Automatic measurement of changes in brain volume on consecutive 3D MR images by segmentation propagation

Venglustat combined with imiglucerase for neurological disease in adults with Gaucher disease type 3: the LEAP trial

Gaucher disease type 3 is a chronic neuronopathic disorder with wide-ranging effects, including hepatosplenomegaly, anaemia, thrombocytopenia, skeletal disease and diverse neurological manifestations. Biallelic mutations in GBA1 reduce lysosomal acid β-glucosidase activity, and its substrates, glucosylceramide and glucosylsphingosine, accumulate. Enzyme replacement therapy and substrate reduction therapy ameliorate systemic features of Gaucher disease, but no therapies are approved for neurological manifestations. Venglustat is an investigational, brain-penetrant, glucosylceramide synthase inhibitor with potential to improve the disease by rebalancing influx of glucosylceramide with impaired lysosomal recycling. The Phase 2, open-label LEAP trial (NCT02843035) evaluated orally administered venglustat 15 mg once-daily in combination with maintenance dose of imiglucerase enzyme replacement therapy during 1 year of treatment in 11 adults with Gaucher disease type 3. Primary endpoints were venglustat safety and tolerability and change in concentration of glucosylceramide and glucosylsphingosine in CSF from baseline to Weeks 26 and 52. Secondary endpoints included change in plasma concentrations of glucosylceramide and glucosylsphingosine, venglustat pharmacokinetics in plasma and CSF, neurologic function, infiltrative lung disease and systemic disease parameters. Exploratory endpoints included changes in brain volume assessed with volumetric MRI using tensor-based morphometry, and resting functional MRI analysis of regional brain activity and connectivity between resting state networks. Mean (SD) plasma venglustat AUC0-24 on Day 1 was 851 (282) ng•h/ml; Cmax of 58.1 (26.4) ng/ml was achieved at a median tmax 2.00 h. After once-daily venglustat, plasma concentrations (4 h post-dose) were higher compared with Day 1, indicating ∼2-fold accumulation. One participant (Patient 9) had low-to-undetectable venglustat exposure at Weeks 26 and 52. Based on mean plasma and CSF venglustat concentrations (excluding Patient 9), steady state appeared to be reached on or before Week 4. Mean (SD) venglustat concentration at Week 52 was 114 (65.8) ng/ml in plasma and 6.14 (3.44) ng/ml in CSF. After 1 year of treatment, median (inter-quartile range) glucosylceramide decreased 78% (72, 84) in plasma and 81% (77, 83) in CSF; median (inter-quartile range) glucosylsphingosine decreased 56% (41, 60) in plasma and 70% (46, 76) in CSF. Ataxia improved slightly in nine patients: mean (SD, range) total modified Scale for Assessment and Rating of Ataxia score decreased from 2.68 [1.54 (0.0 to 5.5)] at baseline to 1.55 [1.88 (0.0 to 5.0)] at Week 52 [mean change: -1.14 (95% CI: -2.06 to -0.21)]. Whole brain volume increased slightly in patients with venglustat exposure and biomarker reduction in CSF (306.7 ± 4253.3 mm3) and declined markedly in Patient 9 (-13894.8 mm3). Functional MRI indicated stronger connectivity at Weeks 26 and 52 relative to baseline between a broadly distributed set of brain regions in patients with venglustat exposure and biomarker reduction but not Patient 9, although neurocognition, assessed by Vineland II, deteriorated in all domains over time, which illustrates disease progression despite the intervention. There were no deaths, serious adverse events or discontinuations. In adults with Gaucher disease type 3 receiving imiglucerase, addition of once-daily venglustat showed acceptable safety and tolerability and preliminary evidence of clinical stability with intriguing but intrinsically inconsistent signals in selected biomarkers, which need to be validated and confirmed in future research.

Magnetic resonance imaging measures of brain volumes across the EXPEDITION trials in mild and moderate Alzheimer's disease dementia

Introduction

Solanezumab is a monoclonal antibody that preferentially binds soluble amyloid beta and promotes its clearance from the brain. The aim of this post hoc analysis was to assess the effect of low-dose solanezumab (400 mg) on global brain volume measures in patients with mild or moderate Alzheimer's disease (AD) dementia quantified using volumetric magnetic resonance imaging (vMRI) data from the EXPEDITION clinical trial program.

Methods

Patients with mild or moderate AD (EXPEDITION and EXPEDITION2) and mild AD (EXPEDITION3), were treated with either placebo or solanezumab (400 mg) every 4 weeks (Q4W) for 76 weeks. vMRI scans were acquired at baseline and at 80 weeks from 427 MRI facilities using a standardized imaging protocol. Whole brain volume (WBV) and ventricle volume (VV) changes were estimated at 80 weeks using either boundary shift integral (EXPEDITION and EXPEDITION2) or tensor-based morphometry (EXPEDITION3).

Results

The pooled cohort used for this study consisted of participants with vMRI at baseline and week 80 across the three trials. Analyzed patient subgroups comprised full patient cohort (N = 2933), apolipoprotein E (APOE) ε4+ carriers (N = 1835), and patients with mild (N = 2497) or moderate AD dementia (N = 428). No significant effect (all P-values ≥.05) of treatment was observed in the pooled sample, individual trials, or subgroups of patients with mild or moderate AD or APOE ε4 carriers, in either WBV or VV change.

Discussion

Analysis of patients with mild or moderate AD dementia from baseline to 80 weeks using vMRI measures of WBV and VV changes suggested that low-dose solanezumab was not linked to changes in volumes at 80 weeks. Analysis of the pooled cohort did not demonstrate an effect on brain volumes with treatment. Evaluation of a higher dose of solanezumab in the preclinical stage of AD is currently being undertaken.

Study the Longitudinal in vivo and Cross-Sectional ex vivo Brain Volume Difference for Disease Progression and Treatment Effect on Mouse Model of Tauopathy Using Automated MRI Structural Parcellation

Brain volume measurements extracted from structural MRI data sets are a widely accepted neuroimaging biomarker to study mouse models of neurodegeneration. Whether to acquire and analyze data in vivo or ex vivo is a crucial decision during the phase of experimental designs, as well as data analysis. In this work, we extracted the brain structures for both longitudinal in vivo and single-time-point ex vivo MRI acquired from the same animals using accurate automatic multi-atlas structural parcellation, and compared the corresponding statistical and classification analysis. We found that most gray matter structures volumes decrease from in vivo to ex vivo, while most white matter structures volume increase. The level of structural volume change also varies between different genetic strains and treatment. In addition, we showed superior statistical and classification power of ex vivo data compared to the in vivo data, even after resampled to the same level of resolution. We further demonstrated that the classification power of the in vivo data can be improved by incorporating longitudinal information, which is not possible for ex vivo data. In conclusion, this paper demonstrates the tissue-specific changes, as well as the difference in statistical and classification power, between the volumetric analysis based on the in vivo and ex vivo structural MRI data. Our results emphasize the importance of longitudinal analysis for in vivo data analysis.

New method for estimating the volume and volume fractions of the nasal structures in the goose (Anser anser domesticus) using computed tomography images

1. The conchae within the nasal cavity of poultry are important for water and energy conservation, but have not been experimentally evaluated. The aim of the present study was to determine the accuracy of volume and volume fraction estimates of the conchae, nasal septum and nasal cavity. 2. The nasal cavities of 7 adult goose heads were scanned using computed tomography (CT), with images sampled randomly at a 1/5 sampling fraction. Physical sections were obtained from the same samples, using an electric saw that had an adjustable section range, and provided 14 to 15 sections with a thickness of 2.5 mm. The section surface areas of the nasal cavity, nasal septum and conchae were estimated using the Cavalieri principle. Results obtained using the CT and physical section images were compared. Volumes and volume fractions obtained from the physical sections were accepted as the gold standard and differences in the CT images were determined. 3. Multiplication of the data obtained on the CT images with the deviation percentage of the physical sections produced normalised values. No differences were observed between the gold standard data and the CT images. While it was possible to normalise the obtained data using the gold standard values, the raw data could also be used for comparative studies because the deviations from normal would be similar for all groups. 4. Our study showed that the nasal structures could be estimated in vivo using CT images.

Automatic structural parcellation of mouse brain MRI using multi-atlas label fusion

Multi-atlas segmentation propagation has evolved quickly in recent years, becoming a state-of-the-art methodology for automatic parcellation of structural images. However, few studies have applied these methods to preclinical research. In this study, we present a fully automatic framework for mouse brain MRI structural parcellation using multi-atlas segmentation propagation. The framework adopts the similarity and truth estimation for propagated segmentations (STEPS) algorithm, which utilises a locally normalised cross correlation similarity metric for atlas selection and an extended simultaneous truth and performance level estimation (STAPLE) framework for multi-label fusion. The segmentation accuracy of the multi-atlas framework was evaluated using publicly available mouse brain atlas databases with pre-segmented manually labelled anatomical structures as the gold standard, and optimised parameters were obtained for the STEPS algorithm in the label fusion to achieve the best segmentation accuracy. We showed that our multi-atlas framework resulted in significantly higher segmentation accuracy compared to single-atlas based segmentation, as well as to the original STAPLE framework.

Reliability of longitudinal brain volume loss measurements between 2 sites in patients with multiple sclerosis: comparison of 7 quantification techniques

BACKGROUND AND PURPOSE

Brain volume loss is currently a MR imaging marker of neurodegeneration in MS. Available quantification algorithms perform either direct (segmentation-based techniques) or indirect (registration-based techniques) measurements. Because there is no reference standard technique, the assessment of their accuracy and reliability remains a difficult goal. Therefore, the purpose of this work was to assess the robustness of 7 different postprocessing algorithms applied to images acquired from different MR imaging systems.

MATERIALS AND METHODS

Nine patients with MS were followed longitudinally over 1 year (3 time points) on two 1.5T MR imaging systems. Brain volume change measures were assessed using 7 segmentation algorithms: a segmentation-classification algorithm, FreeSurfer, BBSI, KN-BSI, SIENA, SIENAX, and JI algorithm.

RESULTS

Intersite variability showed that segmentation-based techniques and SIENAX provided large and heterogeneous values of brain volume changes. A Bland-Altman analysis showed a mean difference of 1.8%, 0.07%, and 0.79% between the 2 sites, and a wide length agreement interval of 11.66%, 7.92%, and 11.94% for the segmentation-classification algorithm, FreeSurfer, and SIENAX, respectively. In contrast, registration-based algorithms showed better reproducibility, with a low mean difference of 0.45% for BBSI, KN-BSI and JI, and a mean length agreement interval of 1.55%. If SIENA obtained a lower mean difference of 0.12%, its agreement interval of 3.29% was wider.

CONCLUSIONS

If brain atrophy estimation remains an open issue, future investigations of the accuracy and reliability of the brain volume quantification algorithms are needed to measure the slow and small brain volume changes occurring in MS.

Development of identification of the central sulcus in brain magnetic resonance imaging

Magnetic resonance imaging (MRI) is useful in the quantitative evaluation of brain atrophy, because the superior contrast resolution facilitates separation of the gray and white matter. Quantitative assessment of brain atrophy has mainly been performed by manual measurement, which requires considerable time and effort to determine the brain volume. Therefore, computer-aided quantitative measurement methods for the diagnosis of brain atrophy are required. We have developed a method of segmenting the cerebrum, cerebellum-brainstem, and temporal lobe simultaneously on MR images obtained in a single sequence. It is important to measure the volume of not only these regions but also the frontal lobe in clinical use. However, for segmenting the frontal lobe, it is necessary to identify the Sylvian fissure and the central sulcus, which represent boundaries. Here, we developed a method of identifying the central sulcus from MR images obtained with a 1.5 T MRI scanner. The brain and the cerebrospinal fluid (CSF) regions were segmented using semiautomated segmentation method on MR images. The central sulcus shows an oblique line from the inside to the outside on the convexity view. The almost straight appearance of the central sulcus was used for segmentation of the central sulcus from the segmented CSF images. The central sulcus was identified with this technique in 77% of the images obtained by all sequences. This technique for identifying the central sulcus is very important not only for volumetry, but also for clinical diagnosis.

Magnetic resonance virtual histology for embryos: 3D atlases for automated high-throughput phenotyping

Ambitious international efforts are underway to produce gene-knockout mice for each of the 25,000 mouse genes, providing a new platform to study mammalian development and disease. Robust, large-scale methods for morphological assessment of prenatal mice will be essential to this work. Embryo phenotyping currently relies on histological techniques but these are not well suited to large volume screening. The qualitative nature of these approaches also limits the potential for detailed group analysis. Advances in non-invasive imaging techniques such as magnetic resonance imaging (MRI) may surmount these barriers. We present a high-throughput approach to generate detailed virtual histology of the whole embryo, combined with the novel use of a whole-embryo atlas for automated phenotypic assessment. Using individual 3D embryo MRI histology, we identified new pituitary phenotypes in Hesx1 mutant mice. Subsequently, we used advanced computational techniques to produce a whole-body embryo atlas from 6 CD-1 embryos, creating an average image with greatly enhanced anatomical detail, particularly in CNS structures. This methodology enabled unsupervised assessment of morphological differences between CD-1 embryos and Chd7 knockout mice (n=5 Chd7(+/+) and n=8 Chd7(+/-), C57BL/6 background). Using a new atlas generated from these three groups, quantitative organ volumes were automatically measured. We demonstrated a difference in mean brain volumes between Chd7(+/+) and Chd7(+/-) mice (42.0 vs. 39.1mm(3), p<0.05). Differences in whole-body, olfactory and normalised pituitary gland volumes were also found between CD-1 and Chd7(+/+) mice (C57BL/6 background). Our work demonstrates the feasibility of combining high-throughput embryo MRI with automated analysis techniques to distinguish novel mouse phenotypes.

Rigid model-based 3D segmentation of the bones of joints in MR and CT images for motion analysis

There are several medical application areas that require the segmentation and separation of the component bones of joints in a sequence of images of the joint acquired under various loading conditions, our own target area being joint motion analysis. This is a challenging problem due to the proximity of bones at the joint, partial volume effects, and other imaging modality-specific factors that confound boundary contrast. In this article, a two-step model-based segmentation strategy is proposed that utilizes the unique context of the current application wherein the shape of each individual bone is preserved in all scans of a particular joint while the spatial arrangement of the bones alters significantly among bones and scans. In the first step, a rigid deterministic model of the bone is generated from a segmentation of the bone in the image corresponding to one position of the joint by using the live wire method. Subsequently, in other images of the same joint, this model is used to search for the same bone by minimizing an energy function that utilizes both boundary- and region-based information. An evaluation of the method by utilizing a total of 60 data sets on MR and CT images of the ankle complex and cervical spine indicates that the segmentations agree very closely with the live wire segmentations, yielding true positive and false positive volume fractions in the range 89%-97% and 0.2%-0.7%. The method requires 1-2 minutes of operator time and 6-7 min of computer time per data set, which makes it significantly more efficient than live wire-the method currently available for the task that can be used routinely.

A review of geometric transformations for nonrigid body registration

This paper provides a comprehensive and quantitative review of spatial transformations models for nonrigid image registration. It explains the theoretical foundation of the models and classifies them according to this basis. This results in two categories, physically based models described by partial differential equations of continuum mechanics (e.g., linear elasticity and fluid flow) and basis function expansions derived from interpolation and approximation theory (e.g., radial basis functions, B-splines and wavelets). Recent work on constraining the transformation so that it preserves the topology or is diffeomorphic is also described. The final section reviews some recent evaluation studies. The paper concludes by explaining under what conditions a particular transformation model is appropriate.

Hierarchical linear modeling (HLM) of longitudinal brain structural and cognitive changes in alcohol-dependent individuals during sobriety

BACKGROUND

Hierarchical linear modeling (HLM) can reveal complex relationships between longitudinal outcome measures and their covariates under proper consideration of potentially unequal error variances. We demonstrate the application of HLM to the study of magnetic resonance imaging (MRI)-derived brain volume changes and cognitive changes in abstinent alcohol-dependent individuals as a function of smoking status, smoking severity, and drinking quantities.

METHODS

Twenty non-smoking recovering alcoholics (nsALC) and 30 age-matched smoking recovering alcoholics (sALC) underwent quantitative MRI and cognitive assessments at 1 week, 1 month, and 7 months of sobriety. Eight non-smoking light drinking controls were studied at baseline and 7 months later. Brain and ventricle volumes at each time point were quantified using MRI masks, while the boundary shift integral method measured volume changes between time points. Using HLM, we modeled volumetric and cognitive outcome measures as a function of cigarette and alcohol use variables.

RESULTS

Different hierarchical linear models with unique model structures are presented and discussed. The results show that smaller brain volumes at baseline predict faster brain volume gains, which were also related to greater smoking and drinking severities. Over 7 months of abstinence from alcohol, sALC compared to nsALC showed less improvements in visuospatial learning and memory despite larger brain volume gains and ventricular shrinkage.

CONCLUSIONS

Different and unique hierarchical linear models allow assessments of the complex relationships among outcome measures of longitudinal data sets. These HLM applications suggest that chronic cigarette smoking modulates the temporal dynamics of brain structural and cognitive changes in alcoholics during prolonged sobriety.

Automatic quantification of changes in bone in serial MR images of joints

Recent innovations in drug therapies have made it highly desirable to obtain sensitive biomarkers of disease progression that can be used to quantify the performance of candidate disease modifying drugs. In order to measure potential image-based biomarkers of disease progression in an experimental model of rheumatoid arthritis (RA), we present two different methods to automatically quantify changes in a bone in in-vivo serial magnetic resonance (MR) images from the model. Both methods are based on rigid and nonrigid image registration to perform the analysis. The first method uses segmentation propagation to delineate a bone from the serial MR images giving a global measure of temporal changes in bone volume. The second method uses rigid body registration to determine intensity change within a bone, and then maps these into a reference coordinate system using nonrigid registration. This gives a local measure of temporal changes in bone lesion volume. We detected significant temporal changes in local bone lesion volume in five out of eight identified candidate bone lesion regions, and significant difference in local bone lesion volume between male and female subjects in three out of eight candidate bone lesion regions. But the global bone volume was found to be fluctuating over time. Finally, we compare our findings with histology of the subjects and the manual segmentation of bone lesions.

Magnetic resonance imaging measures of brain atrophy in multiple sclerosis

Magnetic resonance imaging (MRI) has been widely used to diagnose and monitor multiple sclerosis (MS). Although MRI-visible lesions are a key feature of MS, they are thought to correlate poorly with clinical progression. Neurodegeneration is increasingly being recognized as an important factor in the pathogenesis of MS, and MRI measures of brain atrophy have been suggested as surrogate markers of neuroaxonal loss and disease progression. This pathology may be more relevant to the progression of disability than focal inflammation. A number of MRI-based methods have been developed for the measurement of global and regional brain atrophy. Natural-history studies of MS and clinically isolated syndromes suggestive of MS have observed atrophy in these subjects above that seen in controls, over periods ranging from three months to years. Brain atrophy has also been incorporated as an outcome measure in therapeutic trials of disease-modifying treatments. This paper considers neuroaxonal loss and the pathological basis of brain atrophy, methods developed to quantify brain atrophy, the findings of natural-history and therapeutic studies, the relationship of brain atrophy to disability and cognition, and the future research directions and clinical applications of brain atrophy measurements.

The application of serial MRI analysis techniques to the study of cerebral atrophy in late-onset dementia

We have used a serial MR image analysis technique previously developed for studies of cerebral atrophy in early-onset dementia and applied it to a study of late-onset dementia patients with images acquired using a different scanner and scan sequence. Validation and optimisation tests showed that with only small changes to key analysis parameters the technique can successfully be applied to previously untested data with dissimilar image characteristics. The overall accuracy in estimation of cerebral atrophy using the technique was determined to be between 2 and 4 ml (1sigma) depending on the conditions during image acquisition. By comparing the results of alternative registration techniques we demonstrate the potential of using of fully automated 9 DOF image registration as an effective and efficient means of correcting for scanner pixel size variations, even in the presence of significant cerebral atrophy. Applied to the late-onset dementia study, patients were found to have significantly increased mean atrophy rates (p<0.001) compared to controls. In general the analysis technique is shown to be a robust, accurate and transferable tool of potential value for future studies of dementia and related neuro-degenerative disorders.

Temporal dynamics of brain anatomy

The brain changes profoundly in structure and function during development and as a result of diseases such as the dementias, schizophrenia, multiple sclerosis, and tumor growth. Strategies to measure, map, and visualize these brain changes are of immense value in basic and clinical neuroscience. Algorithms that map brain change with sufficient spatial and temporal sensitivity can also assess drugs that aim to decelerate or arrest these changes. In neuroscience studies, these tools can reveal subtle brain changes in adolescence and old age and link these changes with measurable differences in brain function and cognition. Early detection of brain change in patients at risk for dementia; tumor recurrence; or relapsing-remitting conditions, such as multiple sclerosis, is also vital for optimizing therapy. We review a variety of mathematical and computational approaches to detect structural brain change with unprecedented sensitivity, both spatially and temporally. The resulting four-dimensional (4-D) maps of brain anatomy are warehoused in population-based brain atlases. Here, statistical tools compare brain changes across subjects and across populations, adjusting for complex differences in brain structure. Brain changes in an individual can be compared with a normative database comprised of subjects matched for age, gender, and other demographic factors. These dynamic brain maps offer key biological markers for understanding disease progression and testing therapeutic response. The early detection of disease-related brain changes is also critical for possible pre-emptive intervention before the ravages of disease have set in.

Automatic segmentation of the brain and intracranial cerebrospinal fluid in T1-weighted volume MRI scans of the head, and its application to serial cerebral and intracranial volumetry

A new fully automatic algorithm for the segmentation of the brain and total intracranial cerebrospinal fluid (CSF) from T(1)-weighted volume MRI scans of the head, called Exbrain v.2, is described. The algorithm was developed in the context of serial intracranial volumetry. A brain mask obtained using a previous version of the algorithm forms the basis of the CSF segmentation. Improved brain segmentation is then obtained by iterative tracking of the brain-CSF interface. Gray matter (GM), white matter (WM), and intracranial CSF volumes and probability maps are calculated based on a model of intensity probability distribution (IPD) that includes two partial volume classes: GM-CSF and GM-WM. Accuracy was assessed using the Montreal Neurological Institute's (MNI) digital phantom scan. Reproducibility was assessed using scan pairs from 24 controls and 10 patients with epilepsy. Segmentation overlap with the gold standard was 98% for the brain and 95%, 96%, and 97% for the GM, WM, and total intracranial contents, respectively; CSF overlap was 86%. In the controls, the Bland and Altman coefficient of reliability (CR) was 35.2 cm(3) for the total brain volume (TBV) and 29.0 cm(3) for the intracranial volume (ICV). Scan-matching reduced CR to 25.2 cm(3) and 17.1 cm(3) for the TBV and ICV, respectively. For the patients, similar CR values were obtained for the ICV.

Cerebral atrophy and disability in relapsing-remitting and secondary progressive multiple sclerosis over four years

Pathology and magnetic resonance imaging (MRI) studies have provided evidence of widespread axonal loss and reductions of cerebral and spinal cord volume in multiple sclerosis (MS). Atrophy measures on MRI may be a useful surrogate marker of worsening disability in MS, but the published studies are of relatively short duration. Change in brain volume (atrophy) was measured over a four-year period in 20 patients with relapsing-remitting (RR) and 18 with secondary progressive (SP) MS using three-dimensional (3D) MRI acquired during treatment trials of interferon-beta-1a (Rebif). Brain parenchymal and lateral ventricle volume changes were determined and correlated with clinical measures. Over four years, brain parenchymal volume (BPV) decreased in RRMS and SPMS patients by 0.9% (P = 0.006) and 0.3% (P = 0.118), respectively, and the lateral ventricle volumes increased by 15% (P < 0.0001) and 13% (P < 0.0001), respectively. In RRMS patients both lateral ventricle volume (r = 0.63, P = 0.004) and BPV change (r = -0.47, P = 0.037) were related to disability change, as measured by the Expanded Disability Status Scale. Even though a small study and despite the possible confounding effects of interferon treatment, this study demonstrated an association between measures of cerebral atrophy and worsening disability. The data also provides evidence that brain atrophy can be detected early in the disease course and central white matter atrophy as reflected by ventricle enlargement appears to be a continuous process.

Segmentation of MR images for computer-assisted surgery of the lumbar spine

This paper describes a segmentation algorithm designed to separate bone from soft tissue in magnetic resonance (MR) images developed for computer-assisted surgery of the spine. The algorithm was applied to MR images of the spine of healthy volunteers. Registration experiments were carried out on a physical model of a spine generated from computed tomography (CT) data of a surgical patient. Segmented CT, manually segmented MR and MR images segmented using the developed algorithm were compared. The algorithm performed well at segmenting bone from soft tissue on images taken of healthy volunteers. Registration experiments showed similar results between the CT and MR data. The MR data, which were manually segmented, performed worse on visual verification experiments than both the CT and semi-automatic segmented data. The algorithm developed performs well at segmenting bone from soft tissue in MR images of the spine as measured using registration experiments.

Quantification of small cerebral ventricular volume changes in treated growth hormone patients using nonrigid registration

Nonrigid registration can automatically quantify small changes in volume of anatomical structures over time by means of segmentation propagation. Here, we use a nonrigid registration algorithm based on optimising normalized mutual information to quantify small changes in brain ventricle volume in magnetic resonance (MR) images of a group of five patients treated with growth hormone replacement therapy and a control group of six volunteers. The lateral ventricles are segmented from each subject image by registering with the brainweb image which has this structure delineated. The mean (standard deviation) volume change measurements are 1.09 (0.73) cm3 for the patient group and 0.08 (0.62) cm3 for the volunteer group; this difference is statistically significant at the 1% level. We validate our volume measurements by determining the precision from three consecutive scans of five volunteers and also comparing the measurements to previously published volume change estimates obtained by visual inspection of difference images. Results demonstrate a precision of sigma < or = 0.52 cm3 (n = 5) and a rank correlation coefficient with assessed difference images of p = 0.7 (n = 11). To determine the level of shape correspondence we manually segmented subject's ventricles and compared them to the propagations using a voxel overlap similarity index, this gave a mean similarity index of 0.81 (n = 7).

Automatic volumetric measurement of lateral ventricles on magnetic resonance images with correction of partial volume effects

PURPOSE

To propose a method for the quantification of lateral ventricle (LV) volumes on a single sequence of 3D magnetic resonance (MR) images.

MATERIALS AND METHODS

This algorithm, following a preliminary fuzzy tissue classification step, is based on the development of mathematical morphology processes allowing both the extraction of the LVs and the correction of partial volume effects on their boundaries. The procedure is fast and totally unsupervised. The method is tested on a phantom image, then applied to five patients diagnosed as potentially suffering from Alzheimer's disease, and finally applied on several MR acquisitions to show the genericness of the algorithm.

RESULTS AND CONCLUSION

This technique yielded both an accurate estimation of ventricular volumes intra- and intersubject with respect to published data and a relevant management of partial volume effects. Numerous clinical applications are now expected, from the study of schizophrenia to the longitudinal follow-up of Alzheimer's patients.

Automated hippocampal segmentation by regional fluid registration of serial MRI: validation and application in Alzheimer's disease

The application of voxel-level three-dimensional registration to serial magnetic resonance imaging (MRI) is described. This fluid registration determines deformation fields modeling brain change, which are consistent with a model describing a viscous fluid. The objective was to validate the measurement of hippocampal volumetric change by fluid registration in Alzheimer's disease (AD) against current methodologies. The hippocampus was chosen for this study because it is difficult to measure reproducibly by manual segmentation and is widely studied; however, the technique is applicable to any structure which can be delineated on a scan. First, suitable values for the viscosity-body-force-ratio, alpha (0.01), and the number of iterations (300), were established and the convergence, repeatability, linearity, and accuracy investigated and compared with expert manual segmentation. A simple model of hippocampal atrophy was used to compare simulated volumetric change against that obtained by fluid registration. Finally the serial segmentation was compared with the current gold standard technique-expert human labeling with a volume repeatability of approximately 4%-in 27 subjects (15 normal controls, 12 clinically diagnosed with Alzheimer's disease). The scan-rescan volumetric consistency of serial segmentation by fluid-registration was shown to be superior to human serial segmentors ( approximately 2%). The mean absolute volume difference between fluid and manual segmentation was 0.7%. Fluid registration has potential importance for tracking longitudinal structural changes in brain particularly in the context of the clinical trial where large numbers of subjects may have multiple MR scans.

The benefit of stereology for quantitative radiology

A knowledge of stereology (i.e. proper sampling), the opportunities provided by computers for image analysis (i.e. image segmentation, image registration, data base exploration, 3D reconstruction), and the strengths (i.e. non-invasive) and limitations (i.e. finite resolution, image artefacts) of medical imaging equipment must all be combined for reliable quantitative magnetic resonance imaging (MRI), the goal of which is to obtain a deeper understanding of the structure, function, life cycle and evolution of the human body, especially the brain, and a more objective diagnosis of disease and assessment of its response to treatment. In this article we illustrate the first of these requirements. We describe the application of proper sampling strategies and efficient computer-based counting procedures for obtaining unbiased estimates of volume by the Cavalieri method and of surface area from vertical sections. In particular, we estimate the volume of a brain tumour from Cavalieri sections, the volume of grey matter in the cerebral hemispheres from Cavalieri slices and the surface area of the cerebral cortex from vertical sections. The estimates obtained are mathematically unbiased. In each case, we assess the precision of the estimates empirically. Application of formulae available for predicting the precision of volume estimates obtained using the Cavalieri sections and slices methods is also described.