Filters








97 Hits in 0.96 sec

Explainable Brain Disease Classification and Relevance-Guided Deep Learning [article]

Christian Tinauer, Stefan Heber, Lukas Pirpamer, Anna Damulina, Reinhold Schmidt, Rudolf Stollberger, Stefan Ropele, Christian Langkammer
2021 medRxiv   pre-print
Deep neural networks are increasingly used for neurological disease classification by MRI, but the networks' decisions are not easily interpretable by humans. Heat mapping by deep Taylor decomposition revealed that (potentially misleading) image features even outside of the brain tissue are crucial for the classifier's decision. We propose a regularization technique to train convolutional neural network (CNN) classifiers utilizing relevance-guided heat maps calculated online during training.
more » ... method was applied using T1-weighted MR images from 128 subjects with Alzheimer's disease (mean age=71.9+-8.5 years) and 290 control subjects (mean age=71.3+-6.4 years). The developed relevance-guided framework achieves higher classification accuracies than conventional CNNs but more importantly, it relies on less but more relevant and physiological plausible voxels within brain tissue. Additionally, preprocessing effects from skull stripping and registration are mitigated, rendering this practically useful in deep learning neuroimaging studies. Understanding the decision mechanisms underlying CNNs, these results challenge the notion that unprocessed T1-weighted brain MR images in standard CNNs yield higher classification accuracy in Alzheimer's disease than solely atrophy.
doi:10.1101/2021.09.09.21263013 fatcat:4ybzxjz6bbf5pnkky7jseodzbq

Foundations of advanced magnetic resonance imaging

Roland Bammer, Stefan Skare, Rexford Newbould, Chunlei Liu, Vincent Thijs, Stefan Ropele, David B. Clayton, Gunnar Krueger, Michael E. Moseley, Gary H. Glover
2005 Neurotherapeutics  
During the past decade, major breakthroughs in magnetic resonance imaging (MRI) quality were made by means of quantum leaps in scanner hardware and pulse sequences. Some advanced MRI techniques have truly revolutionized the detection of disease states and MRI can nowwithin a few minutes-acquire important quantitative information noninvasively from an individual in any plane or volume at comparatively high resolution. This article provides an overview of the most common advanced MRI methods
more » ... ding diffusion MRI, perfusion MRI, functional MRI, and the strengths and weaknesses of MRI at high magnetic field strengths.
doi:10.1007/bf03206665 fatcat:a66utm5yn5dwxfvxx5s7hwg6ve

Foundations of advanced magnetic resonance imaging

Roland Bammer, Stefan Skare, Rexford Newbould, Chunlei Liu, Vincent Thijs, Stefan Ropele, David B. Clayton, Gunnar Krueger, Michael E. Moseley, Gary H. Glover
2005 NeuroRx  
During the past decade, major breakthroughs in magnetic resonance imaging (MRI) quality were made by means of quantum leaps in scanner hardware and pulse sequences. Some advanced MRI techniques have truly revolutionized the detection of disease states and MRI can nowwithin a few minutes-acquire important quantitative information noninvasively from an individual in any plane or volume at comparatively high resolution. This article provides an overview of the most common advanced MRI methods
more » ... ding diffusion MRI, perfusion MRI, functional MRI, and the strengths and weaknesses of MRI at high magnetic field strengths.
doi:10.1602/neurorx.2.2.167 pmid:15897944 pmcid:PMC1064985 fatcat:vyfeyhkvtvfe3kg6iyxxgzzrgu

MTI of white matter hyperintensities

Franz Fazekas, Stefan Ropele, Christian Enzinger, Faton Gorani, Alexandra Seewann, Katja Petrovic, Reinhold Schmidt
2005 Brain  
The severity of tissue changes associated with incidental white matter hyperintensities (WMH) in the elderly cannot be sufficiently determined by conventional MRI. We, therefore, performed a regional analysis of the magnetization transfer ratio (MTR) maps obtained on a 1.5 T scanner from 198 neurologically asymptomatic participants of the Austrian Stroke Prevention Study (mean age 70, age range 52-87 years) in regard to WMH and predefined areas of normal appearing brain tissue. Fluid attenuated
more » ... inversion recovery MRI was used to grade lesion severity and for lesion volume measurements. The MTR of WMH was always significantly lower than that of normal appearing white matter (NAWM) with an overall relative reduction of $10% and decreased significantly with increasing scores of WMH severity (P = 0.02) and WMH volume (r = À0.24, P = 0.0016). NAWM MTR was not different between subjects with very few and extensive WMH and the WMH volume was associated with NAWM MTR of the frontal lobes only. Concerning a possible impact on cerebral functioning the MTR of the frontal NAWM was significantly associated with fine motor dexterity (P = 0.04) but not with cognitive performance. A significant decline of the MTR with aging was seen in both NAWM and cortex but not in WMH. We conclude that MTR measurements can serve to quantify WMH associated tissue damage. It is predominantly focal, relatively mild, increases with lesion size and may have remote effects on the frontal white matter. Keywords: ageing; cerebral small vessel disease; magnetization transfer imaging; white matter injury Abbreviations: ASPS = Austrian Stroke Prevention Study; FLAIR = fluid-attenuated inversion recovery; MS = multiple sclerosis; MTI = magnetization transfer imaging; MTR = magnetization transfer ratio; NAGM = normal appearing grey matter; NAWM = normal appearing white matter; PVH = periventricular hyperintensities; ROI = region of interest; WMH = white matter hyperintensities
doi:10.1093/brain/awh567 pmid:15958507 fatcat:khwzhpuchndtbn7oc6tvvgmllu

Multimodal assessment of white matter tracts in amyotrophic lateral sclerosis

Florian Borsodi, Valeriu Culea, Christian Langkammer, Michael Khalil, Lukas Pirpamer, Stefan Quasthoff, Christian Enzinger, Reinhold Schmidt, Franz Fazekas, Stefan Ropele, Yi Wang
2017 PLoS ONE  
doi:10.1371/journal.pone.0178371 pmid:28575122 pmcid:PMC5456080 fatcat:54txlcmgqrdgnpnqgseldyra2e

Quantitative Susceptibility Mapping in Multiple Sclerosis

Christian Langkammer, Tian Liu, Michael Khalil, Christian Enzinger, Margit Jehna, Siegrid Fuchs, Franz Fazekas, Yi Wang, Stefan Ropele
2013 Radiology  
INTRODUCTION: Multiple Sclerosis (MS) affects approximately 1 out of 1000 people in the United States. Patients suffer from a variety of neurologic symptoms including visual deficits, gait impairment, and muscle weakness leading to disabilities. The pathological hallmark of MS is the presence of demyelinating/inflammatory lesions in the brain and spinal cord. The appearance of lesions is best demonstrated in the living patient by magnetic resonance imaging (MRI). Conventional MR T2 hyperintense
more » ... lesion number or lesion volume have not yielded satisfactory correlations and cannot adequately inform clinical diagnosis and decisions. The lack of pathological specificity of T2 hyperintense lesions ranging from edema, mild demyelination, and scar-like myelin gliosis may be responsible for the weak associations. Limited research has been done regarding the iron accumulation of the MS brain. Reports of both MR and histopathology have shown increased iron deposition, particularly in vessel walls of the veins in MS 1,2 and in basal ganglia areas 3,4,5,6 . Whether the study of iron property can afford more specific characterization of MS brain, function and treatment need to be further investigated. A new technique, quantitative susceptibility mapping (QSM), allows quantitative evaluations of iron deposition in the brain. A unique attribute of QSM is the ability of discriminating calcium induced susceptibility, thus it is more intrinsic to iron susceptibility. We performed QSM in patients with MS to characterize lesion burden and demonstrate the potential of QSM as a quantitative imaging marker for monitoring disease progression and/or responses to treatment. METHODS AND MATERIALS: 3 MS patients (1 Male, 2 Females; mean age: 52.8 ± 6.0) and 2 normal control subjects (1 Male, 1 Female; mean age: 73 ± 1.8) were scanned at a 3 Tesla Siemens system (Siemens Verio, Germany). Image Acquisition: 1) For QSM: 3D multi-echo T2*-weighted spoiled gradient echo sequence (TR/TE {min, max} / FA =55 ms/ {3.6, 45} ms/ 15 0 ); Image resolution= 0.9x0.9x1.5 mm 3 ; Bandwidth = 240 Hz/Pixel; scan time= 7:40. 2) FLAIR: 2D FSE (TR/TE/TI/FA=9000ms/96ms/2500/150°/46, resolution=0.9 x 0.9 x 2 mm 3 ) scan time=4:14. 3) Pre-and post-contrast T1 weighted: High resolution none-selective SPGR T1-weighted image 7 , scanning parameters (TR/TE/FA/slices=7.1ms/3.1ms/12°/512, resolution=1.0 x 1.0 x 0.4 mm 3 ) scan time=5:28. QSM images were reconstructed using a morphology-enabled dipole inversion (MEDI) algorithm offline 8 . Images including QSM and SWI were sent to the PACS system where two neuroradiologists reviewed the images independently. Quantitative Measurements: 1) Autoregional quantification was conducted 9,10 . QSM were coregistered to the T1 structural image. Automated segmentation was performed on structural scan using FreeSurfer 11 for masking 3D volume of interests (VOIs) and used to extract mean regional susceptibility values. The VOIs used in this study include: MS lesions, corpus callosum (CC), caudate, putamen, and lateral ventricles. To compare with previous reported normal brain data, regions of interest were manually outlined 8 . Summary statistics mean and standard deviation were reported.
doi:10.1148/radiol.12120707 pmid:23315661 pmcid:PMC3632806 fatcat:mnw7qcwnojdjvk6bs7fdbsdzmu

Susceptibility induced gray–white matter MRI contrast in the human brain

Christian Langkammer, Nikolaus Krebs, Walter Goessler, Eva Scheurer, Kathrin Yen, Franz Fazekas, Stefan Ropele
2012 NeuroImage  
Keywords: MR phase contrast Susceptibility contrast Gray-white matter contrast Brain iron Myelin MR phase images have shown significantly improved contrast between cortical gray and white matter regions compared to magnitude images obtained with gradient echo sequences. A variety of underlying biophysical mechanisms (including iron, blood, myelin content, macromolecular chemical exchange, and fiber orientation) have been suggested to account for this observation but assessing the individual
more » ... ribution of these factors is limited in vivo. For a closer investigation of iron and myelin induced susceptibility changes, postmortem MRI of six human corpses (age range at death: 56-80 years) was acquired in situ. Following autopsy, the iron concentrations in the frontal and occipital cortex as well as in white matter regions were chemically determined. The magnetization transfer ratio (MTR) was used as an indirect measure for myelin content. Susceptibility effects were assessed separately by determining R2* relaxation rates and quantitative phase shifts. Contributions of myelin and iron to local variations of the susceptibility were assessed by univariate and multivariate linear regression analysis. Mean iron concentration was lower in the frontal cortex than in frontal white matter (26 ± 6 vs. 45 ± 6 mg/kg wet tissue) while an inverse relation was found in the occipital lobe (cortical gray matter: 41± 10 vs. white matter: 34± 10 mg/kg wet tissue). Multiple regression analysis revealed iron and MTR as independent predictors of the effective transverse relaxation rate R2* but solely MTR was identified as source of MR phase contrast. R2* was correlated with iron concentrations in cortical gray matter only (r= 0.42, p b 0.05). In conclusion, MR phase contrast between cortical gray and white matter can be mainly attributed to variations in myelin content, but not to iron concentration. Both, myelin and iron impact the effective transverse relaxation rate R2* significantly. Magnitude contrast is limited because it only reflects the extent but not the direction of the susceptibility shift. Abbreviations: MT, magnetization transfer; MTR, magnetization transfer ratio; R2, transverse relaxation rate; R2′, rf-reversible dephasing rate; R2*, effective transverse relaxation rate.
doi:10.1016/j.neuroimage.2011.08.045 pmid:21893208 pmcid:PMC3236994 fatcat:orcimzqftfbo3pegih55gznteu

Changes in cellular ferritin distribution of human brain tissue during autolysis revealed using analytical electron microscopy [article]

Sowmya Sunkara, Snjezana Radulovic, Saska Lipovsek, Christoph Birkl, Stefan Eggenreich, Anna Maria Birkl-Toeglhofer, Maximilian Schinagl, Daniel Funk, Michael Stoeger-Pollach, Johannes Haybaeck, Walter Goessler, Stefan Ropele (+1 others)
2022 bioRxiv   pre-print
Iron in the human brain is stored in the cores of ferritin proteins. Excess iron has been implicated in neurological disorders, so a careful balance of the iron concentration is thought to be essential for human health. Though it is known that oligodendrocytes contain most of the brain ferritin, followed by other glial cells and neurons, little is known about changes in the cellular ferritin distribution when total iron load changes. Moreover, changes in the number of filled ferritins or in
more » ... r cellular distribution during autolysis have not been revealed yet. To reveal these changes, we examined brain samples taken from six deceased humans at different post mortem times. To compare samples with different ferritin content, the samples included the frontal gray and frontal white matter, which store a low amount of iron, and the putamen and globus pallidus, which store a high amount of iron. We quantified the number of loaded ferritin cores from iron L - elemental maps obtained with energy filtered transmission electron microscopy and compared this number to the iron concentrations determined using inductively coupled plasma mass spectrometry and quantitative magnetic resonance imaging. The more loaded ferritin cores we found in the tissues, the higher the proportion of ferritins found in neurons instead of glial cells, indicating that total ferritin load scales with the proportion of ferritin that is located in neurons. Second, we found a statistically significant correlation between the post mortem interval and the amount / concentration of loaded ferritins and found that this concentration is highly dependent on the post mortem interval (PMI): the longer the PMI, the lower the ferritin concentration found in a sample. Third, the longer the post mortem interval, the higher the proportion of ferritins found in neurons instead of glial cells, indicating that ferritin loss during autolysis affects neurons and glial cells at different rates.
doi:10.1101/2022.01.26.477869 fatcat:2ehtaed62jac3nvus5wgnjzyda

Shape of the Central Sulcus and Disability After Subcortical Stroke

Eric Jouvent, Zhong Yi Sun, François De Guio, Edouard Duchesnay, Marco Duering, Stefan Ropele, Martin Dichgans, Jean-François Mangin, Hugues Chabriat
2016 Stroke  
and Purpose-Both brain and cognitive reserves modulate the clinical impact of chronic brain diseases. Whether a motor reserve also modulates the relationships between stroke and disability is unknown. We aimed to determine whether the shape of the central sulcus, a marker of the development of underlying motor connections, is independently associated with disability in patients with a positive history of small subcortical ischemic stroke. Methods-Shapes of central sulci were reconstructed from
more » ... igh-resolution magnetic resonance imaging and ordered without supervision according to a validated algorithm in 166 patients with a positive history of small subcortical ischemic stroke caused by CADASIL (Cerebral Autosomal Dominant Arteriopathy With Subcortical Infarcts and Leukoencephalopathy), a severe monogenic cerebral small vessel disease affecting young patients. Ordinal logistic regression modeling was used to test the relationships between modified Rankin scale, a disability scale strongly weighted toward motor disability, and sulcal shape. Results-Modified Rankin scale was strongly associated with sulcal shape, independent of age, sex, and level of education (proportional odds ratio =1.19, 95% confidence interval =1.06-1.35; P=0.002). Results remained significant after further adjustment for brain atrophy, volume of lacunes, and volume of white matter hyperintensities of presumed vascular origin. Conclusions-The severity of disability in patients with a positive history of small subcortical ischemic stroke caused by a severe cerebral small vessel disease is related to the shape of the central sulcus, independently of the main determinants of disability. These results support the concept of a motor reserve that could modulate the clinical severity in patients with a positive history of small subcortical ischemic stroke.
doi:10.1161/strokeaha.115.012562 pmid:26941259 fatcat:igfxne5jkffohbvnq6jhvcg6ey

Magnetic resonance elastography of the human brain using a multiphase DENSE acquisition

Johannes Strasser, Michaela Tanja Haindl, Rudolf Stollberger, Franz Fazekas, Stefan Ropele
2019 Magnetic Resonance in Medicine  
In magnetic resonance elastography (MRE), a series of time-shifted images is acquired at specific phase offsets in relation to an induced mechanical excitation. To efficiently gather the set of phase offset images and to overcome limitations due to prolonged TEs and related susceptibility artifacts at low-frequency MRE, we developed an improved displacement encoding with a stimulated echoes (DENSE) method.
doi:10.1002/mrm.27672 pmid:30693964 pmcid:PMC6590321 fatcat:k3se2wknxzfq5a4ldr4enqzntm

Prediction of 3-year clinical course in CADASIL

Eric Jouvent, Edouard Duchesnay, Foued Hadj-Selem, François De Guio, Jean-François Mangin, Dominique Hervé, Marco Duering, Stefan Ropele, Reinhold Schmidt, Martin Dichgans, Hugues Chabriat
2016 Neurology  
Stefan Ropele: study design, manuscript correction. Reinhold Schmidt: study design, manuscript correction. Martin Dichgans: study design, data acquisition, data analysis.  ... 
doi:10.1212/wnl.0000000000003252 pmid:27694265 pmcid:PMC5089530 fatcat:uajt7ujh7vg5dhp7z2jazpm7fa

European Study on Intravenous Immunoglobulin in Multiple Sclerosis

Massimo Filippi, Maria A. Rocca, Elisabetta Pagani, Giuseppe Iannucci, Maria Pia Sormani, Franz Fazekas, Stefan Ropele, Otto R. Hommes, Giancarlo Comi
2004 Archives of Neurology  
Author Contributions: Study concept and design: Filippi, Fazekas, Ropele, and Hommes. Acquisition of data: Filippi, Fazekas, Ropele, and Hommes.  ...  Analysis and interpretation of data: Filippi, Assunta, Pagani, Iannucci, Sormani, Fazkas, Ropele, and Comi. Drafting of the manuscript: Filippi.  ... 
doi:10.1001/archneur.61.9.1409 pmid:15364687 fatcat:psvleu2zrvgypcf4qios4wbmx4

In vivo assessment of anisotropy of apparent magnetic susceptibility in white matter from a single orientation acquisition

Renat Sibgatulin, Daniel Güllmar, Andreas Deistung, Stefan Ropele, Jürgen R. Reichenbach
2021 NeuroImage  
Multiple studies have reported a significant dependence of the effective transverse relaxation rate constant (R2*) and the phase of gradient-echo based (GRE) signal on the orientation of white matter fibres in the human brain. It has also been hypothesized that magnetic susceptibility, as obtained by single-orientation quantitative susceptibility mapping (QSM), exhibits such a dependence. In this study, we investigated this hypothesized relationship in a cohort of healthy volunteers. We show
more » ... t R2* follows the predicted orientation dependence consistently across white matter regions, whereas the apparent magnetic susceptibility is related differently to fibre orientation across the brain and often in a complex non-monotonic manner. In addition, we explored the effect of fractional anisotropy measured by diffusion-weighted MRI on the strength of the orientation dependence and observed only a limited influence in many regions. However, with careful consideration of such an impact and the limitations imposed by the ill-posed nature of the dipole inversion process, it is possible to study magnetic susceptibility anisotropy in specific brain regions with a single orientation acquisition.
doi:10.1016/j.neuroimage.2021.118442 pmid:34339831 fatcat:efc3cqyq35bp5ebqxeci5z6vey

Fast quantitative susceptibility mapping using 3D EPI and total generalized variation

Christian Langkammer, Kristian Bredies, Benedikt A. Poser, Markus Barth, Gernot Reishofer, Audrey Peiwen Fan, Berkin Bilgic, Franz Fazekas, Caterina Mainero, Stefan Ropele
2015 NeuroImage  
published version features the final layout of the paper including the volume, issue and page numbers. Link to publication General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal
more » ... r the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal. If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the "Taverne" license above, please follow below link for the End User
doi:10.1016/j.neuroimage.2015.02.041 pmid:25731991 fatcat:mbosuwana5fdjed5otnn7j2kae

Different Types of White Matter Hyperintensities in CADASIL

Edouard Duchesnay, Fouad Hadj Selem, François De Guio, Mathieu Dubois, Jean-François Mangin, Marco Duering, Stefan Ropele, Reinhold Schmidt, Martin Dichgans, Hugues Chabriat, Eric Jouvent
2018 Frontiers in Neurology  
Conclusion: The results of the present study support the hypothesis that the whole extent of WMH results from a combination of different regional populations of WMH, some of which are associated, for yet undetermined reasons, with milder forms of the disease.
doi:10.3389/fneur.2018.00526 pmid:30042721 pmcid:PMC6048276 fatcat:ndl23sxcozf7vphn3ttmr3xa7q
« Previous Showing results 1 — 15 out of 97 results