Validation analysis of the novel imaging-based prognostic radiomic signature in patients undergoing primary surgery for advanced high-grade serous ovarian cancer (HGSOC)

BACKGROUND

Predictive models based on radiomics features are novel, highly promising approaches for gynaecological oncology. Here, we wish to assess the prognostic value of the newly discovered Radiomic Prognostic Vector (RPV) in an independent cohort of high-grade serous ovarian cancer (HGSOC) patients, treated within a Centre of Excellence, thus avoiding any bias in treatment quality.

METHODS

RPV was calculated using standardised algorithms following segmentation of routine preoperative imaging of patients (n = 323) who underwent upfront debulking surgery (01/2011-07/2018). RPV was correlated with operability, survival and adjusted for well-established prognostic factors (age, postoperative residual disease, stage), and compared to previous validation models.

RESULTS

The distribution of low, medium and high RPV scores was 54.2% (n = 175), 33.4% (n = 108) and 12.4% (n = 40) across the cohort, respectively. High RPV scores independently associated with significantly worse progression-free survival (PFS) (HR = 1.69; 95% CI:1.06-2.71; P = 0.038), even after adjusting for stage, age, performance status and residual disease. Moreover, lower RPV was significantly associated with total macroscopic tumour clearance (OR = 2.02; 95% CI:1.56-2.62; P = 0.00647).

CONCLUSIONS

RPV was validated to independently identify those HGSOC patients who will not be operated tumour-free in an optimal setting, and those who will relapse early despite complete tumour clearance upfront. Further prospective, multicentre trials with a translational aspect are warranted for the incorporation of this radiomics approach into clinical routine.

Monitoring Response to Transarterial Chemoembolization in Hepatocellular Carcinoma Using 18F-Fluorothymidine PET

Accurate disease monitoring is essential after transarterial chemoembolization (TACE) in hepatocellular carcinoma (HCC) because of the potential for profound adverse events and large variations in survival outcome. Posttreatment changes on conventional imaging can confound determination of residual or recurrent disease, magnifying the clinical challenge. On the basis of increased expression of thymidylate synthase (TYMS), thymidine kinase 1 (TK-1), and equilibrative nucleoside transporter 1 (SLC29A1) in HCC compared with liver tissue, we conducted a proof-of-concept study evaluating the efficacy of 3'-deoxy-3'-18F-fluorothymidine (18F-FLT) PET to assess response to TACE. Because previous PET studies in HCC have been hampered by high background liver signal, we investigated whether a temporal-intensity voxel clustering (kinetic spatial filtering, or KSF) improved lesion detection. Methods: A tissue microarray was built from 36 HCC samples and from matching surrounding cirrhotic tissue and was stained for TK-1 A prospective study was conducted; 18 patients with a diagnosis of HCC by the criteria of the American Association for the Study of Liver Diseases who were eligible for treatment with TACE were enrolled. The patients underwent baseline conventional imaging and dynamic 18F-FLT PET with KSF followed by TACE. Imaging was repeated 6-8 wk after TACE. The PET parameters were compared with modified enhancement-based RECIST. Results: Cancer Genome Atlas analysis revealed increased RNA expression of TYMS, TK-1, and SLC29A1 in HCC. TK-1 protein expression was significantly higher in HCC (P < 0.05). The sensitivity of 18F-FLT PET for baseline HCC detection was 73% (SUVmax, 9.7 ± 3.0; tumor to liver ratio, 1.2 ± 0.3). Application of KSF did not improve lesion detection. Lesion response after TACE by modified RECIST was 58% (14 patients with 24 lesions). A 30% reduction in mean 18F-FLT PET uptake was observed after TACE, correlating with an observed PET response of 60% (15/25). A significant and profound reduction in the radiotracer delivery parameter K1 after TACE was observed. Conclusion:18F-FLT PET can differentiate HCC from surrounding cirrhotic tissue, with PET parameters correlating with TACE response. KSF did not improve visualization of tumor lesions. These findings warrant further investigation.

Neuroanatomical and Microglial Alterations in the Striatum of Levodopa-Treated, Dyskinetic Hemi-Parkinsonian Rats

Dyskinesia associated with chronic levodopa treatment in Parkinson’s disease is associated with maladaptive striatal plasticity. The objective of this study was to examine whether macroscale structural changes, as captured by magnetic resonance imaging (MRI) accompany this plasticity and to identify plausible cellular contributors in a rodent model of levodopa-induced dyskinesia. Adult male Sprague-Dawley rats were rendered hemi-parkinsonian by stereotaxic injection of 6-hydroxydopamine into the left medial forebrain bundle prior to chronic treatment with saline (control) or levodopa to induce abnormal involuntary movements (AIMs), reflective of dyskinesia. Perfusion-fixed brains underwent ex vivo structural MRI before sectioning and staining for cellular markers. Chronic treatment with levodopa induced significant AIMs (p < 0.0001 versus saline). The absolute volume of the ipsilateral, lesioned striatum was increased in levodopa-treated rats resulting in a significant difference in percentage volume change when compared to saline-treated rats (p < 0.01). Moreover, a significant positive correlation was found between this volume change and AIMs scores for individual levodopa-treated rats (r = 0.96; p < 0.01). The density of Iba1+ cells was increased within the lesioned versus intact striatum (p < 0.01) with no difference between treatment groups. Conversely, Iba1+ microglia soma size was significantly increased (p < 0.01) in the lesioned striatum of levodopa-treated but not saline-treated rats. Soma size was not, however, significantly correlated with either AIMs or MRI volume change. Although GFAP+ astrocytes were elevated in the lesioned versus intact striatum (p < 0.001), there was no difference between treatment groups. No statistically significant effects of either lesion or treatment on RECA1, a marker for blood vessels, were observed. Collectively, these data suggest chronic levodopa treatment in 6-hydroxydopamine lesioned rats is associated with increased striatal volume that correlates with the development of AIMs. The accompanying increase in number and size of microglia, however, cannot alone explain this volume expansion. Further multi-modal studies are warranted to establish the brain-wide effects of chronic levodopa treatment.

Test-retest reliability and longitudinal analysis of automated hippocampal subregion volumes in healthy ageing and Alzheimer's disease populations

The hippocampal formation is a complex brain structure that is important in cognitive processes such as memory, mood, reward processing and other executive functions. Histological and neuroimaging studies have implicated the hippocampal region in neuropsychiatric disorders as well as in neurodegenerative diseases. This highly plastic limbic region is made up of several subregions that are believed to have different functional roles. Therefore, there is a growing interest in imaging the subregions of the hippocampal formation rather than modelling the hippocampus as a homogenous structure, driving the development of new automated analysis tools. Consequently, there is a pressing need to understand the stability of the measures derived from these new techniques. In this study, an automated hippocampal subregion segmentation pipeline, released as a developmental version of Freesurfer (v6.0), was applied to T1-weighted magnetic resonance imaging (MRI) scans of 22 healthy older participants, scanned on 3 separate occasions and a separate longitudinal dataset of 40 Alzheimer's disease (AD) patients. Test-retest reliability of hippocampal subregion volumes was assessed using the intra-class correlation coefficient (ICC), percentage volume difference and percentage volume overlap (Dice). Sensitivity of the regional estimates to longitudinal change was estimated using linear mixed effects (LME) modelling. The results show that out of the 24 hippocampal subregions, 20 had ICC scores of 0.9 or higher in both samples; these regions include the molecular layer, granule cell layer of the dentate gyrus, CA1, CA3 and the subiculum (ICC > 0.9), whilst the hippocampal fissure and fimbria had lower ICC scores (0.73-0.88). Furthermore, LME analysis of the independent AD dataset demonstrated sensitivity to group and individual differences in the rate of volume change over time in several hippocampal subregions (CA1, molecular layer, CA3, hippocampal tail, fissure and presubiculum). These results indicate that this automated segmentation method provides a robust method with which to measure hippocampal subregions, and may be useful in tracking disease progression and measuring the effects of pharmacological intervention.

Evolution of structural abnormalities in the rat brain following in utero exposure to maternal immune activation: A longitudinal in vivo MRI study

Genetic and environmental risk factors for psychiatric disorders are suggested to disrupt the trajectory of brain maturation during adolescence, leading to the development of psychopathology in adulthood. Rodent models are powerful tools to dissect the specific effects of such risk factors on brain maturational profiles, particularly when combined with Magnetic Resonance Imaging (MRI; clinically comparable technology). We therefore investigated the effect of maternal immune activation (MIA), an epidemiological risk factor for adult-onset psychiatric disorders, on rat brain maturation using atlas and tensor-based morphometry analysis of longitudinal in vivo MR images. Exposure to MIA resulted in decreases in the volume of several cortical regions, the hippocampus, amygdala, striatum, nucleus accumbens and unexpectedly, the lateral ventricles, relative to controls. In contrast, the volumes of the thalamus, ventral mesencephalon, brain stem and major white matter tracts were larger, relative to controls. These volumetric changes were maximal between post-natal day 50 and 100 with no differences between the groups thereafter. These data are consistent with and extend prior studies of brain structure in MIA-exposed rodents. Apart from the ventricular findings, these data have robust face validity to clinical imaging findings reported in studies of individuals at high clinical risk for a psychiatric disorder. Further work is now required to address the relationship of these MRI changes to behavioral dysfunction and to establish thier cellular correlates.

Correlations of Behavioral Deficits with Brain Pathology Assessed through Longitudinal MRI and Histopathology in the HdhQ150/Q150 Mouse Model of Huntington's Disease

A variety of mouse models have been developed that express mutant huntingtin (mHTT) leading to aggregates and inclusions that model the molecular pathology observed in Huntington's disease. Here we show that although homozygous HdhQ150 knock-in mice developed motor impairments (rotarod, locomotor activity, grip strength) by 36 weeks of age, cognitive dysfunction (swimming T maze, fear conditioning, odor discrimination, social interaction) was not evident by 94 weeks. Concomitant to behavioral assessments, T2-weighted MRI volume measurements indicated a slower striatal growth with a significant difference between wild type (WT) and HdhQ150 mice being present even at 15 weeks. Indeed, MRI indicated significant volumetric changes prior to the emergence of the "clinical horizon" of motor impairments at 36 weeks of age. A striatal decrease of 27% was observed over 94 weeks with cortex (12%) and hippocampus (21%) also indicating significant atrophy. A hypothesis-free analysis using tensor-based morphometry highlighted further regions undergoing atrophy by contrasting brain growth and regional neurodegeneration. Histology revealed the widespread presence of mHTT aggregates and cellular inclusions. However, there was little evidence of correlations between these outcome measures, potentially indicating that other factors are important in the causal cascade linking the molecular pathology to the emergence of behavioral impairments. In conclusion, the HdhQ150 mouse model replicates many aspects of the human condition, including an extended pre-manifest period prior to the emergence of motor impairments.

Diamagnetic chemical exchange saturation transfer (diaCEST) affords magnetic resonance imaging of extracellular matrix hydrogel implantation in a rat model of stroke

Extracellular matrix (ECM) is widely used as an inductive biological scaffold to repair soft tissue after injury by promoting functional site-appropriate remodeling of the implanted material. However, there is a lack of non-invasive analysis methods to monitor the remodeling characteristics of the ECM material after implantation and its biodegradation over time. We describe the use of diamagnetic chemical exchange saturation transfer (CEST) magnetic resonance imaging to monitor the distribution of an ECM hydrogel after intracerebral implantation into a stroke cavity. In vitro imaging indicated a robust concentration-dependent detection of the ECM precursor and hydrogel at 1.8 and 3.6 ppm, which broadly corresponded to chondroitin sulfate and fibronectin. This detection was robust to changes in pH and improved at 37 °C. In vivo implantation of ECM hydrogel into the stroke cavity in a rat model corresponded macroscopically to the distribution of biomaterial as indicated by histology, but mismatches were also evident. Indeed, CEST imaging detected an endogenous "increased deposition". To account for this endogenous activity, pre-implantation images were subtracted from post-implantation images to yield a selective visualization of hydrogel distribution in the stroke cavity and its evolution over 7 days. The CEST detection of ECM returned to baseline within 3 days due to a decrease in fibronectin and chondroitin sulfate in the hydrogel. The distribution of ECM hydrogel within the stroke cavity is hence feasible in vivo, but further advances are required to warrant a selective long-term monitoring in the context of biodegradation.

Dissociation between iron accumulation and ferritin upregulation in the aged substantia nigra: attenuation by dietary restriction

Despite regulation, brain iron increases with aging and may enhance aging processes including neuroinflammation. Increases in magnetic resonance imaging transverse relaxation rates, R2 and R2*, in the brain have been observed during aging. We show R2 and R2* correlate well with iron content via direct correlation to semi-quantitative synchrotron-based X-ray fluorescence iron mapping, with age-associated R2 and R2* increases reflecting iron accumulation. Iron accumulation was concomitant with increased ferritin immunoreactivity in basal ganglia regions except in the substantia nigra (SN). The unexpected dissociation of iron accumulation from ferritin-upregulation in the SN suggests iron dyshomeostasis in the SN. Occurring alongside microgliosis and astrogliosis, iron dyshomeotasis may contribute to the particular vulnerability of the SN. Dietary restriction (DR) has long been touted to ameliorate brain aging and we show DR attenuated age-related in vivo R2 increases in the SN over ages 7 - 19 months, concomitant with normal iron-induction of ferritin expression and decreased microgliosis. Iron is known to induce microgliosis and conversely, microgliosis can induce iron accumulation, which of these may be the initial pathological aging event warrants further investigation. We suggest iron chelation therapies and anti-inflammatory treatments may be putative 'anti-brain aging' therapies and combining these strategies may be synergistic.

Neurorestoration induced by the HDAC inhibitor sodium valproate in the lactacystin model of Parkinson's is associated with histone acetylation and up-regulation of neurotrophic factors

BACKGROUND AND PURPOSE

Histone hypoacetylation is associated with Parkinson's disease (PD), due possibly to an imbalance in the activities of enzymes responsible for histone (de)acetylation; correction of which may be neuroprotective/neurorestorative. This hypothesis was tested using the anti-epileptic drug sodium valproate, a known histone deacetylase inhibitor (HDACI), utilizing a delayed-start study design in the lactacystin rat model of PD.

EXPERIMENTAL APPROACH

The irreversible proteasome inhibitor lactacystin was unilaterally injected into the substantia nigra of Sprague-Dawley rats that subsequently received valproate for 28 days starting 7 days after lactacystin lesioning. Longitudinal motor behavioural testing, structural MRI and post-mortem assessment of nigrostriatal integrity were used to track changes in this model of PD and quantify neuroprotection/restoration. Subsequent cellular and molecular analyses were performed to elucidate the mechanisms underlying valproate's effects.

KEY RESULTS

Despite producing a distinct pattern of structural re-modelling in the healthy and lactacystin-lesioned brain, delayed-start valproate administration induced dose-dependent neuroprotection/restoration against lactacystin neurotoxicity, characterized by motor deficit alleviation, attenuation of morphological brain changes and restoration of dopaminergic neurons in the substantia nigra. Molecular analyses revealed that valproate alleviated lactacystin-induced histone hypoacetylation and induced up-regulation of brain neurotrophic/neuroprotective factors.

CONCLUSIONS AND IMPLICATIONS

The histone acetylation and up-regulation of neurotrophic/neuroprotective factors associated with valproate treatment culminate in a neuroprotective and neurorestorative phenotype in this animal model of PD. As valproate induced structural re-modelling of the brain, further research is required to determine whether valproate represents a viable candidate for disease treatment; however, the results suggest that HDACIs could hold potential as disease-modifying agents in PD.

Bringing memory fMRI to the clinic: comparison of seven memory fMRI protocols in temporal lobe epilepsy

fMRI is increasingly implemented in the clinic to assess memory function. There are multiple approaches to memory fMRI, but limited data on advantages and reliability of different methods. Here, we compared effect size, activation lateralisation, and between‐sessions reliability of seven memory fMRI protocols: Hometown Walking (block design), Scene encoding (block design and event‐related design), Picture encoding (block and event‐related), and Word encoding (block and event‐related). All protocols were performed on three occasions in 16 patients with temporal lobe epilepsy (TLE). Group T‐maps showed activity bilaterally in medial temporal lobe for all protocols. Using ANOVA, there was an interaction between hemisphere and seizure‐onset lateralisation (P = 0.009) and between hemisphere, protocol and seizure‐onset lateralisation (P = 0.002), showing that the distribution of memory‐related activity between left and right temporal lobes differed between protocols and between patients with left‐onset and right‐onset seizures. Using voxelwise intraclass Correlation Coefficient, between‐sessions reliability was best for Hometown and Scenes (block and event). The between‐sessions spatial overlap of activated voxels was also greatest for Hometown and Scenes. Lateralisation of activity between hemispheres was most reliable for Scenes (block and event) and Words (event). Using receiver operating characteristic analysis to explore the ability of each fMRI protocol to classify patients as left‐onset or right‐onset TLE, only the Words (event) protocol achieved a significantly above‐chance classification of patients at all three sessions. We conclude that Words (event) protocol shows the best combination of between‐sessions reliability of the distribution of activity between hemispheres and reliable ability to distinguish between left‐onset and right‐onset patients. Hum Brain Mapp 36:1595–1608, 2015. © 2015 The Authors Human Brain Mapping Published by Wiley Periodicals, Inc.

Reduced cortical volume and elevated astrocyte density in rats chronically treated with antipsychotic drugs-linking magnetic resonance imaging findings to cellular pathology

BACKGROUND

Increasing evidence suggests that antipsychotic drugs (APD) might affect brain structure directly, particularly the cerebral cortex. However, the precise anatomical loci of these effects and their underlying cellular basis remain unclear.

METHODS

With ex vivo magnetic resonance imaging in rats treated chronically with APDs, we used automated analysis techniques to map the regions that show maximal impact of chronic (8 weeks) treatment with either haloperidol or olanzapine on the rat cortex. Guided by these imaging findings, we undertook a focused postmortem investigation with stereology.

RESULTS

We identified decreases in the volume and thickness of the anterior cingulate cortex (ACC) after chronic APD treatment, regardless of the APD administered. Postmortem analysis confirmed these volumetric findings and demonstrated that chronic APD treatment had no effect on the total number of neurons or S100β+ astrocytes in the ACC. In contrast, an increase in the density of these cells was observed.

CONCLUSIONS

This study demonstrates region-specific structural effects of chronic APD treatment on the rat cortex, primarily but not exclusively localized to the ACC. At least in the rat, these changes are not due to a loss of either neurons or astrocytes and are likely to reflect a loss of neuropil. Although caution needs to be exerted when extrapolating results from animals to patients, this study highlights the power of this approach to link magnetic resonance imaging findings to their histopathological origins.

Correlations of behavioral deficits with brain pathology assessed through longitudinal MRI and histopathology in the R6/1 mouse model of Huntington's disease

Huntington's disease (HD) is caused by the expansion of a CAG repeat in the huntingtin (HTT) gene. The R6 mouse models of HD express a mutant version of exon 1 HTT and typically develop motor and cognitive impairments, a widespread huntingtin (HTT) aggregate pathology and brain atrophy. Unlike the more commonly used R6/2 mouse line, R6/1 mice have fewer CAG repeats and, subsequently, a less rapid pathological decline. Compared to the R6/2 line, fewer descriptions of the progressive pathologies exhibited by R6/1 mice exist. The association between the molecular and cellular neuropathology with brain atrophy, and with the development of behavioral phenotypes remains poorly understood in many models of HD. In attempt to link these factors in the R6/1 mouse line, we have performed detailed assessments of behavior and of regional brain abnormalities determined through longitudinal, in vivo magnetic resonance imaging (MRI), as well as an end-stage, ex vivo MRI study and histological assessment. We found progressive decline in both motor and non-motor related behavioral tasks in R6/1 mice, first evident at 11 weeks of age. Regional brain volumes were generally unaffected at 9 weeks, but by 17 weeks there was significant grey matter atrophy. This age-related brain volume loss was validated using a more precise, semi-automated Tensor Based morphometry assessment. As well as these clear progressive phenotypes, mutant HTT (mHTT) protein, the hallmark of HD molecular pathology, was widely distributed throughout the R6/1 brain and was accompanied by neuronal loss. Despite these seemingly concomitant, robust pathological phenotypes, there appeared to be little correlation between the three main outcome measures: behavioral performance, MRI-detected brain atrophy and histopathology. In conclusion, R6/1 mice exhibit many features of HD, but the underlying mechanisms driving these clear behavioral disturbances and the brain volume loss, still remain unclear.

Hippocampal hypertrophy and sleep apnea: a role for the ischemic preconditioning?

The full impact of multisystem disease such as obstructive sleep apnoea (OSA) on regions of the central nervous system is debated, as the subsequent neurocognitive sequelae are unclear. Several preclinical studies suggest that its purported major culprits, intermittent hypoxia and sleep fragmentation, can differentially affect adult hippocampal neurogenesis. Although the prospective biphasic nature of chronic intermittent hypoxia in animal models of OSA has been acknowledged, so far the evidence for increased ‘compensatory’ neurogenesis in humans is uncertain. In a cross-sectional study of 32 patients with mixed severity OSA and 32 non-apnoeic matched controls inferential analysis showed bilateral enlargement of hippocampi in the OSA group. Conversely, a trend for smaller thalami in the OSA group was noted. Furthermore, aberrant connectivity between the hippocampus and the cerebellum in the OSA group was also suggested by the correlation analysis. The role for the ischemia/hypoxia preconditioning in the neuropathology of OSA is herein indicated, with possible further reaching clinical implications.

3D reconstruction of 2D fluorescence histology images and registration with in vivo MR images: application in a rodent stroke model

To validate and add value to non-invasive imaging techniques, the corresponding histology is required to establish biological correlates. We present an efficient, semi-automated image-processing pipeline that uses immunohistochemically stained sections to reconstruct a 3D brain volume from 2D histological images before registering these with the corresponding 3D in vivo magnetic resonance images (MRI). A multistep registration procedure that first aligns the "global" volume by using the centre of mass and then applies a rigid and affine alignment based on signal intensities is described. This technique was applied to a training set of three rat brain volumes before being validated on three normal brains. Application of the approach to register "abnormal" images from a rat model of stroke allowed the neurobiological correlates of the variations in the hyper-intense MRI signal intensity caused by infarction to be investigated. For evaluation, the corresponding anatomical landmarks in MR and histology were defined to measure the registration accuracy. A registration error of 0.249 mm (approximately one in-plane voxel dimension) was evident in healthy rat brains and of 0.323 mm in a rodent model of stroke. The proposed reconstruction and registration pipeline allowed for the precise analysis of non-invasive MRI and corresponding microstructural histological features in 3D. We were thus able to interrogate histology to deduce the cause of MRI signal variations in the lesion cavity and the peri-infarct area.

Registration of challenging pre-clinical brain images

The size and complexity of brain imaging studies in pre-clinical populations are increasing, and automated image analysis pipelines are urgently required. Pre-clinical populations can be subjected to controlled interventions (e.g., targeted lesions), which significantly change the appearance of the brain obtained by imaging. Existing systems for registration (the systematic alignment of scans into a consistent anatomical coordinate system), which assume image similarity to a reference scan, may fail when applied to these images. However, affine registration is a particularly vital pre-processing step for subsequent image analysis which is assumed to be an effective procedure in recent literature describing sophisticated techniques such as manifold learning. Therefore, in this paper, we present an affine registration solution that uses a graphical model of a population to decompose difficult pairwise registrations into a composition of steps using other members of the population. We developed this methodology in the context of a pre-clinical model of stroke in which large, variable hyper-intense lesions significantly impact registration performance. We tested this technique systematically in a simulated human population of brain tumour images before applying it to pre-clinical models of Parkinson's disease and stroke.

A comparison of automated anatomical-behavioural mapping methods in a rodent model of stroke

•

The first application of voxel lesion symptom mapping (VLSM) in rodents.

•

A comparison of VLSM with tensor based morphometry (TBM) methods in a stroke model.

•

Comparison of automated techniques with manual measurements and model power calculations.

•

Analysis of both local and non-local lesion effects.

•

Correlation of structural change with behaviour.

Neurological damage, due to conditions such as stroke, results in a complex pattern of structural changes and significant behavioural dysfunctions; the automated analysis of magnetic resonance imaging (MRI) and discovery of structural–behavioural correlates associated with these disorders remains challenging. Voxel lesion symptom mapping (VLSM) has been used to associate behaviour with lesion location in MRI, but this analysis requires the definition of lesion masks on each subject and does not exploit the rich structural information in the images. Tensor-based morphometry (TBM) has been used to perform voxel-wise structural analyses over the entire brain; however, a combination of lesion hyper-intensities and subtle structural remodelling away from the lesion might confound the interpretation of TBM. In this study, we compared and contrasted these techniques in a rodent model of stroke (n = 58) to assess the efficacy of these techniques in a challenging pre-clinical application. The results from the automated techniques were compared using manually derived region-of-interest measures of the lesion, cortex, striatum, ventricle and hippocampus, and considered against model power calculations. The automated TBM techniques successfully detect both lesion and non-lesion effects, consistent with manual measurements. These techniques do not require manual segmentation to the same extent as VLSM and should be considered part of the toolkit for the unbiased analysis of pre-clinical imaging-based studies.

White matter integrity assessed by diffusion tensor tractography in a patient with a large tumor mass but minimal clinical and neuropsychological deficits

Diffusion tensor imaging (DTI) tractography and image registration were used to investigate a patient with a massive left-sided brain tumor, whose size was largely disproportionate to his subtle neurological deficits. MRI was obtained from the patient and his healthy identical twin, who acted as anatomical reference for DTI and as a control for quantitative measures. To compensate for the patient's altered anatomy, seed and way points for probabilistic tractography were drawn on the color-coded direction maps of the healthy twin. Registration, based on the combination of b0-images, T2-weighted and T1-weighted images, was used to identify the corresponding regions in the patient. The corticospinal tract (CST), the superior longitudinal fasciculus (SLF), and the cingulum bundle (CB) showed displaced anatomy. A significant difference was found between fractional anisotropy distribution along the left SLF and CB, but not along the CST. These findings fit well with the patient's substantial preservation of his motor abilities, while abnormalities of the SLF and CB could explain the subtle but detectable cognitive deficits.

Contrasting effects of haloperidol and lithium on rodent brain structure: a magnetic resonance imaging study with postmortem confirmation

BACKGROUND

Magnetic resonance imaging (MRI) studies suggest that antipsychotic -treated patients with schizophrenia show a decrease in gray-matter volumes, whereas lithium-treated patients with bipolar disorder show marginal increases in gray-matter volumes. Although these clinical data are confounded by illness, chronicity, and other medications, they do suggest that typical antipsychotic drugs and lithium have contrasting effects on brain volume.

METHODS

Rodent models offer a tractable system to test this hypothesis, and we therefore examined the effect of chronic treatment (8 weeks) and subsequent withdrawal (8 weeks) with clinically relevant dosing of an antipsychotic (haloperidol, HAL) or lithium (Li) on brain volume using longitudinal in vivo structural MRI and confirmed the findings postmortem using unbiased stereology.

RESULTS

Chronic HAL treatment induced decreases in whole brain volume (-4%) and cortical gray matter (-6%), accompanied by hypertrophy of the corpus striatum (+14%). In contrast, chronic Li treatment induced increases in whole-brain volume (+5%) and cortical gray matter (+3%) without a significant effect on striatal volume. Following 8 weeks of drug withdrawal, HAL-induced changes in brain volumes normalized, whereas Li-treated animals retained significantly greater total brain volumes, as confirmed postmortem. However, the distribution of these contrasting changes was topographically distinct: with the haloperidol decreases more prominent rostral, the lithium increases were more prominent caudal.

CONCLUSIONS

The implications of these findings for the clinic, potential mitigation strategies, and further drug development are discussed.

A comprehensive testing protocol for MRI neuroanatomical segmentation techniques: Evaluation of a novel lateral ventricle segmentation method

Although a wide range of approaches have been developed to automatically assess the volume of brain regions from MRI, the reproducibility of these algorithms across different scanners and pulse sequences, their accuracy in different clinical populations and sensitivity to real changes in brain volume have not always been comprehensively examined. Firstly we present a comprehensive testing protocol which comprises 312 freely available MR images to assess the accuracy, reproducibility and sensitivity of automated brain segmentation techniques. Accuracy is assessed in infants, young adults and patients with Alzheimer's disease in comparison to gold standard measures by expert observers using a manual technique based on Cavalieri's principle. The protocol determines the reliability of segmentation between scanning sessions, different MRI pulse sequences and 1.5T and 3T field strengths and examines their sensitivity to small changes in volume using a large longitudinal dataset. Secondly we apply this testing protocol to a novel algorithm for segmenting the lateral ventricles and compare its performance to the widely used FSL FIRST and FreeSurfer methods. The testing protocol produced quantitative measures of accuracy, reliability and sensitivity of lateral ventricle volume estimates for each segmentation method. The novel algorithm showed high accuracy in all populations (intraclass correlation coefficient, ICC>0.95), good reproducibility between MRI pulse sequences (ICC>0.99) and was sensitive to age related changes in longitudinal data. FreeSurfer demonstrated high accuracy (ICC>0.95), good reproducibility (ICC>0.99) and sensitivity whilst FSL FIRST showed good accuracy in young adults and infants (ICC>0.90) and good reproducibility (ICC=0.98), but was unable to segment ventricular volume in patients with Alzheimer's disease or healthy subjects with large ventricles. Using the same computer system, the novel algorithm and FSL FIRST processed a single MRI image in less than 10min while FreeSurfer took approximately 7h. The testing protocol presented enables the accuracy, reproducibility and sensitivity of different algorithms to be compared. We also demonstrate that the novel segmentation algorithm and FreeSurfer are both effective in determining lateral ventricular volume and are well suited for multicentre and longitudinal MRI studies.

Evolution of extra-nigral damage predicts behavioural deficits in a rat proteasome inhibitor model of Parkinson's disease

Establishing the neurological basis of behavioural dysfunction is key to provide a better understanding of Parkinson's disease (PD) and facilitate development of effective novel therapies. For this, the relationships between longitudinal structural brain changes associated with motor behaviour were determined in a rat model of PD and validated by post-mortem immunohistochemistry. Rats bearing a nigrostriatal lesion induced by infusion of the proteasome inhibitor lactacystin into the left-medial forebrain bundle and saline-injected controls underwent magnetic resonance imaging (MRI) at baseline (prior to surgery) and 1, 3 and 5 weeks post-surgery with concomitant motor assessments consisting of forelimb grip strength, accelerating rotarod, and apormorphine-induced rotation. Lactacystin-injected rats developed early motor deficits alongside decreased ipsilateral cortical volumes, specifically thinning of the primary motor (M1) and somatosensory cortices and lateral ventricle hypertrophy (as determined by manual segmentation and deformation-based morphometry). Although sustained, motor dysfunction and nigrostriatal damage were maximal by 1 week post-surgery. Additional volume decreases in the ipsilateral ventral midbrain; corpus striatum and thalamus were only evident by week 3 and 5. Whilst cortical MRI volume changes best predicted the degree of motor impairment, post-mortem tyrosine hydroxylase immunoreactivity in the striatum was a better predictor of motor behaviour overall, with the notable exception of performance in the accelerating rotarod, in which, M1 cortical thickness remained the best predictor. These results highlight the importance of identifying extra-nigral regions of damage that impact on behavioural dysfunction from damage to the nigrostriatal system.

Segmentation of the thalamus in MRI based on T1 and T2

Reliable identification of thalamic nuclei is required to improve positioning of electrodes in Deep Brain Stimulation (DBS), and to allow the role of individual thalamic nuclei in health and disease to be fully investigated. In this work, a previously proposed method for identifying sub-regions within the thalamus based on differences in their T1 and T2 values is explored in detail. The effect on the segmentation of T1 and T2 dependence weighted against priors for spatial position and extent was investigated. When T1 and T2 dependence was highly weighted, good distinction between identified regions was obtained in T1/T2 feature-space, but no contiguous anatomically distinct regions were identified within the thalamus. Incorporating spatial priors was necessary to ensure anatomically distinct regions were defined. Optimal values for segmentation parameters were obtained by assessing performance on a 'synthetic thalamus'. Using these optimum input parameters, within- and between-subjects reproducibility was assessed. Good reproducibility was obtained when six regions were specified to be identified in the thalamus. The six regions identified were similar in the majority of the normal subject group. However, intriguingly these regions were different from those obtained in the same subjects using a well-known connectivity-based segmentation technique. This method shows promise to identify intrathalamic structures on the basis of T1 and T2 signal. A comprehensive characterisation of thalamic nuclei may require a fully multi-modal approach.

Magnetic resonance brain image processing and arithmetic with FSL

Medical imaging has been transformed by a move from qualitative to quantitative approaches where image processing is used to enhance visual information and image analysis is used to derive structural and functional measurements. The ideal quantitative analysis methods are automatic and require no user intervention, and so-called image analysis pipelines exist for some applications. However, in the majority of cases automatic methods seldom live up to their name, may fail when prior assumptions are not met, and may not exist at all for new applications. The identification and careful use of well-known image processing and analysis techniques is a vital part of imaging and invaluable when problems arise with automatic methods. Here a number of key image analysis tasks in brain imaging are presented with particular reference to the freely available FMRIB Software Library.

Reproducibility of thalamic segmentation based on probabilistic tractography

Reliable identification of thalamic nuclei is required to improve targeting of electrodes used in Deep Brain Stimulation (DBS), and for exploring the role of thalamus in health and disease. A previously described method using probabilistic tractography to segment the thalamus based on connections to cortical target regions was implemented. Both within- and between-subject reproducibility were quantitatively assessed by the overlap of the resulting segmentations; the effect of two different numbers of target regions (6 and 31) on reproducibility of the segmentation results was also investigated. Very high reproducibility was observed when a single dataset was processed multiple times using different starting conditions. Thalamic segmentation was also very reproducible when multiple datasets from the same subject were processed using six cortical target regions. Within-subject reproducibility was reduced when the number of target regions was increased, particularly in medial and posterior regions of the thalamus. A large degree of overlap in segmentation results from different subjects was obtained, particularly in thalamic regions classified as connecting to frontal, parietal, temporal and pre-central cortical target regions.

Mitochondrial displacements in response to nanomechanical forces

Mechanical stress affects and regulates many aspects of the cell, including morphology, growth, differentiation, gene expression and apoptosis. In this study we show how mechanical stress perturbs the intracellular structures of the cell and induces mechanical responses. In order to correlate mechanical perturbations to cellular responses, we used a combined fluorescence-atomic force microscope (AFM) to produce well defined nanomechanical perturbations of 10 nN while simultaneously tracking the real-time motion of fluorescently labelled mitochondria in live cells. The spatial displacement of the organelles in response to applied loads demonstrates the highly dynamic mechanical response of mitochondria in fibroblast cells. The average displacement of all mitochondrial structures analysed showed an increase of approximately 40%, post-perturbation ( approximately 160 nm in comparison to basal displacements of approximately 110 nm). These results show that local forces can produce organelle displacements at locations far from the initial point of contact (up to approximately 40 microm). In order to examine the role of the cytoskeleton in force transmission and its effect on mitochondrial displacements, both the actin and microtubule cytoskeleton were disrupted using Cytochalasin D and Nocodazole, respectively. Our results show that there is no significant change in mitochondrial displacement following indentation after such treatments. These results demonstrate the role of the cytoskeleton in force transmission through the cell and on mitochondrial displacements. In addition, it is suggested that care must be taken when performing mechanical experiments on living cells with the AFM, as these local mechanical perturbations may have significant structural and even biochemical effects on the cell.

Accuracy assessment of global and local atrophy measurement techniques with realistic simulated longitudinal Alzheimer's disease images

The evaluation of atrophy quantification methods based on magnetic resonance imaging have been usually hindered by the lack of realistic gold standard data against which to judge these methods or to help refine them. Recently [Camara, O., Schweiger, M., Scahill, R., Crum, W., Sneller, B., Schnabel, J., Ridgway, G., Cash, D., Hill, D., Fox, N., 2006. Phenomenological model of diffuse global and regional atrophy using finite-element methods. IEEE Trans. Med.l Imaging 25, 1417-1430], we presented a technique in which atrophy is realistically simulated in different tissue compartments or neuroanatomical structures with a phenomenological model. In this study, we have generated a cohort of realistic simulated Alzheimer's disease (AD) images with known amounts of atrophy, mimicking a set of 19 real controls and 27 probable AD subjects, with an improved version of our atrophy simulation methodology. This database was then used to assess the accuracy of several well-known computational anatomy methods which provide global (BSI and SIENA) or local (Jacobian integration) estimates of longitudinal atrophy in brain structures using MR images. SIENA and BSI results correlated very well with gold standard data (Pearson coefficient of 0.962 and 0.969 respectively), achieving small mean absolute differences with respect to the gold standard (percentage change from baseline volume): BSI of 0.23%+/-0.26%; SIENA of 0.22%+/-0.28%. Jacobian integration was guided by both fluid and FFD-based registration techniques and resulting deformation fields and associated Jacobians were compared, region by region, with gold standard ones. The FFD-based technique outperformed the fluid one in all evaluated structures (mean absolute differences from the gold standard in percentage change from baseline volume): whole brain, FFD=0.31%, fluid=0.58%; lateral ventricles, FFD=0.79%; fluid=1.45%; left hippocampus, FFD=0.82%; fluid=1.42%; right hippocampus, FFD=0.95%; fluid=1.62%. The largest errors for both local techniques occurred in the sulcal CSF (FFD=2.27%; fluid=3.55%) regions. For large structures such as the whole brain, these mean absolute differences, relative to the applied atrophy, represented similar percentages for the BSI, SIENA and FFD techniques (controls/patients): BSI, 51.99%/16.36%; SIENA, 62.34%/21.59%; FFD, 41.02%/24.95%. For small structures such as the hippocampi, these percentages were larger, especially for controls where errors were approximately equal to the small applied changes (controls/patients): FFD, 92.82%/43.61%. However, these apparently large relative errors have not prevented the global or hippocampal measures from finding significant group separation in our study. The evaluation framework presented here will help in quantifying whether the accuracy of future methodological developments is sufficient for analysing change in smaller or less atrophied local brain regions. Results obtained in our experiments with realistic simulated data confirm previously published estimates of accuracy for both evaluated global techniques. Regarding Jacobian Integration methods, the FFD-based one demonstrated promising results and potential for being used in clinical studies alongside (or in place of) the more common global methods. The generated gold standard data has also allowed us to identify some stages and sets of parameters in the evaluated techniques--the brain extraction step in the global techniques and the number of multi-resolution levels and the stopping criteria in the registration-based methods--that are critical for their accuracy.

A new validation method for X-ray mammogram registration algorithms using a projection model of breast X-ray compression

Establishing spatial correspondence between features visible in X-ray mammograms obtained at different times has great potential to aid assessment and quantitation of change in the breast indicative of malignancy. The literature contains numerous nonrigid registration algorithms developed for this purpose, but existing approaches are flawed by the assumption of inappropriate 2-D transformation models and quantitative estimation of registration accuracy is limited. In this paper, we describe a novel validation method which simulates plausible mammographic compressions of the breast using a magnetic resonance imaging (MRI) derived finite element model. By projecting the resulting known 3-D displacements into 2-D and generating pseudo-mammograms from these same compressed magnetic resonance (MR) volumes, we can generate convincing images with known 2-D displacements with which to validate a registration algorithm. We illustrate this approach by computing the accuracy for two conventional nonrigid 2-D registration algorithms applied to mammographic test images generated from three patient MR datasets. We show that the accuracy of these algorithms is close to the best achievable using a 2-D one-to-one correspondence model but that new algorithms incorporating more representative transformation models are required to achieve sufficiently accurate registrations for this application.

Correlations between apolipoprotein E epsilon4 gene dose and whole brain atrophy rates

OBJECTIVE

The purpose of this study was to characterize the relationship between whole brain atrophy rates and three levels of genetic risk for Alzheimer's disease in cognitively normal persons. The authors previously found accelerated whole brain atrophy rates in patients with probable Alzheimer's disease by computing changes in brain volume from sequential magnetic resonance images (MRIs).

METHODS

The authors assessed 36 late-middle-aged persons from three genetic groups: those with two, one, and no copies of the apolipoprotein E (APOE) epsilon4 allele, a common Alzheimer's disease susceptibility gene. The participants had clinical ratings, neuropsychological tests, and volumetric T1-weighted MRIs during a baseline visit and again approximately 2 years later. Two different image-analysis techniques, brain boundary shift integration and iterative principal component analysis, were used to compute whole brain atrophy rates.

RESULTS

While there were no baseline, follow-up, or between-visit differences in the clinical ratings or neuropsychological test scores among the three subject groups, whole brain atrophy rates were significantly greater in the epsilon4 homozygote group than in noncarriers and were significantly correlated with epsilon4 gene dose (i.e., the number of epsilon4 alleles in a person's APOE genotype).

CONCLUSION

Since APOE epsilon4 gene dose is associated with an increased risk of Alzheimer's disease and a younger median age at dementia onset, this study suggests an association between the risk of Alzheimer's disease and accelerated brain atrophy rates before the onset of cognitive impairment.

Phenomenological model of diffuse global and regional atrophy using finite-element methods

The main goal of this work is the generation of ground-truth data for the validation of atrophy measurement techniques, commonly used in the study of neurodegenerative diseases such as dementia. Several techniques have been used to measure atrophy in cross-sectional and longitudinal studies, but it is extremely difficult to compare their performance since they have been applied to different patient populations. Furthermore, assessment of performance based on phantom measurements or simple scaled images overestimates these techniques' ability to capture the complexity of neurodegeneration of the human brain. We propose a method for atrophy simulation in structural magnetic resonance (MR) images based on finite-element methods. The method produces cohorts of brain images with known change that is physically and clinically plausible, providing data for objective evaluation of atrophy measurement techniques. Atrophy is simulated in different tissue compartments or in different neuroanatomical structures with a phenomenological model. This model of diffuse global and regional atrophy is based on volumetric measurements such as the brain or the hippocampus, from patients with known disease and guided by clinical knowledge of the relative pathological involvement of regions and tissues. The consequent biomechanical readjustment of structures is modelled using conventional physics-based techniques based on biomechanical tissue properties and simulating plausible tissue deformations with finite-element methods. A thermoelastic model of tissue deformation is employed, controlling the rate of progression of atrophy by means of a set of thermal coefficients, each one corresponding to a different type of tissue. Tissue characterization is performed by means of the meshing of a labelled brain atlas, creating a reference volumetric mesh that will be introduced to a finite-element solver to create the simulated deformations. Preliminary work on the simulation of acquisition artefacts is also presented. Cross-sectional and longitudinal sets of simulated data are shown and a visual classification protocol has been used by experts to rate real and simulated scans according to their degree of atrophy. Results confirm the potential of the proposed methodology.

Generalized overlap measures for evaluation and validation in medical image analysis

Measures of overlap of labelled regions of images, such as the Dice and Tanimoto coefficients, have been extensively used to evaluate image registration and segmentation algorithms. Modern studies can include multiple labels defined on multiple images yet most evaluation schemes report one overlap per labelled region, simply averaged over multiple images. In this paper, common overlap measures are generalized to measure the total overlap of ensembles of labels defined on multiple test images and account for fractional labels using fuzzy set theory. This framework allows a single "figure-of-merit" to be reported which summarises the results of a complex experiment by image pair, by label or overall. A complementary measure of error, the overlap distance, is defined which captures the spatial extent of the nonoverlapping part and is related to the Hausdorff distance computed on grey level images. The generalized overlap measures are validated on synthetic images for which the overlap can be computed analytically and used as similarity measures in nonrigid registration of three-dimensional magnetic resonance imaging (MRI) brain images. Finally, a pragmatic segmentation ground truth is constructed by registering a magnetic resonance atlas brain to 20 individual scans, and used with the overlap measures to evaluate publicly available brain segmentation algorithms.

Generalised overlap measures for assessment of pairwise and groupwise image registration and segmentation

Effective validation techniques are an essential pre-requisite for segmentation and non-rigid registration techniques to enter clinical use. These algorithms can be evaluated by calculating the overlap of corresponding test and gold-standard regions. Common overlap measures compare pairs of binary labels but it is now common for multiple labels to exist and for fractional (partial volume) labels to be used to describe multiple tissue types contributing to a single voxel. Evaluation studies may involve multiple image pairs. In this paper we use results from fuzzy set theory and fuzzy morphology to extend the definitions of existing overlap measures to accommodate multiple fractional labels. Simple formulas are provided which define single figures of merit to quantify the total overlap for ensembles of pairwise or groupwise label comparisons. A quantitative link between overlap and registration error is established by defining the overlap tolerance. Experiments are performed on publicly available labeled brain data to demonstrate the new measures in a comparison of pairwise and groupwise registration.

Simulation of local and global atrophy in Alzheimer's disease studies

We propose a method for atrophy simulation in structural MR images based on finite-element methods, providing data for objective evaluation of atrophy measurement techniques. The modelling of diffuse global and regional atrophy is based on volumetric measurements from patients with known disease and guided by clinical knowledge of the relative pathological involvement of regions. The consequent biomechanical readjustment of structures is modelled using conventional physics-based techniques based on tissue properties and simulating plausible deformations with finite-element methods. Tissue characterization is performed by means of the meshing of a labelled brain atlas, creating a reference volumetric mesh, and a partial volume tissue model is used to reduce the impact of the mesh discretization. An example of simulated data is shown and a visual evaluation protocol used by experts has been developed to assess the degree of realism of the simulated images. First results demonstrate the potential of the proposed methodology.

Anisotropic multi-scale fluid registration: evaluation in magnetic resonance breast imaging

Registration using models of compressible viscous fluids has not found the general application of some other techniques (e.g., free-form-deformation (FFD)) despite its ability to model large diffeomorphic deformations. We report on a multi-resolution fluid registration algorithm which improves on previous work by (a) directly solving the Navier-Stokes equation at the resolution of the images, (b) accommodating image sampling anisotropy using semi-coarsening and implicit smoothing in a full multi-grid (FMG) solver and (c) exploiting the inherent multi-resolution nature of FMG to implement a multi-scale approach. Evaluation is on five magnetic resonance (MR) breast images subject to six biomechanical deformation fields over 11 multi-resolution schemes. Quantitative assessment is by tissue overlaps and target registration errors and by registering using the known correspondences rather than image features to validate the fluid model. Context is given by comparison with a validated FFD algorithm and by application to images of volunteers subjected to large applied deformation. The results show that fluid registration of 3D breast MR images to sub-voxel accuracy is possible in minutes on a 1.6 GHz Linux-based Athlon processor with coarse solutions obtainable in a few tens of seconds. Accuracy and computation time are comparable to FFD techniques validated for this application.

Non-rigid image registration: theory and practice

Image registration is an important enabling technology in medical image analysis. The current emphasis is on development and validation of application-specific non-rigid techniques, but there is already a plethora of techniques and terminology in use. In this paper we discuss the current state of the art of non-rigid registration to put on-going research in context and to highlight current and future clinical applications that might benefit from this technology. The philosophy and motivation underlying non-rigid registration is discussed and a guide to common terminology is presented. The core components of registration systems are described and outstanding issues of validity and validation are confronted.

An inverse problem approach to the estimation of volume change

We present a new technique for determining structure-by-structure volume changes, using an inverse problem approach. Given a pre-labelled brain and a series of images at different time-points, we generate finite element meshes from the image data, with volume change modelled by means of an unknown coefficient of expansion on a per-structure basis. We can then determine the volume change in each structure of interest using inverse problem optimization techniques. The proposed method has been tested with simulated and clinical data. Results suggest that the presented technique can be seen as an alternative for volume change estimation.

Free-breathing radial acquisitions of the heart

There is considerable interest in performing free-breathing acquisitions of the heart in order to obtain high-quality images without the need for multiple, long breathholds. In this article a 3D motion-correction method is described that is based on image registration of in-plane data and through-plane slice tracking. A number of fast radial undersampled images are acquired, each of which is free of motion artifacts. Initially, in-plane translational and rotational motion between each image was corrected before combining the data to give a fully sampled image. At the next stage, correction of in-plane deformation, in addition to translations and rotations, was performed in the image domain. Through-plane translational motion was compensated using a navigator echo to move the acquisition plane. Using this method, information on the motion of the heart was captured at the same time as acquiring the image data. No motion model, assumptions about the motion, or training data are required. The method is demonstrated on phantom data and cardiac images acquired on free-breathing volunteers.

Information theoretic similarity measures in non-rigid registration

Mutual Information (MI) and Normalised Mutual Information (NMI) have enjoyed success as image similarity measures in medical image registration. More recently, they have been used for non-rigid registration, most often evaluated empirically as functions of changing registration parameter. In this paper we present expressions derived from intensity histogram representations of these measures, for their change in response to a local perturbation of a deformation field linking two images. These expressions give some insight into the operation of NMI in registration and are implemented as driving forces within a fluid registration framework. The performance of the measures is tested on publicly available simulated multi-spectral MR brain images.

An automated algorithm for the computation of brain volume change from sequential MRIs using an iterative principal component analysis and its evaluation for the assessment of whole-brain atrophy rates in patients with probable Alzheimer's disease

This article introduces an automated method for the computation of changes in brain volume from sequential magnetic resonance images (MRIs) using an iterative principal component analysis (IPCA) and demonstrates its ability to characterize whole-brain atrophy rates in patients with Alzheimer's disease (AD). The IPCA considers the voxel intensity pairs from coregistered MRIs and identifies those pairs a sufficiently large distance away from the iteratively determined PCA major axis. Analyses of simulated and real MRI data support the underlying assumption of a linear relationship in paired voxel intensities, identify an outlier distance threshold that optimizes the trade-off between sensitivity and specificity in the detection of small volume changes while accounting for global intensity changes, and demonstrate an ability to detect changes as small as 0.04% of brain volume without confounding effects of between-scan shifts in voxel intensity. In eight patients with probable AD and eight age-matched normal control subjects, the IPCA was comparable to the established but partly manual digital subtraction (DS) method in characterizing annual rates of whole-brain atrophy: resulting rates were correlated (Spearman rank correlation = 0.94, P < 0.0005) and comparable in distinguishing probable AD from normal aging (IPCA-detected atrophy rates: 2.17 +/- 0.52% per year in the patients vs. 0.41 +/- 0.22% per year in the controls [Wilcoxon-Mann-Whitney test P = 7.8 x 10(-4)]; DS-detected atrophy rates: 3.51 +/- 1.31% per year in the patients vs. 0.48 +/- 0.29% per year in the controls [P = 7.8 x 10(-4)]). The IPCA could be used in tracking the progression of AD, evaluating the disease-modifying effects of putative treatments, and investigating the course of other normal and pathological changes in brain morphology.

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.

Zen and the art of medical image registration: correspondence, homology, and quality

Nonrigid registration (NRR) is routinely used in the study of neuroanatomy and function and is a standard component of analysis packages such as SPM. There remain many unresolved correspondence problems that arise from attempts to associate functional areas with specific neuroanatomy and to compare both function and anatomy across patient groups. Problems can result from ignorance of the underlying neurology which is then compounded by unjustified inferences drawn from the results of NRR. Usually the magnitude, distribution, and significance of errors in NRR are unknown so the errors in correspondences determined by NRR are also unknown and their effect on experimental results cannot easily be quantified. In this paper we review the principles by which the presumed correspondence and homology of structures is used to drive registration and identify the conceptual and algorithmic areas where current techniques are lacking. We suggest that for applications using NRR to be robust and achieve their potential, context-specific definitions of correspondence must be developed which properly characterise error. Prior knowledge of image content must be utilised to monitor and guide registration and gauge the degree of success. The use of NRR in voxel-based morphometry is examined from this context and found wanting. We conclude that a move away from increasingly sophisticated but context-free registration technology is required and that the veracity of studies that rely on NRR should be keenly questioned when the error distribution is unknown and the results are unsupported by other contextual information.

Transient ischaemic attacks are associated with increased rates of global cerebral atrophy

OBJECTIVES

To determine whether patients presenting with a first transient ischaemic attack (TIA) subsequently show increased rates of brain atrophy compared with age matched controls; and to assess potential risk factors for brain atrophy in this group.

METHODS

60 patients with a first, isolated TIA and 26 age and sex matched controls were recruited. None had evidence of cognitive impairment. Vascular risk factors were treated appropriately. All subjects had volumetric imaging at the start of the study and one year later, when they were clinically reassessed. TIA patients also had serial dual echo brain imaging. Rates of whole brain atrophy were calculated from the registered volumetric scans, as was the incidence of new ischaemic lesions. In the TIA group, the degree of white matter disease was assessed. Atrophy rates and blood pressure were compared between patients and controls.

RESULTS

22 patients (37%) developed new "clinically silent" infarcts during follow up. The mean (SD) annualised percentage atrophy rate in the TIA group was significantly higher than in the controls, at 0.82 (0.39)% v 0.33 (0.3)% (p < 0.0001). In the TIA group, diastolic blood pressure (p = 0.004) and white matter disease severity (p < 0.001) were correlated with cerebral atrophy rate. Increased white matter disease was found in patients in whom new ischaemic lesions developed (p < 0.001).

CONCLUSIONS

Patients presenting with a first TIA have excess global brain atrophy compared with age matched controls over the subsequent year. Increased atrophy rates following a TIA may be directly or indirectly related to increasing white matter disease and diastolic hypertension. Future studies should assess whether this atrophy inevitably leads to cognitive decline, and whether aggressive treatment of risk factors for cerebrovascular disease (particularly hypertension) after a TIA can influence outcome.

Automatic differentiation of anatomical patterns in the human brain: validation with studies of degenerative dementias

We compared voxel-based morphometry (VBM) with independent accurate region-of-interest (ROI) measurements of temporal lobe structures in order to validate the usefulness of this fully automated and unbiased technique in Alzheimer's disease (AD) and semantic dementia (SD). In AD, ROI analyses appear more sensitive to volume loss in the amygdalae, whereas VBM analyses appear more sensitive to right middle temporal gyrus and regional hippocampal volume loss. In SD, ROI analyses appear more sensitive to left middle and inferior temporal gyrus volume loss, whereas VBM appears more sensitive to regional hippocampal volume loss. In addition the significance of volume reductions was generally less in VBM owing to more stringent corrections for multiple comparisons. In conclusion, the automated technique detects a general trend of atrophy similar to that of expertly labeled ROI measurements in AD and SD, although there are discrepancies in the ranking of severity and in the significance of volume reductions that are more marked in AD.

Progressive ventricular enlargement in patients with clinically isolated syndromes is associated with the early development of multiple sclerosis

BACKGROUND

In patients with clinically isolated syndromes (CIS) suggestive of multiple sclerosis (MS), the extent of brain magnetic resonance imaging (MRI) lesion load influences the probability and time to development of clinically definite MS. Cerebral atrophy is recognised in established MS, but its time of onset and whether, in early disease, it is related to MRI lesion load or clinical outcome is less certain.

OBJECTIVES

This study investigated ventricular enlargement over one year in CIS patients and explored its relation with lesion load and clinical outcome.

METHODS

A semi-automated thresholding technique for measuring ventricular volume (MIDAS) was applied to MRI scans in a cohort of 55 patients with CIS, recruited consecutively and imaged within three months of the onset of symptoms and again after one year.

RESULTS

Clinical MS had developed after one year in 16 of 40 patients with an abnormal baseline T2 scan and 2 of 15 with a normal scan. Significant ventricular enlargement was seen in 27 of 55 patients who fulfilled the new McDonald MRI criteria for MS using all available MRI at clinical follow up (median increase 0.3 cm(3), p=0.005) Significant increase in ventricular volume was also seen in the 18 of 55 patients who developed clinical MS over the follow up period (median increase 0.5 cm(3), p=0.006). There were significant but modest correlations between baseline lesion measures and subsequent ventricular enlargement.

CONCLUSIONS

(1) Lesions and atrophy are both associated with early relapse leading to a diagnosis of clinical MS; (2) while lesions contribute to the development of atrophy, atrophy may also develop by other mechanisms. This suggests that MR measures have a complementary role in monitoring the course of MS, even from the earliest clinical stage.

Rates of global and regional cerebral atrophy in AD and frontotemporal dementia

OBJECTIVE

Serial registered MRI provides a reproducible technique for detecting progressive cerebral atrophy in vivo and was used to determine if there were differences between the rates of cerebral atrophy in AD and frontotemporal dementia (FTD).

METHODS

Eighty-four patients with dementia (54 AD and 30 FTD) and 27 age-matched control subjects each had at least two volumetric MR scans. Serial scans were positionally matched (registered), and brain volume loss was determined by calculation of the brain boundary shift integral.

RESULTS

There was a difference between the rates of whole-brain atrophy in patients (mean annual volume loss 2.7% of total brain volume) and in control subjects (mean annual volume loss 0.5%). AD and FTD were associated with different rates of atrophy (mean annual losses 2.4 and 3.2%). The range of atrophy rates in the FTD group (0.3 to 8.0%) greatly exceeded that in the AD group (0.5 to 4.7%). Frontal-variant FTD was associated with a wider range of atrophy rates than temporal-variant FTD. Analysis of regional brain atrophy rates revealed that there was widespread symmetrically distributed cerebral volume loss in AD, whereas in frontal FTD there was greater atrophy anteriorly and in temporal FTD the atrophy rate was greatest in the left anterior cerebral cortex.

CONCLUSIONS

Both AD and FTD patients had increased rates of brain atrophy. Whereas the patients with AD were associated with a relatively restricted spread of atrophy rates, the greater spread of rates observed in the patients with FTD may reflect the heterogeneity of disease in FTD, with differences observed between frontal and temporal FTD. Increased rates of whole-brain atrophy did not discriminate AD from FTD, but analysis of regional atrophy rates revealed marked differences between patient groups.

Normalization of cerebral volumes by use of intracranial volume: implications for longitudinal quantitative MR imaging

BACKGROUND AND PURPOSE

MR-based volumetric measures of cerebral structures are increasingly used for diagnostic purposes and to measure progression of atrophy. Variations in individual head size may be corrected by normalization with use of a total intracranial volume (TIV) measurement. The TIV also may be used to correct for voxel size fluctuations in serial studies. The TIV should be measured from the same images used for structural volumetry, usually T1-weighted imaging. The objectives were to show that normalization with TIV reduces interindividual variation, to develop and validate a simple protocol for measuring TIV from T1-weighted MR images, and to apply TIV normalization to serial brain measures in controls and subjects with Alzheimer disease (AD).

METHODS

We measured TIV with a semiautomated segmentation technique on T1- and T2-weighted MR images in 55 controls, 10 AD patients, and two persons at risk of familial AD. Whole-brain volumes also were measured and normalized with TIVs.

RESULTS

The TIV normalization of cross-sectional brain volumes significantly reduced interindividual variation; the coefficient of variation (CV) was reduced from 10.0% to 6.0% in controls (P <.001). The TIVs measured on T1-weighted images had low variability (CV, 0.16%) and did not differ significantly from those measured on T2-weighted images (P =.16). The TIV normalization of serial brain-volume measurements reduced interimage differences caused by voxel-scaling variations (CV reduced from 1.3% to 0.5%, P =.002) in 10 controls and five AD patients.

CONCLUSION

Structural volumes should be normalized with a TIV, measured cross-sectionally, to reduce interindividual variation, and longitudinally with a concurrent measurement, to reduce subtle interimage differences. This may have important implications in progression studies.

Imaging of onset and progression of Alzheimer's disease with voxel-compression mapping of serial magnetic resonance images

BACKGROUND

Early diagnosis and monitoring of the progression of Alzheimer's disease is important for the development of therapeutic strategies. To detect the earliest structural brain changes, individuals need to be studied before symptom onset. We used an imaging technique known as voxel-compression mapping to localise progressive atrophy in patients with preclinical Alzheimer's disease.

METHODS

Four symptom-free individuals from families with early-onset Alzheimer's disease with known autosomal dominant mutations underwent serial magnetic resonance imaging (MRI) over 5-8 years. All four became symptomatic during follow-up. 20 individuals with a clinical diagnosis of probable Alzheimer's disease and 20 control participants also underwent serial MR imaging. A non-linear fluid matching algorithm was applied to register repeat scans onto baseline imaging. Jacobian determinants were used to create the voxel-compression maps.

FINDINGS

Progressive atrophy was revealed in presymptomatic individuals, with posterior cingulate and neocortical temporoparietal cortical losses, and medial temporal-lobe atrophy. In patients with known Alzheimer's disease, atrophy was widespread apart from in the primary motor and sensory cortices and cerebellum, reflecting the clinical phenomenology.

INTERPRETATION

Voxel-compression maps confirmed early involvement of the medial temporal lobes, but also showed posterior cingulate and temporoparietal cortical losses at presymptomatic stage. This technique could be applied diagnostically and used to monitor the effects of therapeutic intervention.

Patterns of temporal lobe atrophy in semantic dementia and Alzheimer's disease

Volumetric magnetic resonance imaging analyses of 30 subjects were undertaken to quantify the global and temporal lobe atrophy in semantic dementia and Alzheimer's disease. Three groups of 10 subjects were studied: semantic dementia patients, Alzheimer's disease patients, and control subjects. The temporal lobe structures measured were the amygdala, hippocampus, entorhinal cortex, parahippocampal gyrus, fusiform gyrus, and superior, middle, and inferior temporal gyri. Semantic dementia and Alzheimer's disease groups did not differ significantly on global atrophy measures. In semantic dementia, there was asymmetrical temporal lobe atrophy, with greater left-sided damage. There was an anteroposterior gradient in the distribution of temporal lobe atrophy, with more marked atrophy anteriorly. All left anterior temporal lobe structures were affected in semantic dementia, with the entorhinal cortex, amygdala, middle and inferior temporal gyri, and fusiform gyrus the most severely damaged. Asymmetrical, predominantly anterior hippocampal atrophy was also present. In Alzheimer's disease, there was symmetrical atrophy of the entorhinal cortex, hippocampus, and amygdala, with no evidence of an anteroposterior gradient in the distribution of temporal lobe or hippocampal atrophy. These data demonstrate that there is a marked difference in the distribution of temporal lobe atrophy in semantic dementia and Alzheimer's disease. In addition, the pattern of atrophy in semantic dementia suggests that semantic memory is subserved by anterior temporal lobe structures, within which the middle and inferior temporal gyri may play a key role.

Progressive brain atrophy on serial MRI in dementia with Lewy bodies, AD, and vascular dementia

The authors determined rates of brain atrophy, as assessed by the boundary shift integral on serial MRI, in patients with dementia with Lewy Bodies (DLB, n = 10), AD (n = 9), vascular dementia (VaD, n = 9), and age-matched controls (n = 20). Mean % +/- SD atrophy rates per year were as follows: DLB, 1.4 +/- 1.1; AD, 2.0 +/- 0.9; VaD, 1.9 +/- 1.1; and controls, 0.5 +/- 0.7. Dementia subjects had higher rates than controls (p < 0.001), but there were no significant differences between the three dementia groups. The authors found accelerating atrophy with increasing severity of cognitive impairment, further emphasizing the need for early diagnosis and intervention in dementia.