Search

Esclerosis múltiple; Brote; Resonancia magnética Esclerosi múltiple; Brot; Imatge per ressonància magnètica Multiple sclerosis; Outbreak; Magnetic resonance imaging Objective To study the clinico-radiological paradox in multiple sclerosis (MS) relapse by analyzing the number and location of gadolinium-enhanced (Gd+) lesions on brain MRI before methylprednisolone (MP) treatment. Methods We analyzed brain MRI from 90 relapsed MS patients in two Phase IV multicenter double-blind randomized clinical trials that showed the noninferiority of different routes and doses of MP administration. A 1.5- or 3-T brain MRI was performed at baseline before MP treatment and within 15 days of symptom onset. The number and location of Gd+ lesions were analyzed. Associations were studied using univariate analysis. Results Sixty-two percent of patients had at least 1 Gd+ brain lesion; the median number was 1 (interquartile range 0–4), and 41% of patients had 2 or more lesions. The most frequent location of Gd+ lesions was subcortical (41.4%). Gd+ brain lesions were found in 71.4% of patients with brainstem-cerebellum symptoms, 57.1% with spinal cord symptoms and 55.5% with optic neuritis (ON). Thirty percent of patients with brain symptoms did not have Gd+ lesions, and only 43.6% of patients had symptomatic Gd+ lesions. The univariate analysis showed a negative correlation between age and the number of Gd+ lesions (p = 0.002). Conclusion Most patients with relapse showed several Gd+ lesions on brain MRI, even when the clinical manifestation was outside of the brain. Our findings illustrate the clinico-radiological paradox in MS relapse and support the value of brain MRI in this scenario. Objetivo Estudiar la paradoja clínico-radiológica en el brote de la esclerosis múltiple (EM) mediante el análisis de lesiones captantes de gadolinio (Gd+) en la RM cerebral antes del tratamiento con metilprednisolona (MP). Métodos Analizamos la RM cerebral basal de 90 pacientes con EM en brote de 2 ensayos clínicos aleatorizados multicéntricos fase IV que demostraron la no inferioridad de diferentes vías y dosis de MP, realizadas antes del tratamiento con MP y en los 15 días siguientes a la aparición de los síntomas. Se analizaron el número y la localización de las lesiones Gd+. Se estudiaron las asociaciones mediante análisis univariado. Resultados El 62% de los pacientes tenía al menos una lesión Gd+ cerebral y el 41% de los pacientes tenía 2 o más lesiones. La localización más frecuente fue la subcortical (41,4%). Se encontraron lesiones Gd+ cerebrales en el 71,4% de los pacientes con síntomas de tronco cerebral o cerebelo, en el 57,1% con síntomas medulares y en el 55,5% con neuritis óptica. El 30% de los pacientes con síntomas cerebrales no tenían lesiones Gd+ y sólo el 4,.6% de los pacientes tenían lesiones Gd+ sintomáticas. El análisis univariante mostró una correlación negativa entre la edad y el número de lesiones Gd+ (p = 0,002). Conclusiones La mayoría de los pacientes en brote mostraron varias lesiones Gd+ en la RM cerebral, incluso cuando la manifestación clínica fue medular u óptica. Nuestros hallazgos ilustran la paradoja clínico-radiológica en el brote de la EM y apoyan el valor de la RM cerebral en este escenario. This work was supported in part by the Ministry of Health of Spain (grant numbers EC07/90278 and EC11/132) and personal grant Rio Hortega CM19/00042 to LMA.

Brain; Protocol design; Quantitative MRI (qMRI) Cerebro; Diseño de protocolo; Resonancia magnética cuantitativa (qMRI) Cervell; Disseny del protocol; Ressonància magnètica quantitativa (qMRI) Purpose: We investigate the feasibility of data-driven, model-free quantitative MRI (qMRI) protocol design on in vivo brain and prostate diffusion-relaxation imaging (DRI). Methods: We select subsets of measurements within lengthy pilot scans, without identifying tissue parameters for which to optimise for. We use the “select and retrieve via direct upsampling” (SARDU-Net) algorithm, made of a selector, identifying measurement subsets, and a predictor, estimating fully-sampled signals from the subsets. We implement both using artificial neural networks, which are trained jointly end-to-end. We deploy the algorithm on brain (32 diffusion-/T1-weightings) and prostate (16 diffusion-/T2-weightings) DRI scans acquired on three healthy volunteers on two separate 3T Philips systems each. We used SARDU-Net to identify sub-protocols of fixed size, assessing reproducibility and testing sub-protocols for their potential to inform multi-contrast analyses via the T1-weighted spherical mean diffusion tensor (T1-SMDT, brain) and hybrid multi-dimensional MRI (HM-MRI, prostate) models, for which sub-protocol selection was not optimised explicitly. Results: In both brain and prostate, SARDU-Net identifies sub-protocols that maximise information content in a reproducible manner across training instantiations using a small number of pilot scans. The sub-protocols support T1-SMDT and HM-MRI multi-contrast modelling for which they were not optimised explicitly, providing signal quality-of-fit in the top 5% against extensive sub-protocol comparisons. Conclusions: Identifying economical but informative qMRI protocols from subsets of rich pilot scans is feasible and potentially useful in acquisition-time-sensitive applications in which there is not a qMRI model of choice. SARDU-Net is demonstrated to be a robust algorithm for data-driven, model-free protocol design. This project was funded by the Engineering and Physical Sciences Research Council (EPSRC EP/R006032/1, M020533/1, G007748, I027084, N018702). This project has received funding under the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 634541 and 666992, and from: Rosetrees Trust (United Kingdom, funding FG); Prostate Cancer United Kingdom Targeted Call 2014 (Translational Research St.2, project reference PG14-018-TR2); Cancer Research United Kingdom grant ref. A21099; Spinal Research (United Kingdom), Wings for Life (Austria), Craig H. Neilsen Foundation (United States) for jointly funding the INSPIRED study; Wings for Life (#169111); United Kingdom Multiple Sclerosis Society (grants 892/08 and 77/2017); the Department of Health’s National Institute for Health Research (NIHR) Biomedical Research Centres and UCLH NIHR Biomedical Research Centre; Champalimaud Centre for the Unknown, Lisbon (Portugal); European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 101003390. FG is currently supported by the investigator-initiated PREdICT study at the Vall d’Hebron Institute of Oncology (Barcelona), funded by AstraZeneca and CRIS Cancer Foundation.

Cervell; Imatge per ressonància magnètica; Aprenentatge transductiu Cerebro; Imagen de resonancia magnética; Aprendizaje transductivo Brain; Magnetic resonance imaging; Transductive learning Segmentation of brain images from Magnetic Resonance Images (MRI) is an indispensable step in clinical practice. Morphological changes of sub-cortical brain structures and quantification of brain lesions are considered biomarkers of neurological and neurodegenerative disorders and used for diagnosis, treatment planning, and monitoring disease progression. In recent years, deep learning methods showed an outstanding performance in medical image segmentation. However, these methods suffer from generalisability problem due to inter-centre and inter-scanner variabilities of the MRI images. The main objective of the study is to develop an automated deep learning segmentation approach that is accurate and robust to the variabilities in scanner and acquisition protocols. In this paper, we propose a transductive transfer learning approach for domain adaptation to reduce the domain-shift effect in brain MRI segmentation. The transductive scenario assumes that there are sets of images from two different domains: (1) source—images with manually annotated labels; and (2) target—images without expert annotations. Then, the network is jointly optimised integrating both source and target images into the transductive training process to segment the regions of interest and to minimise the domain-shift effect. We proposed to use a histogram loss in the feature level to carry out the latter optimisation problem. In order to demonstrate the benefit of the proposed approach, the method has been tested in two different brain MRI image segmentation problems using multi-centre and multi-scanner databases for: (1) sub-cortical brain structure segmentation; and (2) white matter hyperintensities segmentation. The experiments showed that the segmentation performance of a pre-trained model could be significantly improved by up to 10%. For the first segmentation problem it was possible to achieve a maximum improvement from 0.680 to 0.799 in average Dice Similarity Coefficient (DSC) metric and for the second problem the average DSC improved from 0.504 to 0.602. Moreover, the improvements after domain adaptation were on par or showed better performance compared to the commonly used traditional unsupervised segmentation methods (FIRST and LST), also achieving faster execution time. Taking this into account, this work presents one more step toward the practical implementation of deep learning algorithms into the clinical routine. KK holds FI-DGR2017 grant from the Catalan Government with reference number 2017FI_B00372. This work has been supported by DPI2017-86696-R from the Ministerio de Ciencia y Tecnologia.

Body mass index; Structural brain abnormalities Índex de massa corporal; Anormalitats estructurals del cervell Índice de masa corporal; Anomalías estructurales del cerebro We aimed to evaluate the relationship between obesity and structural brain abnormalities assessed by magnetic resonance imaging using data from 45 observational epidemiological studies, where five articles reported prospective longitudinal results. In cross-sectional studies’ analyses, the pooled weighted mean difference for total brain volume (TBV) and gray matter volume (GMV) in obese/overweight participants was -11.59 (95 % CI: -23.17 to -0.02) and -10.98 (95 % CI: -20.78 to -1.18), respectively. TBV was adversely associated with BMI and WC, GMV with BMI, and hippocampal volume with BMI, WC, and WHR. WC/WHR are associated with a risk of lacunar and white matter hyperintensity (WMH). In longitudinal studies’ analyses, BMI was not statistically associated with the overall structural brain abnormalities (for continuous BMI: RR = 1.02, 95 % CI: 0.94–1.12; for categorial BMI: RR = 1.18, 95 % CI: 0.75–1.85). Small sample size of prospective longitudinal studies limited the power of its pooled estimates. A higher BMI is associated with lower brain volume while greater WC/WHR, but not BMI, is related to a risk of lacunar infarct and WMH. Future longitudinal research is needed to further elucidate the specific causal relationships and explore preventive measures. This work was supported by the National Natural Science Foundation of China (No. 82070851, 81870556, 81930019, 81770686, 81970591), Beijing Municipal Administration of Hospital’s Youth Program (QML20170204), Excellent Talents in Dongcheng District of Beijing.

Esquizofrènia; Volum cerebral; Imatges per ressonància magnètica Esquizofrenia; Volumen cerebral; Imagen de resonancia magnética Schizophrenia; Brain volume; Magnetic Resonance Imaging Reductions in whole brain and grey matter volumes are robust features of schizophrenia, yet their etiological influences are unclear. We investigated the association between the genetic and environmental risk for schizophrenia and brain volumes. Whole brain, grey matter and white matter volumes were established from structural MRIs from twins varying in their zygosity and concordance for schizophrenia. Hippocampal volumes were measured manually. We conducted between-group testing and full genetic modelling. Results: We included 168 twins in our study. Whole brain, grey matter, white matter and right hippocampal volumes were smaller in twins with schizophrenia. Twin correlations were larger for whole brain, grey matter and white matter volumes in monozygotic than dizygotic twins and were significantly heritable, whereas hippocampal volume was the most environmentally sensitive. There was a significant phenotypic correlation between schizophrenia and reductions in all the brain volumes except for that of the left hippocampus. For whole brain, grey matter and the right hippocampus the etiological links with schizophrenia were principally associated with the shared familial environment. Lower birth weight and perinatal hypoxia were both associated with lower whole brain volume and with lower white matter and grey matter volumes, respectively. Scan data were collected across 2 sites, and some groups were modest in size. Whole brain, grey matter and right hippocampal volume reductions are linked to schizophrenia through correlated familial risk (i.e., the shared familial environment). The degree of influence of etiological factors varies between brain structures, leading to the possibility of a neuroanatomically specific etiological imprint. The study was in part funded by a Wellcome Trust Research Training Fellowship to M. Picchioni (064971), a NARSAD Young Investigator Award to T. Toulopoulou and by the European Community’s Sixth Framework Programme through a Marie Curie Training Network called the European Twin Study Network on Schizophrenia (EUTwinsS). U. Ettinger was funded by the Deutsche Forschungsgemeinschaft (ET 31/2-1).