ISLES 2015 - A public evaluation benchmark for ischemic stroke lesion segmentation from multispectral MRI.

Ischemic stroke is the most common cerebrovascular disease, and its diagnosis, treatment, and study relies on non-invasive imaging. Algorithms for stroke lesion segmentation from magnetic resonance imaging (MRI) volumes are intensely researched, but the reported results are largely incomparable due to different datasets and evaluation schemes. We approached this urgent problem of comparability with the Ischemic Stroke Lesion Segmentation (ISLES) challenge organized in conjunction with the MICCAI 2015 conference. In this paper we propose a common evaluation framework, describe the publicly available datasets, and present the results of the two sub-challenges: Sub-Acute Stroke Lesion Segmentation (SISS) and Stroke Perfusion Estimation (SPES). A total of 16 research groups participated with a wide range of state-of-the-art automatic segmentation algorithms. A thorough analysis of the obtained data enables a critical evaluation of the current state-of-the-art, recommendations for further developments, and the identification of remaining challenges. The segmentation of acute perfusion lesions addressed in SPES was found to be feasible. However, algorithms applied to sub-acute lesion segmentation in SISS still lack accuracy. Overall, no algorithmic characteristic of any method was found to perform superior to the others. Instead, the characteristics of stroke lesion appearances, their evolution, and the observed challenges should be studied in detail. The annotated ISLES image datasets continue to be publicly available through an online evaluation system to serve as an ongoing benchmarking resource (www.isles-challenge.org).

Image-based in vivo assessment of targeting accuracy of stereotactic brain surgery in experimental rodent models.

Stereotactic neurosurgery is used in pre-clinical research of neurological and psychiatric disorders in experimental rat and mouse models to engraft a needle or electrode at a pre-defined location in the brain. However, inaccurate targeting may confound the results of such experiments. In contrast to the clinical practice, inaccurate targeting in rodents remains usually unnoticed until assessed by ex vivo end-point histology. We here propose a workflow for in vivo assessment of stereotactic targeting accuracy in small animal studies based on multi-modal post-operative imaging. The surgical trajectory in each individual animal is reconstructed in 3D from the physical implant imaged in post-operative CT and/or its trace as visible in post-operative MRI. By co-registering post-operative images of individual animals to a common stereotaxic template, targeting accuracy is quantified. Two commonly used neuromodulation regions were used as targets. Target localization errors showed not only variability, but also inaccuracy in targeting. Only about 30% of electrodes were within the subnucleus structure that was targeted and a-specific adverse effects were also noted. Shifting from invasive/subjective 2D histology towards objective in vivo 3D imaging-based assessment of targeting accuracy may benefit a more effective use of the experimental data by excluding off-target cases early in the study.

Persistent Impact of In utero Irradiation on Mouse Brain Structure and Function Characterized by MR Imaging and Behavioral Analysis.

Prenatal irradiation is known to perturb brain development. Epidemiological studies revealed that radiation exposure during weeks 8-15 of pregnancy was associated with an increased occurrence of mental disability and microcephaly. Such neurological deficits were reproduced in animal models, in which rodent behavioral testing is an often used tool to evaluate radiation-induced defective brain functionality. However, up to now, animal studies suggested a threshold dose of around 0.30 Gray (Gy) below which no behavioral alterations can be observed, while human studies hinted at late defects after exposure to doses as low as 0.10 Gy. Here, we acutely irradiated pregnant mice at embryonic day 11 with doses ranging from 0.10 to 1.00 Gy. A thorough investigation of the dose-response relationship of altered brain function and architecture following in utero irradiation was achieved using a behavioral test battery and volumetric 3D T2-weighted magnetic resonance imaging (MRI). We found dose-dependent changes in cage activity, social behavior, anxiety-related exploration, and spatio-cognitive performance. Although behavioral alterations in low-dose exposed animals were mild, we did unveil that both emotionality and higher cognitive abilities were affected in mice exposed to ≥0.10 Gy. Microcephaly was apparent from 0.33 Gy onwards and accompanied by deviations in regional brain volumes as compared to controls. Of note, total brain volume and the relative volume of the ventricles, frontal and posterior cerebral cortex, cerebellum, and striatum were most strongly correlated to altered behavioral parameters. Taken together, we present conclusive evidence for persistent low-dose effects after prenatal irradiation in mice and provide a better understanding of the correlation between their brain size and performance in behavioral tests.

Impaired recognition of body expressions in the behavioral variant of frontotemporal dementia.

Progressive deterioration of social cognition and emotion processing are core symptoms of the behavioral variant of frontotemporal dementia (bvFTD). Here we investigate whether bvFTD is also associated with impaired recognition of static (Experiment 1) and dynamic (Experiment 2) bodily expressions. In addition, we compared body expression processing with processing of static (Experiment 3) and dynamic (Experiment 4) facial expressions, as well as with face identity processing (Experiment 5). The results reveal that bvFTD is associated with impaired recognition of static and dynamic bodily and facial expressions, while identity processing was intact. No differential impairments were observed regarding motion (static vs. dynamic) or category (body vs. face). Within the bvFTD group, we observed a significant partial correlation between body and face expression recognition, when controlling for performance on the identity task. Voxel-Based Morphometry (VBM) analysis revealed that body emotion recognition was positively associated with gray matter volume in a region of the inferior frontal gyrus (pars orbitalis/triangularis). The results are in line with a supramodal emotion recognition deficit in bvFTD.

Early decrease of type 1 cannabinoid receptor binding and phosphodiesterase 10A activity in vivo in R6/2 Huntington mice.

Several lines of evidence imply early alterations in endocannabinoid and phosphodiesterase 10A (PDE10A) signaling in Huntington disease (HD). Using [(18)F]MK-9470 and [(18)F]JNJ42259152 small-animal positron emission tomography (PET), we investigated for the first time cerebral changes in type 1 cannabinoid (CB1) receptor binding and PDE10A levels in vivo in presymptomatic, early symptomatic, and late symptomatic HD (R6/2) mice, in relation to glucose metabolism ([(18)F]FDG PET), brain morphology (magnetic resonance imaging) and motor function. Ten R6/2 and 16 wild-type (WT) mice were investigated at 3 different time points between the age of 4 and 13 weeks. Parametric CB1 receptor and PDE10A images were anatomically standardized to Paxinos space and analyzed voxelwise. Volumetric microMRI imaging was performed to assess HD pathology. In R6/2 mice, CB1 receptor binding was decreased in comparison with WT in a cluster comprising the bilateral caudate-putamen, globus pallidus, and thalamic nucleus at week 5 (-8.1% ± 2.6%, p = 1.7 × 10(-5)). Longitudinal follow-up showed further progressive decline compared with controls in a cluster comprising the bilateral hippocampus, caudate-putamen, globus pallidus, superior colliculus, thalamic nucleus, and cerebellum (late vs. presymptomatic age: -13.7% ± 3.1% for R6/2 and +1.5% ± 4.0% for WT, p = 1.9 × 10(-5)). In R6/2 mice, PDE10A binding potential also decreased over time to reach significance at early and late symptomatic HD (late vs. presymptomatic age: -79.1% ± 1.9% for R6/2 and +2.1% ± 2.7% for WT, p = 1.5 × 10(-4)). The observed changes in CB1 receptor and PDE10A binding were correlated to anomalies exhibited by R6/2 animals in motor function, whereas no correlation was found with magnetic resonance imaging-based striatal volume. Our findings point to early regional dysfunctions in endocannabinoid and PDE10A signaling, involving the caudate-putamen and lateral globus pallidus, which may play a role in the progression of the disease in R6/2 animals. PET quantification of in vivo CB1 and/or PDE10A binding may thus be useful early biomarkers for HD. Our results also provide evidence of subtle motor deficits at earlier stages than previously described.

Multimodal imaging of subventricular zone neural stem/progenitor cells in the cuprizone mouse model reveals increased neurogenic potential for the olfactory bulb pathway, but no contribution to remyelination of the corpus callosum.

Multiple sclerosis is a devastating demyelinating disease of the central nervous system (CNS) in which endogenous remyelination, and thus recovery, often fails. Although the cuprizone mouse model allowed elucidation of many molecular factors governing remyelination, currently very little is known about the spatial origin of the oligodendrocyte progenitor cells that initiate remyelination in this model. Therefore, we here investigated in this model whether subventricular zone (SVZ) neural stem/progenitor cells (NSPCs) contribute to remyelination of the splenium following cuprizone-induced demyelination. Experimentally, from the day of in situ NSPC labeling, C57BL/6J mice were fed a 0.2% cuprizone diet during a 4-week period and then left to recover on a normal diet for 8weeks. Two in situ labeling strategies were employed: (i) NSPCs were labeled by intraventricular injection of micron-sized iron oxide particles and then followed up longitudinally by means of magnetic resonance imaging (MRI), and (ii) SVZ NSPCs were transduced with a lentiviral vector encoding the eGFP and Luciferase reporter proteins for longitudinal monitoring by means of in vivo bioluminescence imaging (BLI). In contrast to preceding suggestions, no migration of SVZ NSPC towards the demyelinated splenium was observed using both MRI and BLI, and further validated by histological analysis, thereby demonstrating that SVZ NSPCs are unable to contribute directly to remyelination of the splenium in the cuprizone model. Interestingly, using longitudinal BLI analysis and confirmed by histological analysis, an increased migration of SVZ NSPC-derived neuroblasts towards the olfactory bulb was observed following cuprizone treatment, indicative for a potential link between CNS inflammation and increased neurogenesis.

A posteriori registration and subtraction of panoramic compared with intraoral radiography.

OBJECTIVES: To demonstrate the feasibility of panoramic image subtraction for implant assessment. STUDY DESIGN: Three titanium implants were inserted into a fresh pig mandible. One intraoral and 2 panoramic images were obtained at baseline and after each of 6 incremental (0.3, 0.6, 1.0, 1.5, 2.0, 2.5 mm) removals of bone. For each incremental removal of bone, the mandible was removed from and replaced in the holding device. Images representing incremental bone removals were registered by computer with the baseline images and subtracted. Assessment of the subtraction images was based on visual inspection and analysis of structured noise. RESULTS: Incremental bone removals were more visible in intraoral than in panoramic subtraction images; however, computer-based registration of panoramic images reduced the structured noise and enhanced the visibility of incremental removals. CONCLUSION: The feasibility of panoramic image subtraction for implant assessment was demonstrated.