresponded to subjects with high motion. Downstream, these surfaces will enhance comparisons of functional and diffusion MRI datasets, supporting the modelling of emerging patterns of brain connectivity.

Stuttering is a disorder in which the smooth flow of speech is interrupted. People who stutter show structural and functional abnormalities in the speech and motor system. It is unclear whether functional differences reflect general traits of the disorder or are specifically related to the dysfluent speech state. We used a hierarchical approach to separate state and trait effects within stuttering. We collected sparse‐sampled functional MRI during two overt speech tasks (sentence reading and picture description) in 17 people who stutter and 16 fluent controls. Separate analyses identified indicators of: (1) general traits of people who stutter; (2) frequency of dysfluent speech states in subgroups of people who stutter; and (3) the differences between fluent and dysfluent states in people who stutter. We found that reduced activation of left auditory cortex, inferior frontal cortex bilaterally, and medial cerebellum were general traits that distinguished fluent speech in people who stutter from that of controls. The stuttering subgroup with higher frequency of dysfluent states during scanning (n = 9) had reduced activation in the right subcortical grey matter, left temporo‐occipital cortex, the cingulate cortex, and medial parieto‐occipital cortex relative to the subgroup who were more fluent (n = 8). Finally, during dysfluent states relative to fluent ones, there was greater activation of inferior frontal and premotor cortex extending into the frontal operculum, bilaterally. The above differences were seen across both tasks. Subcortical state effects differed according to the task. Overall, our data emphasise the independence of trait and state effects in stuttering.

Highlights • We explored various domain adaptation methods for robust WM lesion segmentation. • We used a triplanar U-net ensemble network (TrUE-Net) as our baseline model. • Transfer learning: fine-tuning from the coarsest encoder layer gave good results. • Semi-supervised domain adversarial training of NNs (DANN) performed the best. • Among unsupervised methods, DANN performed better than domain unlearning., Graphical abstract, Robust automated segmentation of white matter hyperintensities (WMHs) in different datasets (domains) is highly challenging due to differences in acquisition (scanner, sequence), population (WMH amount and location) and limited availability of manual segmentations to train supervised algorithms. In this work we explore various domain adaptation techniques such as transfer learning and domain adversarial learning methods, including domain adversarial neural networks and domain unlearning, to improve the generalisability of our recently proposed triplanar ensemble network, which is our baseline model. We used datasets with variations in intensity profile, lesion characteristics and acquired using different scanners. For the source domain, we considered a dataset consisting of data acquired from 3 different scanners, while the target domain consisted of 2 datasets. We evaluated the domain adaptation techniques on the target domain datasets, and additionally evaluated the performance on the source domain test dataset for the adversarial techniques. For transfer learning, we also studied various training options such as minimal number of unfrozen layers and subjects required for fine-tuning in the target domain. On comparing the performance of different techniques on the target dataset, domain adversarial training of neural network gave the best performance, making the technique promising for robust WMH segmentation.

In this article, SIENA-XL, a new segmentation-based longitudinal pipeline is introduced, for: (i) increasing the precision of longitudinal volume change estimation for white (WM) and gray (GM) matter separately, compared with cross-sectional segmentation methods such as SIENAX; and (ii) avoiding potential biases in registration-based methods when Jacobians are used, with a smoothing extent larger than spatial scale between tissue-interfaces, which is where atrophy usually occurs. SIENA-XL implements a new brain extraction procedure and a multi-time-point intensity equalization step before performing the final segmentation that also includes separate segmentation of deep GM structures by using FMRIB's Integrated Registration and Segmentation Tool. The detection of GM and WM volume changes with SIENA-XL was evaluated using different healthy control (HC) and multiple sclerosis (MS) MRI datasets and compared with the traditional SIENAX and two Jacobian-based approaches, SPM12 and SIENAX-JI (a version of SIENAX including Jacobian integration - JI). In scan-rescan data from HCs, SIENA-XL showed: (i) a significant decrease in error, of 50-70% when compared with SIENAX; (ii) no significant differences in error when compared with SIENAX-JI and SPM12 in a scan-rescan HC dataset that included repositioning. When tested in a HC dataset with scan-rescan both at baseline and after 1 year of follow-up, SIENA-XL showed: (i) significantly higher precision (P < 0.01) than SIENAX; (ii) no significant differences to SIENAX-JI and SPM12. Finally, in a dataset of 79 MS patients with a 2 years follow-up, SIENA-XL showed a substantial reduction of sample size, by comparison with SIENAX, SIENAX-JI, and SPM12, for detecting treatment effects of 25, 30, and 50%. Hum Brain Mapp 39:1063-1077, 2018. © 2017 Wiley Periodicals, Inc.

Rationale and Objectives T2-weighted MRI hyperintensity assessed visually in the corticospinal tract (CST) lacks sensitivity for a diagnosis of amyotrophic lateral sclerosis (ALS). We sought to explore a quantitative approach to FLAIR MRI intensity across a range of ALS phenotypes. Materials & Methods Thirty three classical ALS patients, 10 with a flail arm presentation, and 6 with primary lateral sclerosis underwent MRI at 3 Tesla. Comparisons of quantitative FLAIR intensity in the corticospinal tract (CST) and corpus callosum were made between 21 healthy controls and within patient phenotypic sub-groups, some of who were studied longitudinally. Results Mean FLAIR intensity was greater in patient groups. The cerebral peduncle intensity provided the strongest sub-group classification. FLAIR intensity increased longitudinally. The rate of change of FLAIR within CST correlated with rate of decline in executive function and ALS Functional Rating Score. Conclusion FLAIR MRI encodes quantifiable information of potential diagnostic, stratification and monitoring value.