Measurement of brain structures with artificial neural networks: two- and three-dimensional applications

Detection of Marchiafava Bignami disease using distinct deep learning techniques in medical diagnostics

PURPOSE

To detect the Marchiafava Bignami Disease (MBD) using a distinct deep learning technique.

BACKGROUND

Advanced deep learning methods are becoming more crucial in contemporary medical diagnostics, particularly for detecting intricate and uncommon neurological illnesses such as MBD. This rare neurodegenerative disorder, sometimes associated with persistent alcoholism, is characterized by the loss of myelin or tissue death in the corpus callosum. It poses significant diagnostic difficulties owing to its infrequency and the subtle signs it exhibits in its first stages, both clinically and on radiological scans.

METHODS

The novel method of Variational Autoencoders (VAEs) in conjunction with attention mechanisms is used to identify MBD peculiar diseases accurately. VAEs are well-known for their proficiency in unsupervised learning and anomaly detection. They excel at analyzing extensive brain imaging datasets to uncover subtle patterns and abnormalities that traditional diagnostic approaches may overlook, especially those related to specific diseases. The use of attention mechanisms enhances this technique, enabling the model to concentrate on the most crucial elements of the imaging data, similar to the discerning observation of a skilled radiologist. Thus, we utilized the VAE with attention mechanisms in this study to detect MBD. Such a combination enables the prompt identification of MBD and assists in formulating more customized and efficient treatment strategies.

RESULTS

A significant breakthrough in this field is the creation of a VAE equipped with attention mechanisms, which has shown outstanding performance by achieving accuracy rates of over 90% in accurately differentiating MBD from other neurodegenerative disorders.

CONCLUSION

This model, which underwent training using a diverse range of MRI images, has shown a notable level of sensitivity and specificity, significantly minimizing the frequency of false positive results and strengthening the confidence and dependability of these sophisticated automated diagnostic tools.

Radiologists and Clinical Trials: Part 2: Practical Statistical Methods for Understanding and Monitoring Independent Reader Performance

Though many clinical trials rely on medical image evaluations for primary or key secondary endpoints, the methods to monitor reader performance are all too often mired in the legacy use of adjudication rates. If misused, this simple metric can be misleading and sometimes entirely contradictory. Furthermore, attempts to overcome the limitations of adjudication rates using de novo or ad hoc methods often ignore well-established research conducted over the last half-century and can lead to inaccurate conclusions or variable interpretations. Underperforming readers can be missed, expert readers retrained, or worse, replaced. This paper aims to standardize reader performance evaluations using proven statistical methods. Additionally, these methods will describe how to discriminate between scenarios of concern and normal medical interpretation variability. Statistical methods are provided for inter-reader and intra-reader variability and bias, including the adjudicator's bias. Finally, we have compiled guidelines for calculating correct sample sizes, considerations for intra-reader memory recall, and applying alternative designs for independent readers.

Deep Learning Corpus Callosum Segmentation as a Neurodegenerative Marker in Multiple Sclerosis

BACKGROUND AND PURPOSE

Corpus callosum atrophy is a sensitive biomarker of multiple sclerosis (MS) neurodegeneration but typically requires manual 2D or volumetric 3D-based segmentations. We developed a supervised machine learning algorithm, DeepnCCA, for corpus callosum segmentation and relate callosal morphology to clinical disability using conventional MRI scans collected in clinical routine.

METHODS

In a prospective study of 553 MS patients with 704 acquisitions, 200 unique 2D T₂ -weighted MRI scans were delineated to develop, train, and validate DeepnCCA. Comparative FreeSurfer segmentations were obtained in 504 3D T₁ -weighted scans. Both FreeSurfer and DeepnCCA outputs were correlated with clinical disability. Using principal component analysis of the DeepnCCA output, the morphological changes were explored in relation to clinical disease burden.

RESULTS

DeepnCCA and manual segmentations had high similarity (Dice coefficients 98.1 ± .11%, 89.3 ± .76%, for intracranial and corpus callosum area, respectively through 10-fold cross-validation). DeepnCCA had numerically stronger correlations with cognitive and physical disability as compared to FreeSurfer: Expanded disability status scale (EDSS) ±6 months (r = -.22 P = .002; r = -.17, P = .013), future EDSS (r = -.26, P<.001; r = -.17, P = .012), and future symbol digit modalities test (r = .26, P = .001; r = .24, P = .003). The corpus callosum became thinner with increasing cognitive and physical disability. Increasing physical disability, additionally, significantly correlated with a more angled corpus callosum.

CONCLUSIONS

DeepnCCA (https://github.com/plattenmichael/DeepnCCA/) is an openly available tool that can provide fast and accurate corpus callosum measurements applicable to large MS cohorts, potentially suitable for monitoring disease progression and therapy response.

Altered brain structures in the dorsal and ventral language pathways in individuals with and without developmental language disorder (DLD)

Developmental Language Disorder (DLD) is a neurodevelopmental disorder characterized by difficulty learning and using language, and this difficulty cannot be attributed to other developmental conditions. The aim of the current study was to examine structural differences in dorsal and ventral language pathways between adolescents and young adults with and without DLD (age range: 14-27 years) using anatomical magnetic resonance imaging (MRI) and diffusion tensor imaging (DTI). Results showed age-related structural brain differences in both dorsal and ventral pathways in individuals with DLD. These findings provide evidence for neuroanatomical correlates of persistent language deficits in adolescents/young adults with DLD, and further suggest that this brain-language relationship in DLD is better characterized by taking account the dynamic course of the disorder along development.

Deep neural network to locate and segment brain tumors outperformed the expert technicians who created the training data

Purpose: Deep learning (DL) algorithms have shown promising results for brain tumor segmentation in MRI. However, validation is required prior to routine clinical use. We report the first randomized and blinded comparison of DL and trained technician segmentations.

Approach: We compiled a multi-institutional database of 741 pretreatment MRI exams. Each contained a postcontrast T1-weighted exam, a T2-weighted fluid-attenuated inversion recovery exam, and at least one technician-derived tumor segmentation. The database included 729 unique patients (470 males and 259 females). Of these exams, 641 were used for training the DL system, and 100 were reserved for testing. We developed a platform to enable qualitative, blinded, controlled assessment of lesion segmentations made by technicians and the DL method. On this platform, 20 neuroradiologists performed 400 side-by-side comparisons of segmentations on 100 test cases. They scored each segmentation between 0 (poor) and 10 (perfect). Agreement between segmentations from technicians and the DL method was also evaluated quantitatively using the Dice coefficient, which produces values between 0 (no overlap) and 1 (perfect overlap).

Results: The neuroradiologists gave technician and DL segmentations mean scores of 6.97 and 7.31, respectively (p<0.00007). The DL method achieved a mean Dice coefficient of 0.87 on the test cases.

Conclusions: This was the first objective comparison of automated and human segmentation using a blinded controlled assessment study. Our DL system learned to outperform its “human teachers” and produced output that was better, on average, than its training data.

Review on 2D and 3D MRI Image Segmentation Techniques

BACKGROUND

Magnetic Resonance Imaging is most widely used for early diagnosis of abnormalities in human organs. Due to the technical advancement in digital image processing, automatic computer aided medical image segmentation has been widely used in medical diagnostics.

DISCUSSION

Image segmentation is an image processing technique which is used for extracting image features, searching and mining the medical image records for better and accurate medical diagnostics. Commonly used segmentation techniques are threshold based image segmentation, clustering based image segmentation, edge based image segmentation, region based image segmentation, atlas based image segmentation, and artificial neural network based image segmentation.

CONCLUSION

This survey aims at providing an insight about different 2-Dimensional and 3- Dimensional MRI image segmentation techniques and to facilitate better understanding to the people who are new in this field. This comparative study summarizes the benefits and limitations of various segmentation techniques.

Procedural and declarative memory brain systems in developmental language disorder (DLD)

The aim of the current study was to examine microstructural differences in white matter relevant to procedural and declarative memory between adolescents/young adults with and without Developmental Language Disorder (DLD) using diffusion tensor imaging (DTI). The findings showed atypical age-related changes in white matter structures in the corticostriatal system, in the corticocerebellar system, and in the medial temporal region in individuals with DLD. Results highlight the importance of considering the age factor in research on DLD. Future studies are needed to examine the developmental relationship between long-term memory and individual differences in language development and learning.

Quantifying deep grey matter atrophy using automated segmentation approaches: A systematic review of structural MRI studies

The deep grey matter (DGM) nuclei of the brain play a crucial role in learning, behaviour, cognition, movement and memory. Although automated segmentation strategies can provide insight into the impact of multiple neurological conditions affecting these structures, such as Multiple Sclerosis (MS), Huntington's disease (HD), Alzheimer's disease (AD), Parkinson's disease (PD) and Cerebral Palsy (CP), there are a number of technical challenges limiting an accurate automated segmentation of the DGM. Namely, the insufficient contrast of T1 sequences to completely identify the boundaries of these structures, as well as the presence of iso-intense white matter lesions or extensive tissue loss caused by brain injury. Therefore in this systematic review, 269 eligible studies were analysed and compared to determine the optimal approaches for addressing these technical challenges. The automated approaches used among the reviewed studies fall into three broad categories, atlas-based approaches focusing on the accurate alignment of atlas priors, algorithmic approaches which utilise intensity information to a greater extent, and learning-based approaches that require an annotated training set. Studies that utilise freely available software packages such as FIRST, FreeSurfer and LesionTOADS were also eligible, and their performance compared. Overall, deep learning approaches achieved the best overall performance, however these strategies are currently hampered by the lack of large-scale annotated data. Improving model generalisability to new datasets could be achieved in future studies with data augmentation and transfer learning. Multi-atlas approaches provided the second-best performance overall, and may be utilised to construct a "silver standard" annotated training set for deep learning. To address the technical challenges, providing robustness to injury can be improved by using multiple channels, highly elastic diffeomorphic transformations such as LDDMM, and by following atlas-based approaches with an intensity driven refinement of the segmentation, which has been done with the Expectation Maximisation (EM) and level sets methods. Accounting for potential lesions should be achieved with a separate lesion segmentation approach, as in LesionTOADS. Finally, to address the issue of limited contrast, R2*, T2* and QSM sequences could be used to better highlight the DGM due to its higher iron content. Future studies could look to additionally acquire these sequences by retaining the phase information from standard structural scans, or alternatively acquiring these sequences for only a training set, allowing models to learn the "improved" segmentation from T1-sequences alone.

Voxel Deconvolutional Networks for 3D Brain Image Labeling

Deep learning methods have shown great success in pixel-wise prediction tasks. One of the most popular methods employs an encoder-decoder network in which deconvolutional layers are used for up-sampling feature maps. However, a key limitation of the deconvolutional layer is that it suffers from the checkerboard artifact problem, which harms the prediction accuracy. This is caused by the independency among adjacent pixels on the output feature maps. Previous work only solved the checkerboard artifact issue of deconvolutional layers in the 2D space. Since the number of intermediate feature maps needed to generate a deconvolutional layer grows exponentially with dimensionality, it is more challenging to solve this issue in higher dimensions. In this work, we propose the voxel deconvolutional layer (VoxelDCL) to solve the checkerboard artifact problem of deconvolutional layers in 3D space. We also provide an efficient approach to implement VoxelDCL. To demonstrate the effectiveness of VoxelDCL, we build four variations of voxel deconvolutional networks (VoxelDCN) based on the U-Net architecture with VoxelDCL. We apply our networks to address volumetric brain images labeling tasks using the ADNI and LONI LPBA40 datasets. The experimental results show that the proposed iVoxelDCNa achieves improved performance in all experiments. It reaches 83.34% in terms of dice ratio on the ADNI dataset and 79.12% on the LONI LPBA40 dataset, which increases 1.39% and 2.21% respectively compared with the baseline. In addition, all the variations of VoxelDCN we proposed outperform the baseline methods on the above datasets, which demonstrates the effectiveness of our methods.

Computational methods for corpus callosum segmentation on MRI: A systematic literature review

BACKGROUND AND OBJECTIVE

The corpus callosum (CC) is the largest white matter structure in the brain and has a significant role in central nervous system diseases. Its volume correlates with the severity and/or extent of neurodegenerative disease. Even though the CC's role has been extensively studied over the last decades, and different algorithms and methods have been published regarding CC segmentation and parcellation, no reviews or surveys covering such developments have been reported so far. To bridge this gap, this paper presents a systematic literature review of computational methods focusing on CC segmentation and parcellation acquired on magnetic resonance imaging.

METHODS

IEEExplore, PubMed, EBSCO Host, and Scopus database were searched with the following search terms: ((Segmentation OR Parcellation) AND (Corpus Callosum) AND (DTI OR MRI OR Diffusion Tensor Imag* OR Diffusion Tractography OR Magnetic Resonance Imag*)), resulting in 802 publications. Two reviewers independently evaluated all articles and 36 studies were selected through the systematic literature review process.

RESULTS

This work reviewed four main segmentation methods groups: model-based, region-based, thresholding, and machine learning; 32 different validity metrics were reported. Even though model-based techniques are the most recurrently used for the segmentation task (13 articles), machine learning approaches achieved better outcomes of 95% when analyzing mean values for segmentation and classification metrics results. Moreover, CC segmentation is better established in T₁-weighted images, having more methods implemented and also being tested in larger datasets, compared with diffusion tensor images.

CONCLUSIONS

The analyzed computational methods used to perform CC segmentation on magnetic resonance imaging have not yet overcome all presented challenges owing to metrics variability and lack of traceable materials.

Are Anesthesia and Surgery during Infancy Associated with Decreased White Matter Integrity and Volume during Childhood?

Background

Anesthetics have neurotoxic effects in neonatal animals. Relevant human evidence is limited. We sought such evidence in a structural neuroimaging study.

Methods

Two groups of children underwent structural magnetic resonance imaging: patients who, during infancy, had one of four operations commonly performed in otherwise healthy children and comparable, nonexposed control subjects. Total and regional brain tissue composition and volume, as well as regional indicators of white matter integrity (fractional anisotropy and mean diffusivity), were analyzed.

Results

Analyses included 17 patients, without potential confounding central nervous system problems or risk factors, who had general anesthesia and surgery during infancy and 17 control subjects (age ranges, 12.3 to 15.2 yr and 12.6 to 15.1 yr, respectively). Whole brain white matter volume, as a percentage of total intracranial volume, was lower for the exposed than the nonexposed group, 37.3 ± 0.4% and 38.9 ± 0.4% (least squares mean ± SE), respectively, a difference of 1.5 percentage points (95% CI, 0.3 to 2.8; P = 0.016). Corresponding decreases were statistically significant for parietal and occipital lobes, infratentorium, and brainstem separately. White matter integrity was lower for the exposed than the nonexposed group in superior cerebellar peduncle, cerebral peduncle, external capsule, cingulum (cingulate gyrus), and fornix (cres) and/or stria terminalis. The groups did not differ in total intracranial, gray matter, and cerebrospinal fluid volumes.

Conclusions

Children who had anesthesia and surgery during infancy showed broadly distributed, decreased white matter integrity and volume. Although the findings may be related to anesthesia and surgery during infancy, other explanations are possible.

Nonlocal atlas-guided multi-channel forest learning for human brain labeling

PURPOSE

It is important for many quantitative brain studies to label meaningful anatomical regions in MR brain images. However, due to high complexity of brain structures and ambiguous boundaries between different anatomical regions, the anatomical labeling of MR brain images is still quite a challenging task. In many existing label fusion methods, appearance information is widely used. However, since local anatomy in the human brain is often complex, the appearance information alone is limited in characterizing each image point, especially for identifying the same anatomical structure across different subjects. Recent progress in computer vision suggests that the context features can be very useful in identifying an object from a complex scene. In light of this, the authors propose a novel learning-based label fusion method by using both low-level appearance features (computed from the target image) and high-level context features (computed from warped atlases or tentative labeling maps of the target image).

METHODS

In particular, the authors employ a multi-channel random forest to learn the nonlinear relationship between these hybrid features and target labels (i.e., corresponding to certain anatomical structures). Specifically, at each of the iterations, the random forest will output tentative labeling maps of the target image, from which the authors compute spatial label context features and then use in combination with original appearance features of the target image to refine the labeling. Moreover, to accommodate the high inter-subject variations, the authors further extend their learning-based label fusion to a multi-atlas scenario, i.e., they train a random forest for each atlas and then obtain the final labeling result according to the consensus of results from all atlases.

RESULTS

The authors have comprehensively evaluated their method on both public LONI_LBPA40 and IXI datasets. To quantitatively evaluate the labeling accuracy, the authors use the dice similarity coefficient to measure the overlap degree. Their method achieves average overlaps of 82.56% on 54 regions of interest (ROIs) and 79.78% on 80 ROIs, respectively, which significantly outperform the baseline method (random forests), with the average overlaps of 72.48% on 54 ROIs and 72.09% on 80 ROIs, respectively.

CONCLUSIONS

The proposed methods have achieved the highest labeling accuracy, compared to several state-of-the-art methods in the literature.

Implementation of Talairach Atlas Based Automated Brain Segmentation for Radiation Therapy Dosimetry

Radiotherapy for brain cancer inevitably results in irradiation of uninvolved brain. While it has been demonstrated that irradiation of the brain can result in cognitive deficits, dose-volume relationships are not well established. There is little work correlating a particular cognitive deficit with dose received by the region of the brain responsible for the specific cognitive function. One obstacle to such studies is that identification of brain anatomy is both labor intensive and dependent on the individual performing the segmentation. Automatic segmentation has the potential to be both efficient and consistent. Brains2 is a software package developed by the University of Iowa for MRI volumetric studies. It utilizes MR images, the Talairach atlas, and an artificial neural network (ANN) to segment brain images into substructures in a standardized manner. We have developed a software package, Brains2DICOM, that converts the regions of interest identified by Brains2 into a DICOM radiotherapy structure set. The structure set can be imported into a treatment planning system for dosimetry. We demonstrated the utility of Brains2DICOM using a test case, a 34-year-old man with diffuse astrocytoma treated with three-dimensional conformal radiotherapy. Brains2 successfully applied the Talairach atlas to identify the right and left frontal, parietal, temporal, occipital, subcortical, and cerebellum regions. Brains2 was not successful in applying the ANN to identify small structures, such as the hippocampus and caudate. Further work is necessary to revise the ANN or to develop new methods for identification of small structures in the presence of disease and radiation induced changes. The segmented regions-of-interest were transferred to our commercial treatment planning system using DICOM and dose-volume histograms were constructed. This method will facilitate the acquisition of data necessary for the development of normal tissue complication probability (NTCP) models that assess the probability of cognitive complications secondary to radiotherapy for intracranial and head and neck neoplasms.

Automatic segmentation of the striatum and globus pallidus using MIST: Multimodal Image Segmentation Tool

Accurate segmentation of the subcortical structures is frequently required in neuroimaging studies. Most existing methods use only a T1-weighted MRI volume to segment all supported structures and usually rely on a database of training data. We propose a new method that can use multiple image modalities simultaneously and a single reference segmentation for initialisation, without the need for a manually labelled training set. The method models intensity profiles in multiple images around the boundaries of the structure after nonlinear registration. It is trained using a set of unlabelled training data, which may be the same images that are to be segmented, and it can automatically infer the location of the physical boundary using user-specified priors. We show that the method produces high-quality segmentations of the striatum, which is clearly visible on T1-weighted scans, and the globus pallidus, which has poor contrast on such scans. The method compares favourably to existing methods, showing greater overlap with manual segmentations and better consistency.

Examining Procedural Learning and Corticostriatal Pathways for Individual Differences in Language: Testing Endophenotypes of DRD2/ANKK1

The aim of the study was to explore whether genetic variation in the dopaminergic system is associated with procedural learning and the corticostriatal pathways in individuals with developmental language impairment (DLI). We viewed these two systems as endophenotypes and hypothesized that they would be more sensitive indicators of genetic effects than the language phenotype itself. Thus, we genotyped two SNPs in the DRD2/ANKK1 gene complex, and tested for their associations to the phenotype of DLI and the two endophenotypes. Results showed that individuals with DLI revealed poor procedural learning abilities and abnormal structures of the basal ganglia. Genetic variation in DRD2/ANKK1 was associated with procedural learning abilities and with microstructural differences of the caudate nucleus. The association of the language phenotype with these DRD2/ANKK1 polymorphisms was not significant, but the phenotype was significantly associated with the two endophenotypes. We suggest that procedural learning and the corticostriatal pathways could be used as effective endophenotypes to aid molecular genetic studies searching for genes predisposing to DLI.

Dissociable contributions of amygdala and hippocampus to emotion and memory in patients with Alzheimer's disease

The amygdala and the hippocampus are associated with emotional processing and declarative memory, respectively. Studies have shown that patients with bilateral hippocampal damage caused by anoxia/ischemia, and patients with probable Alzheimer's disease (AD), can experience emotions for prolonged periods of time, even when they cannot remember what caused the emotion in the first place (Feinstein et al. (2010) Proc Natl Acad Sci USA 107:7674-7679; Guzmán-Vélez et al. (2014) Cogn Behav Neurol 27:117-129). This study aimed to investigate, for the first time, the roles of the amygdala and hippocampus in the dissociation between feelings of emotion and declarative memory for emotion-inducing events in patients with AD. Individuals with probable AD (N = 12) and age-matched healthy comparisons participants (HCP; N = 12) completed a high-resolution (0.44 × 0.44 × 0.80 mm) T2-weighted structural MR scan of the medial temporal lobe. Each of these individuals also completed two separate emotion induction procedures (sadness and happiness) using film clips. We collected real-time emotion ratings at baseline and multiple times postinduction, and administered a test of declarative memory shortly after each induction. Consistent with previous research, hippocampal volume was significantly smaller in patients with AD compared with HCP, and was positively correlated with memory for the film clips. Sustained feelings of emotion and amygdala volume did not significantly differ between patients with AD and HCP. Follow-up analyses showed a significant negative correlation between amygdala volume and sustained sadness, and a significant positive correlation between amygdala volume and sustained happiness. Our findings suggest that the amygdala is important for regulating and sustaining an emotion independent of hippocampal function and declarative memory for the emotion-inducing event. © 2015 Wiley Periodicals, Inc.

Soft-Split Random Forest for Anatomy Labeling

Random Forest (RF) has been widely used in the learning-based labeling. In RF, each sample is directed from the root to each leaf based on the decisions made in the interior nodes, also called splitting nodes. The splitting nodes assign a testing sample to either left or right child based on the learned splitting function. The final prediction is determined as the average of label probability distributions stored in all arrived leaf nodes. For ambiguous testing samples, which often lie near the splitting boundaries, the conventional splitting function, also referred to as hard split function, tends to make wrong assignments, hence leading to wrong predictions. To overcome this limitation, we propose a novel soft-split random forest (SSRF) framework to improve the reliability of node splitting and finally the accuracy of classification. Specifically, a soft split function is employed to assign a testing sample into both left and right child nodes with their certain probabilities, which can effectively reduce influence of the wrong node assignment on the prediction accuracy. As a result, each testing sample can arrive at multiple leaf nodes, and their respective results can be fused to obtain the final prediction according to the weights accumulated along the path from the root node to each leaf node. Besides, considering the importance of context information, we also adopt a Haar-features based context model to iteratively refine the classification map. We have comprehensively evaluated our method on two public datasets, respectively, for labeling hippocampus in MR images and also labeling three organs in Head & Neck CT images. Compared with the hard-split RF (HSRF), our method achieved a notable improvement in labeling accuracy.

Multi-atlas learner fusion: An efficient segmentation approach for large-scale data

We propose multi-atlas learner fusion (MLF), a framework for rapidly and accurately replicating the highly accurate, yet computationally expensive, multi-atlas segmentation framework based on fusing local learners. In the largest whole-brain multi-atlas study yet reported, multi-atlas segmentations are estimated for a training set of 3464 MR brain images. Using these multi-atlas estimates we (1) estimate a low-dimensional representation for selecting locally appropriate example images, and (2) build AdaBoost learners that map a weak initial segmentation to the multi-atlas segmentation result. Thus, to segment a new target image we project the image into the low-dimensional space, construct a weak initial segmentation, and fuse the trained, locally selected, learners. The MLF framework cuts the runtime on a modern computer from 36 h down to 3-8 min - a 270× speedup - by completely bypassing the need for deformable atlas-target registrations. Additionally, we (1) describe a technique for optimizing the weak initial segmentation and the AdaBoost learning parameters, (2) quantify the ability to replicate the multi-atlas result with mean accuracies approaching the multi-atlas intra-subject reproducibility on a testing set of 380 images, (3) demonstrate significant increases in the reproducibility of intra-subject segmentations when compared to a state-of-the-art multi-atlas framework on a separate reproducibility dataset, (4) show that under the MLF framework the large-scale data model significantly improve the segmentation over the small-scale model under the MLF framework, and (5) indicate that the MLF framework has comparable performance as state-of-the-art multi-atlas segmentation algorithms without using non-local information.

Supervised machine learning-based classification scheme to segment the brainstem on MRI in multicenter brain tumor treatment context

PURPOSE

To constrain the risk of severe toxicity in radiotherapy and radiosurgery, precise volume delineation of organs at risk is required. This task is still manually performed, which is time-consuming and prone to observer variability. To address these issues, and as alternative to atlas-based segmentation methods, machine learning techniques, such as support vector machines (SVM), have been recently presented to segment subcortical structures on magnetic resonance images (MRI).

METHODS

SVM is proposed to segment the brainstem on MRI in multicenter brain cancer context. A dataset composed by 14 adult brain MRI scans is used to evaluate its performance. In addition to spatial and probabilistic information, five different image intensity values (IIVs) configurations are evaluated as features to train the SVM classifier. Segmentation accuracy is evaluated by computing the Dice similarity coefficient (DSC), absolute volumes difference (AVD) and percentage volume difference between automatic and manual contours.

RESULTS

Mean DSC for all proposed IIVs configurations ranged from 0.89 to 0.90. Mean AVD values were below 1.5 cm(3), where the value for best performing IIVs configuration was 0.85 cm(3), representing an absolute mean difference of 3.99% with respect to the manual segmented volumes.

CONCLUSION

Results suggest consistent volume estimation and high spatial similarity with respect to expert delineations. The proposed approach outperformed presented methods to segment the brainstem, not only in volume similarity metrics, but also in segmentation time. Preliminary results showed that the approach might be promising for adoption in clinical use.

Progressive Structural Brain Changes and NRG1 Gene Variants in First-Episode Nonaffective Psychosis

BACKGROUND

Structural brain abnormalities are already present during the early phases of psychosis, but factors underlying brain volume changes are still not well understood. The neuregulin 1 gene (NRG1), influencing neurodevelopment and neuroplasticity, has been associated with schizophrenia. Our aim was to examine whether variations in the NRG1 gene (SNP8NRG221132, SNP8NRG6221533 and SNP8NRG243177 polymorphisms) influence longitudinal changes in the brain during a first episode of psychosis (FEP).

METHODS

A 3-year follow-up magnetic resonance imaging (MRI) study was performed. Fifty-nine minimally medicated patients who were experiencing FEP and 14 healthy control individuals underwent genotyping and structural brain MRI at baseline and at 1- and 3-year follow-up. A comparison of brain volumes, gray matter, white matter (WM), lateral ventricles (LV), cortical cerebrospinal fluid, and thalamus and caudate was made between the groups according to their genotype.

RESULTS

In patients, the SNP8NRG6221533 risk C allele was significantly associated with increased LV volume across time. C allele carriers had significantly less WM compared with subjects homozygous for the T allele after the follow-up. No other significant differences were observed among subgroups. No significant changes according to the genotypes were found in healthy individuals.

CONCLUSION

Our findings suggest that variations of neurodevelopment-related genes, such as the NRG1 gene, can contribute to brain abnormalities described in early phases of schizophrenia and progressive changes during the initial years of the illness. To our knowledge, it is the first time that a relation between NRG1 polymorphisms and longitudinal brain changes is reported. © 2015 S. Karger AG, Basel.

T1ρ imaging in premanifest Huntington disease reveals changes associated with disease progression

BACKGROUND

Imaging biomarkers sensitive to Huntington's disease (HD) during the premanifest phase preceding motor diagnosis may accelerate identification and evaluation of potential therapies. For this purpose, quantitative MRI sensitive to tissue microstructure and metabolism may hold great potential. We investigated the potential value of T1ρ relaxation to detect pathological changes in premanifest HD (preHD) relative to other quantitative relaxation parameters.

METHODS

Quantitative MR parametric mapping was used to assess differences between 50 preHD subjects and 26 age- and sex-matched controls. Subjects with preHD were classified into two progression groups based on their CAG-age product (CAP) score; a high and a low/moderate CAP group. Voxel-wise and region-of-interest analyses were used to assess changes in the quantitative relaxation times.

RESULTS

T1ρ showed a significant increase in the relaxation times in the high-CAP group, as compared to controls, largely in the striatum. The T1ρ changes in the preHD subjects showed a significant relationship with CAP score. No significant changes in T2 or T2* relaxation times were found in the striatum. T2* relaxation changes were found in the globus pallidus, but no significant changes with disease progression were found.

CONCLUSION

These data suggest that quantitative T1ρ mapping may provide a useful marker for assessing disease progression in HD. The absence of T2 changes suggests that the T1ρ abnormalities are unlikely owing to altered water content or tissue structure. The established sensitivity of T1ρ to pH and glucose suggests that these factors are altered in HD perhaps owing to abnormal mitochondrial function.

Regional atrophy associated with cognitive and motor function in prodromal Huntington disease

BACKGROUND

Neuroimaging studies suggest that volumetric MRI measures of specific brain structures may serve as excellent biomarkers in future clinical trials of Huntington disease (HD).

OBJECTIVE

Demonstration of the clinical significance of these measures is an important step in determining their appropriateness as potential outcome measures.

METHODS

Measures of gray- and white-matter lobular volumes and subcortical volumes (caudate, putamen, globus pallidus, thalamus, nucleus accumbens, hippocampus) were obtained from MRI scans of 516 individuals who tested positive for the HD gene expansion, but were not yet exhibiting signs or symptoms severe enough to warrant diagnosis ("pre-HD"). MRI volumes (corrected for intracranial volume) were correlated with cognitive, motor, psychiatric, and functional measures known to be sensitive to subtle changes in pre-HD.

RESULTS

Caudate, putamen, and globus pallidus volumes consistently correlated with cognitive and motor, but not psychiatric or functional measures in pre-HD. Volumes of white matter, nucleus accumbens, and thalamus, but not cortical gray matter, also correlated with some of the motor and cognitive measures.

CONCLUSIONS

Results of regression analyses suggest that volumes of basal ganglia structures contributed more highly to the prediction of most motor and cognitive variables than volumes of other brain regions. These results support the use of volumetric measures, especially of the basal ganglia, as outcome measures in future clinical trials in pre-HD. Results may also assist investigators in selecting the most appropriate measures for treatment trials that target specific clinical features or regions of neuropathology.

Automated MRI cerebellar size measurements using active appearance modeling

Although the human cerebellum has been increasingly identified as an important hub that shows potential for helping in the diagnosis of a large spectrum of disorders, such as alcoholism, autism, and fetal alcohol spectrum disorder, the high costs associated with manual segmentation, and low availability of reliable automated cerebellar segmentation tools, has resulted in a limited focus on cerebellar measurement in human neuroimaging studies. We present here the CATK (Cerebellar Analysis Toolkit), which is based on the Bayesian framework implemented in FMRIB's FIRST. This approach involves training Active Appearance Models (AAMs) using hand-delineated examples. CATK can currently delineate the cerebellar hemispheres and three vermal groups (lobules I-V, VI-VII, and VIII-X). Linear registration with the low-resolution MNI152 template is used to provide initial alignment, and Point Distribution Models (PDM) are parameterized using stellar sampling. The Bayesian approach models the relationship between shape and texture through computation of conditionals in the training set. Our method varies from the FIRST framework in that initial fitting is driven by 1D intensity profile matching, and the conditional likelihood function is subsequently used to refine fitting. The method was developed using T1-weighted images from 63 subjects that were imaged and manually labeled: 43 subjects were scanned once and were used for training models, and 20 subjects were imaged twice (with manual labeling applied to both runs) and used to assess reliability and validity. Intraclass correlation analysis shows that CATK is highly reliable (average test-retest ICCs of 0.96), and offers excellent agreement with the gold standard (average validity ICC of 0.87 against manual labels). Comparisons against an alternative atlas-based approach, SUIT (Spatially Unbiased Infratentorial Template), that registers images with a high-resolution template of the cerebellum, show that our AAM approach offers superior reliability and validity. Extensions of CATK to cerebellar hemisphere parcels are envisioned.

Automatic hippocampus segmentation of 7.0 Tesla MR images by combining multiple atlases and auto-context models

In many neuroscience and clinical studies, accurate measurement of hippocampus is very important to reveal the inter-subject anatomical differences or the subtle intra-subject longitudinal changes due to aging or dementia. Although many automatic segmentation methods have been developed, their performances are still challenged by the poor image contrast of hippocampus in the MR images acquired especially from 1.5 or 3.0 Tesla (T) scanners. With the recent advance of imaging technology, 7.0 T scanner provides much higher image contrast and resolution for hippocampus study. However, the previous methods developed for segmentation of hippocampus from 1.5 T or 3.0 T images do not work for the 7.0 T images, due to different levels of imaging contrast and texture information. In this paper, we present a learning-based algorithm for automatic segmentation of hippocampi from 7.0 T images, by taking advantages of the state-of-the-art multi-atlas framework and also the auto-context model (ACM). Specifically, ACM is performed in each atlas domain to iteratively construct sequences of location-adaptive classifiers by integrating both image appearance and local context features. Due to the plenty texture information in 7.0 T images, more advanced texture features are also extracted and incorporated into the ACM during the training stage. Then, under the multi-atlas segmentation framework, multiple sequences of ACM-based classifiers are trained for all atlases to incorporate the anatomical variability. In the application stage, for a new image, its hippocampus segmentation can be achieved by fusing the labeling results from all atlases, each of which is obtained by applying the atlas-specific ACM-based classifiers. Experimental results on twenty 7.0 T images with the voxel size of 0.35×0.35×0.35 mm3 show very promising hippocampus segmentations (in terms of Dice overlap ratio 89.1±0.020), indicating high applicability for the future clinical and neuroscience studies.

Local label learning (LLL) for subcortical structure segmentation: application to hippocampus segmentation

Automatic and reliable segmentation of subcortical structures is an important but difficult task in quantitative brain image analysis. Multi-atlas based segmentation methods have attracted great interest due to their promising performance. Under the multi-atlas based segmentation framework, using deformation fields generated for registering atlas images onto a target image to be segmented, labels of the atlases are first propagated to the target image space and then fused to get the target image segmentation based on a label fusion strategy. While many label fusion strategies have been developed, most of these methods adopt predefined weighting models that are not necessarily optimal. In this study, we propose a novel local label learning strategy to estimate the target image's segmentation label using statistical machine learning techniques. In particular, we use a L1-regularized support vector machine (SVM) with a k nearest neighbor (kNN) based training sample selection strategy to learn a classifier for each of the target image voxel from its neighboring voxels in the atlases based on both image intensity and texture features. Our method has produced segmentation results consistently better than state-of-the-art label fusion methods in validation experiments on hippocampal segmentation of over 100 MR images obtained from publicly available and in-house datasets. Volumetric analysis has also demonstrated the capability of our method in detecting hippocampal volume changes due to Alzheimer's disease.

Neuroanatomical Differences between First-Episode Psychosis Patients with and without Neurocognitive Deficit: A 3-Year Longitudinal Study

BACKGROUND

The course of cognitive function in first-episode psychosis (FEP) patients suggests that some individuals are normal or near normal whereas some cases present a marked decline. The goal of the present longitudinal study was to identify neuroanatomical differences between deficit and non-deficit patients.

METHODS

Fifty nine FEP patients with neuroimage and neurocognitive information were studied at baseline and 3 year after illness onset. A global cognitive function score was used to classify deficit and non-deficit patients at baseline. Analysis of covariances and repeated-measures analysis were performed to evaluate differences in brain volumes. Age, premorbid IQ, and intracranial volume were used as covariates. We examined only volumes of whole brain, whole brain gray and white matter, cortical CSF and lateral ventricles, lobular volumes of gray and white matter, and subcortical (caudate nucleus and thalamus) regions.

RESULTS

At illness onset 50.8% of patients presented global cognitive deficit. There were no significant differences between neuropsychological subgroups in any of the brain regions studied at baseline [all F(1, 54) ≤ 3.42; all p ≥ 0.07] and follow-up [all F(1, 54) ≤ 3.43; all p ≥ 0.07] time points. There was a significant time by group interaction for the parietal tissue volume [F(1, 54) = 4.97, p = 0.030] and the total gray matter volume [F(1, 54) = 4.31, p = 0.042], with the deficit group showing a greater volume decrease.

CONCLUSION

Our results did not confirm the presence of significant morphometric differences in the brain regions evaluated between cognitively impaired and cognitively preserved schizophrenia patients at the early stages of the illness. However, there were significant time by group interactions for the parietal tissue volume and the total gray matter volume during the 3-year follow-up period, which might indicate that cognitive deficit in schizophrenia would be associated with progressive brain volume loss.

Unawareness of motor phenoconversion in Huntington disease

OBJECTIVE

To determine whether Huntington disease (HD) mutation carriers have motor symptoms (complaints) when definite motor onset (motor phenoconversion) is diagnosed and document differences between the groups with and without unawareness of motor signs.

METHODS

We analyzed data from 550 HD mutation carriers participating in the multicenter PREDICT-HD Study followed through the HD prodrome. Data analysis included demographics, the Unified Huntington's Disease Rating Scale (UHDRS) and the Participant HD History of symptoms, self-report of progression, and cognitive, behavioral, and imaging measures. Unawareness was identified when no motor symptoms were self-reported but when definite motor HD was diagnosed.

RESULTS

Of 38 (6.91%) with onset of motor HD, almost half (18/38 = 47.36%) had no motor symptoms despite signs of disease on the UHDRS motor rating and consistent with unawareness. A group with motor symptoms and signs was similar on a range of measures to the unaware group. Those with unawareness of HD signs reported less depression. Patients with symptoms had more striatal atrophy on imaging measures.

CONCLUSIONS

Only half of the patients with newly diagnosed motor HD had motor symptoms. Unaware patients were less likely to be depressed. Self-report of symptoms may be inaccurate in HD at the earliest stage.

BDNF Val66Met variants and brain volume changes in non-affective psychosis patients and healthy controls: a 3 year follow-up study

INTRODUCTION

Functional gene polymorphisms modulating neuroplasticity might mediate brain longitudinal structural changes in schizophrenia. The present study aimed to explore possible effects of BDNF Val66Met polymorphism variations on progressive structural brain changes after 3 years from the first episode of psychosis.

METHOD

Patients were part of a large epidemiological and longitudinal intervention program of first-episode psychosis, carried out at the University Hospital Marqués de Valdecilla, Cantabria, Spain. Eighty first-episode patients and 54 healthy controls were included in the final analyses. Brain magnetic resonance imaging (baseline and 3-year follow-up) and BDNF genotype, and clinical and functional outcome were investigated.

RESULTS

We did not detect significant association between brain changes and BDNF Val66Met polymorphism variations in patients and controls (all p>0.060). At baseline, there were no significant associations between brain anomalies and BDNF genotype. Functional deficits were similar in Met-carrier and Val homozygote patients after 3-year follow-up (X(2) = 0.66; p = 0.564); there was no relationship between significant volume change across time and functional outcome. Otherwise, Met-carrier controls had significant high rates of alcohol-consumption (p = 0.019) compared to Val homozygote controls.

CONCLUSION

Our findings do not support the notion that BDNF genotype variations may mediate brain macroscopic morphological changes across time.

Abnormal subcortical components of the corticostriatal system in young adults with DLI: a combined structural MRI and DTI study

Developmental Language Impairment (DLI) is a neurodevelopmental disorder affecting 12% to 14% of the school-age children in the United States. While substantial studies have shown a wide range of linguistic and non-linguistic difficulty in individuals with DLI, very little is known about the neuroanatomical mechanisms underlying this disorder. In the current study, we examined the subcortical components of the corticostriatal system in young adults with DLI, including the caudate nucleus, the putamen, the nucleus accumbens, the globus pallidus, and the thalamus. Additionally, the four cerebral lobes and the hippocampus were also comprised for an exploratory analysis. We used conventional magnetic resonance imaging (MRI) to measure regional brain volumes, as well as diffusion tensor imaging (DTI) to assess water diffusion anisotropy as quantified by fractional anisotropy (FA). Two groups of participants, one with DLI (n=12) and the other without (n=12), were recruited from a prior behavioral study, and all were matched on age, gender, and handedness. Volumetric analyses revealed region-specific abnormalities in individuals with DLI, showing pathological enlargement bilaterally in the putamen and the nucleus accumbens, and unilaterally in the right globus pallidus after the intracranial volumes were controlled. Regarding the DTI findings, the DLI group showed decreased FA values in the globus pallidus and the thalamus but these significant differences disappeared after controlling for the whole-brain FA value, indicating that microstructural abnormality is diffuse and affects other regions of the brain. Taken together, these results suggest region-specific corticostriatal abnormalities in DLI at the macrostructural level, but corticostriatal abnormalities at the microstructural level may be a part of a diffuse pattern of brain development. Future work is suggested to investigate the relationship between corticostriatal connectivity and individual differences in language development.

Relapse duration, treatment intensity, and brain tissue loss in schizophrenia: a prospective longitudinal MRI study

OBJECTIVE

Longitudinal structural MRI studies have shown that patients with schizophrenia have progressive brain tissue loss after onset. Recurrent relapses are believed to play a role in this loss, but the relationship between relapse and structural MRI measures has not been rigorously assessed. The authors analyzed longitudinal data to examine this question.

METHODS

The authors studied data from 202 patients drawn from the Iowa Longitudinal Study of first-episode schizophrenia for whom adequate structural MRI data were available (N=659 scans) from scans obtained at regular intervals over an average of 7 years. Because clinical follow-up data were obtained at 6-month intervals, the authors were able to compute measures of relapse number and duration and relate them to structural MRI measures. Because higher treatment intensity has been associated with smaller brain tissue volumes, the authors also examined this countereffect in terms of dose-years.

RESULTS

Relapse duration was related to significant decreases in both general (e.g., total cerebral volume) and regional (e.g., frontal) brain measures. Number of relapses was unrelated to brain measures. Significant effects were also observed for treatment intensity.

CONCLUSIONS

Extended periods of relapse may have a negative effect on brain integrity in schizophrenia, suggesting the importance of implementing proactive measures that may prevent relapse and improve treatment adherence. By examining the relative balance of effects, that is, relapse duration versus antipsychotic treatment intensity, this study sheds light on a troublesome dilemma that clinicians face. Relapse prevention is important, but it should be sustained using the lowest possible medication dosages that will control symptoms.

Statistical epistasis and progressive brain change in schizophrenia: an approach for examining the relationships between multiple genes

Although schizophrenia is generally considered to occur as a consequence of multiple genes that interact with one another, very few methods have been developed to model epistasis. Phenotype definition has also been a major challenge for research on the genetics of schizophrenia. In this report, we use novel statistical techniques to address the high dimensionality of genomic data, and we apply a refinement in phenotype definition by basing it on the occurrence of brain changes during the early course of the illness, as measured by repeated magnetic resonance scans (i.e., an 'intermediate phenotype.') The method combines a machine-learning algorithm, the ensemble method using stochastic gradient boosting, with traditional general linear model statistics. We began with 14 genes that are relevant to schizophrenia, based on association studies or their role in neurodevelopment, and then used statistical techniques to reduce them to five genes and 17 single nucleotide polymorphisms (SNPs) that had a significant statistical interaction: five for PDE4B, four for RELN, four for ERBB4, three for DISC1 and one for NRG1. Five of the SNPs involved in these interactions replicate previous research in that, these five SNPs have previously been identified as schizophrenia vulnerability markers or implicate cognitive processes relevant to schizophrenia. This ability to replicate previous work suggests that our method has potential for detecting a meaningful epistatic relationship among the genes that influence brain abnormalities in schizophrenia.

Multicenter reliability of diffusion tensor imaging

A number of studies are now collecting diffusion tensor imaging (DTI) data across sites. While the reliability of anatomical images has been established by a number of groups, the reliability of DTI data has not been studied as extensively. In this study, five healthy controls were recruited and imaged at eight imaging centers. Repeated measures were obtained across two imaging protocols allowing intra-subject and inter-site variability to be assessed. Regional measures within white matter were obtained for standard rotationally invariant measures: fractional anisotropy, mean diffusivity, radial diffusivity, and axial diffusivity. Intra-subject coefficient of variation (CV) was typically <1% for all scalars and regions. Inter-site CV increased to ~1%-3%. Inter-vendor variation was similar to inter-site variability. This variability includes differences in the actual implementation of the sequence.

Progressive brain change in schizophrenia: a prospective longitudinal study of first-episode schizophrenia

BACKGROUND

Schizophrenia has a characteristic onset during adolescence or young adulthood but also tends to persist throughout life. Structural magnetic resonance studies indicate that brain abnormalities are present at onset, but longitudinal studies to assess neuroprogression have been limited by small samples and short or infrequent follow-up intervals.

METHODS

The Iowa Longitudinal Study is a prospective study of 542 first-episode patients who have been followed up to 18 years. In this report, we focus on those patients (n = 202) and control subjects (n = 125) for whom we have adequate structural magnetic resonance data (n = 952 scans) to provide a relatively definitive determination of whether progressive brain change occurs over a time interval of up to 15 years after intake.

RESULTS

A repeated-measures analysis showed significant age-by-group interaction main effects that represent a significant decrease in multiple gray matter regions (total cerebral, frontal, thalamus), multiple white matter regions (total cerebral, frontal, temporal, parietal), and a corresponding increase in cerebrospinal fluid (lateral ventricles and frontal, temporal, and parietal sulci). These changes were most severe during the early years after onset. They occur at severe levels only in a subset of patients. They are correlated with cognitive impairment but only weakly with other clinical measures.

CONCLUSIONS

Progressive brain change occurs in schizophrenia, affects both gray matter and white matter, is most severe during the early stages of the illness, and occurs only in a subset of patients. Measuring severity of progressive brain change offers a promising new avenue for phenotype definition in genetic studies of schizophrenia.

Neural network approaches to grade adult depression

Depression is a common but worrying psychological disorder that adversely affects one's quality of life. It is more ominous to note that its incidence is increasing. On the other hand, screening and grading of depression is still a manual and time consuming process that might be biased. In addition, grades of depression are often determined in continuous ranges, e.g., 'mild to moderate' and 'moderate to severe' instead of making them more discrete as 'mild', 'moderate', and 'severe'. Grading as a continuous range is confusing to the doctors and thus affecting the management plan at large. Given this practical issue, the present paper attempts to differentiate depression grades more accurately using two neural net learning approaches-'supervised', i.e., classification with Back propagation neural network (BPNN) and Adaptive Network-based Fuzzy Inference System (ANFIS) classifiers, and 'unsupervised', i.e., 'clustering' technique with Self-organizing map (SOM), built in MATLAB 7. The reason for using the supervised and unsupervised learning approaches is that, supervised learning depends exclusively on domain knowledge. Supervision may induce biasness and subjectivities related to the decision-making. Finally, the performance of BPNN and ANFIS are compared and discussed. It was observed that ANFIS, being a hybrid system, performed much better compared to the BPNN classifier. On the other hand, SOM-based clustering technique is also found useful in constructing three distinct clusters. It also assists visualizing the multidimensional data clusters into 2-D.

The nature and extent of cerebellar atrophy in chronic temporal lobe epilepsy

PURPOSE

Research indicates that patients with chronic temporal lobe epilepsy (TLE) exhibit cerebellar atrophy compared to healthy controls, but the degree to which specific regions of the cerebellum are affected remains unclear. The purpose of this study was to characterize the extent and lateralization of atrophy in individual cerebellar lobes and subregions in unilateral TLE using advanced quantitative magnetic resonance imaging (MRI) techniques.

METHODS

Study participants were 46 persons with TLE and 31 age- and gender- matched healthy controls. All participants underwent high-resolution MRI with manual tracing of the cerebellum yielding gray and white matter volumes of the right and left anterior lobes, superior posterior lobes, inferior posterior lobes, and corpus medullare. The degree to which asymmetric versus generalized abnormalities was evident in unilateral chronic TLE was determined and related to selected clinical seizure features (age of onset, duration of disorder).

KEY FINDINGS

There were no lateralized abnormalities in cerebellar gray matter or white matter in patients with right or left TLE (all p's > 0.2). Compared with controls, unilateral TLE was associated with significant bilateral reductions in the superior (p = 0.032) and inferior (p = 0.023) posterior lobes, whereas volume was significantly increased in the anterior lobes (p = 0.002), especially in patients with early onset TLE, and not significantly different in the corpus medullare (p = 0.71). Total superior cerebellar tissue volumes were reduced in association with increasing duration of epilepsy.

SIGNIFICANCE

Patients with unilateral TLE exhibit a pattern of bilateral cerebellar pathology characterized by atrophy of the superior and inferior posterior lobes, hypertrophy of the anterior lobe, and no effect on the corpus medullare. Cross-sectional analyses show that specific aspects of cerebellar pathology are associated with neurodevelopmental (anterior lobe) or chronicity-related (superior posterior lobe) features of the disorder.

Brain development in children with new onset epilepsy: a prospective controlled cohort investigation

PURPOSE

To characterize prospective neurodevelopmental changes in brain structure in children with new and recent-onset epilepsy compared to healthy controls.

METHODS

Thirty-four healthy controls (mean age 12.9 years) and 38 children with new/recent-onset idiopathic epilepsy (mean age 12.9 years) underwent 1.5 T magnetic resonance imaging (MRI) at baseline and 2 years later. Prospective changes in total cerebral and lobar gray and white matter volumes were compared within and between groups.

RESULTS

Prospective changes in gray matter volume were comparable for the epilepsy and control groups, with significant (p < 0.0001) reduction in total cerebral gray matter, due primarily to significant (p < 0.001) reductions in frontal and parietal gray matter. Prospective white matter volume changes differed between groups. Controls exhibited a significant (p = 0.0012) increase in total cerebral white matter volume due to significant (p < 0.001) volume increases in the frontal, parietal, and temporal lobes. In contrast, the epilepsy group exhibited nonsignificant white matter volume change in the total cerebrum (p = 0.51) as well as across all lobes (all p's > 0.06). The group by white matter volume change interactions were significant for total cerebrum (p = 0.04) and frontal lobe (p = 0.04).

DISCUSSION

Children with new and recent-onset epilepsy exhibit an altered pattern of brain development characterized by delayed age-appropriate increase in white matter volume. These findings may affect cognitive development through reduced brain connectivity and may also be related to the impairments in executive function commonly reported in this population.

Fully automated analysis using BRAINS: AutoWorkup

The BRAINS (Brain Research: Analysis of Images, Networks, and Systems) image analysis software has been in use, and in constant development, for over 20 years. The original neuroimage analysis pipeline using BRAINS was designed as a semiautomated procedure to measure volumes of the cerebral lobes and subcortical structures, requiring manual intervention at several stages in the process. Through use of advanced image processing algorithms the need for manual intervention at stages of image realignment, tissue sampling, and mask editing have been eliminated. In addition, inhomogeneity correction, intensity normalization, and mask cleaning routines have been added to improve the accuracy and consistency of the results. The fully automated method, AutoWorkup, is shown in this study to be more reliable (ICC ≥ 0.96, Jaccard index ≥ 0.80, and Dice index ≥ 0.89 for all tissues in all regions) than the average of 18 manual raters. On a set of 1130 good quality scans, the failure rate for correct realignment was 1.1%, and manual editing of the brain mask was required on 4% of the scans. In other tests, AutoWorkup is shown to produce measures that are reliable for data acquired across scanners, scanner vendors, and across sequences. Application of AutoWorkup for the analysis of data from the 32-site, multivendor PREDICT-HD study yield estimates of reliability to be greater than or equal to 0.90 for all tissues and regions.

Three-dimensional morphometric analysis of brain shape in nonsyndromic orofacial clefting

Previous studies report structural brain differences in individuals with nonsyndromic orofacial clefts (NSOFC) compared with healthy controls. These changes involve non-uniform shifts in tissue volume within the cerebral cortex and cerebellum, suggesting that the shape of the brain may be altered in cleft-affected individuals. To test this hypothesis, a landmark-based morphometric approach was utilized to quantify and compare brain shape in a sample of 31 adult males with cleft lip with or without cleft palate (CL/P), 14 adult males with cleft palate only (CPO) and 41 matched healthy controls. Fifteen midline and surface landmarks were collected from MRI brain scans and the resulting 3D coordinates were subjected to statistical shape analysis. First, a geometric morphometric analysis was performed in three steps: Procrustes superimposition of raw landmark coordinates, omnibus testing for group difference in shape, followed by canonical variates analysis (CVA) of shape coordinates. Secondly, Euclidean distance matrix analysis (EDMA) was carried out on scaled inter-landmark distances to identify localized shape differences throughout the brain. The geometric morphometric analysis revealed significant differences in brain shape among all three groups (P < 0.001). From CVA, the major brain shape changes associated with clefting included selective enlargement of the anterior cerebrum coupled with a relative reduction in posterior and/or inferior cerebral portions, changes in the medio-lateral position of the cerebral poles, posterior displacement of the corpus callosum, and reorientation of the cerebellum. EDMA revealed largely similar brain shape changes. Thus, compared with controls, major brain shape differences were present in adult males with CL/P and CPO. These results both confirm and expand previous findings from traditional volumetric studies of the brain in clefting and provide further evidence that the neuroanatomical phenotype in individuals with NSOFC is a primary manifestation of the defect and not a secondarily acquired characteristic.

Insular cortex morphometry in first-episode schizophrenia-spectrum patients: Diagnostic specificity and clinical correlations

Evidence so far indicates that the consistent association between insular cortex abnormalities and schizophrenia is already present at early phases of the illness. In the present investigation we aimed to study the specificity of insular structural abnormalities in schizophrenia by using region-of-interest morphometry to assess insular cortex morphological characteristics in the same heterogeneous sample of schizophrenia-spectrum patients. The 225 subjects, comprising 82 schizophrenia patients, 36 schizophreniform disorder patients and 24 patients with nonschizophrenic non-affective psychoses, and 83 healthy individuals were investigated. Magnetic resonance imaging brain scans (1.5T) were obtained and images analysed to evaluate insular cortex morphometric variables. The main resulting measurements were for insular gray matter volume and cortical surface area. The contribution of sociodemographic and clinical characteristics was controlled. Patients with schizophrenia-spectrum disorders did not significantly differ from controls in the insular cortex morphometric variables evaluated (all P's>0.11). Clinical variables were not significantly related with insular morphological changes. Noteworthy is the fact that none of the group morphological measurements varied significantly by gender or hemisphere. Neither did we find significant differences when patients with schizophrenia and with other non-affective psychoses were compared. Contrary to our initial hypotheses, we were unable to demonstrate significant morphometric anomalies in a large and heterogeneous sample of patients with a first-episode of schizophrenia-spectrum disorders.

Extrahippocampal integrity in temporal lobe epilepsy and cognition: thalamus and executive functioning

Chronic temporal lobe epilepsy (TLE) is characterized by the presence of extra-hippocampal brain abnormality and cognitive impairment in both memory and nonmemory domains. However, the link between structural integrity and cognition has not frequently been studied. Forty-six patients with TLE and 61 age-matched controls were studied to determine the predictive relationship between baseline thalamic volume and performance on measures of executive functioning evaluated 4 years later. As expected, the TLE group had lower baseline thalamic volumes than controls and also performed more poorly on measures of executive functioning. Total thalamic volume significantly predicted subsequent performance on all three measures of executive functioning. These findings were maintained when both hippocampal volume and frontal lobe volume were taken into account. These findings add to a growing literature demonstrating a link between extra-hippocampal volume abnormalities and cognitive functioning in TLE.

Toward the development of virtual surgical tools to aid orthopaedic FE analyses

Computational models of joint anatomy and function provide a means for biomechanists, physicians, and physical therapists to understand the effects of repetitive motion, acute injury, and degenerative diseases. Finite element models, for example, may be used to predict the outcome of a surgical intervention or to improve the design of prosthetic implants. Countless models have been developed over the years to address a myriad of orthopaedic procedures. Unfortunately, few studies have incorporated patient-specific models. Historically, baseline anatomic models have been used due to the demands associated with model development. Moreover, surgical simulations impose additional modeling challenges. Current meshing practices do not readily accommodate the inclusion of implants. Our goal is to develop a suite of tools (virtual instruments and guides) which enable surgical procedures to be readily simulated and to facilitate the development of all-hexahedral finite element mesh definitions.

Hippocampal volume deficits and shape deformities in young biological relatives of schizophrenia probands

Hippocampal volume decrement may be one of the changes that most closely pre-date schizophrenia onset. Studying hippocampal developmental morphology in adolescent or young adult biological relatives of schizophrenia probands has the potential to further our understanding of the neurodevelopmental etiology of schizophrenia and to discover biomarkers that may aid its early identification. We utilized an artificial neural network segmentation algorithm to automatically define and reliably measure MRI hippocampus volumes. We compared 46 young, nonpsychotic biological relatives of probands against 46 healthy controls without family history of schizophrenia and 46 schizophrenia probands (age range=13 to 28 years). We further contrasted hippocampal shape differences using spherical harmonic functions and assessed how obstetric complications (a trigger for aberrant in utero neurodevelopment) may contribute to hippocampal abnormalities. Similar to schizophrenia probands, unaffected biological relatives of probands had significantly smaller hippocampus volumes than controls; which correspond to inward displacements in shape deformities principally in the anterior hippocampal subregions. Examination of hippocampus volume-age relationships indicate that hippocampus volume normally decreases with age during late adolescence through early adulthood. In contrast, relatives of probands did not show these age-expected changes. Deviant hippocampus volume-age relationships suggest aberrant hippocampal neurodevelopment among biological relatives. Relatives with a history of obstetric complications had significantly smaller left and right hippocampi than relatives without obstetrics complications, including a dose relationship such that greater number of birth complications correlated with smaller hippocampus. Similar hippocampal volume deficits-obstetric complications relationships were observed among schizophrenia probands. Hippocampal abnormalities in schizophrenia are likely to be mediated by different neurobiological mechanisms, including factors associated with obstetric complications which occur during early neurodevelopment. Other brain maturational anomalies affecting the hippocampus in schizophrenia may manifest closer to illness onset in adolescence/early adulthood.

Automatic segmentation of brain structures using geometric moment invariants and artificial neural networks

We propose an automatic method for the segmentation of the brain structures in three dimensional (3D) Magnetic Resonance Images (MRI). The proposed method consists of two stages. In the first stage, we represent the shape of the structure using Geometric Moment Invariants (GMIs) in 8 scales. For each scale, an Artificial Neural Network (ANN) is designed to approximate the signed distance function of a desired structure. The GMIs along with the voxel intensities and coordinates are used as the input features of the ANN and the signed distance function as its output. In the second stage, we combine the outputs of the ANNs of the first stage and design another ANN to classify the image voxels into two classes, inside or outside of the structure. We introduce a fast method for moment calculations. The proposed method is applied to the segmentation of caudate, putamen, and thalamus in MRI where it has outperformed other methods in the literature.

A preliminary longitudinal magnetic resonance imaging study of brain volume and cortical thickness in autism

BACKGROUND

Autism is a developmental neurobiologic disorder associated with structural and functional abnormalities in several brain regions including the cerebral cortex. This longitudinal study examined developmental changes in brain volume and cortical thickness (CT) using magnetic resonance imaging (MRI) in children with autism.

METHODS

MRI scans and behavioral measures were obtained at baseline and after a 30-month interval in a sample of male subjects with autism (n = 18) and healthy age-, and sex-matched control subjects (n = 16) between ages 8 and 12 years at baseline.

RESULTS

No differences in brain volumes were observed between the autism and control subjects at baseline or follow-up. However, differences in total gray matter volumes were observed over time with significantly greater decreases in the autism group compared with control subjects. Differences in CT were observed over time with greater decreases in the autism group compared with control subjects in several brain regions including the frontal lobe. When accounting for multiple comparisons, differences between the two groups became nonsignificant except for changes in occipital CT. Furthermore, associations were observed between several clinical features and changes in CT with greater thinning of the cortex being correlated with more severe symptomatology.

CONCLUSIONS

Findings from this study provide preliminary evidence for age-related changes in gray matter volume and CT in children with autism that are associated with symptoms severity. Future longitudinal studies of larger sample sizes are needed to evaluate developmental changes and examine the relationships between structural abnormalities and clinical expressions of the disorder.

The relationship between symptom severity and regional cortical and grey matter volumes in schizophrenia

OBJECTIVE

To investigate the relationship between symptom severity and cortical and grey matter volumes in schizophrenia.

METHOD

Fifty-three outpatients with schizophrenia were assessed by the Scale for the Assessment of Negative Symptoms and the Scale for the Assessment of Positive Symptoms. Symptoms were grouped into five factors (negative, relational, inattention, disorganization, and reality distortion). Cortical and lobar grey matter volumes within all regions of the brain were obtained from magnetic resonance images using two independent software tools. The relationships between brain volumes and symptom factors were analyzed by partial correlations controlling for age, gender, dose and type of antipsychotic medication, and intracranial volume.

RESULTS

Negative symptoms were generally associated with larger cortical volumes in all regions of the brain, and the relational and inattention factors were associated with larger frontal grey matter volumes. The reality distortion factor was associated with smaller cortical volumes throughout the brain and with smaller frontal and temporal grey matter volumes.

CONCLUSION

Differential contribution of positive and negative symptoms to variation in cortical and grey matter volumes indicates separate neurobiological mechanisms underlying the two major symptom domains in schizophrenia.