China
Africa
Explore chapters and articles related to this topic
Brain Imaging Data
Published in Atsushi Kawaguchi, Multivariate Analysis for Neuroimaging Data, 2021
fMRI measures changes in blood flow in the brain. The increase in hemoglobin to compensate for that consumed in the blood (oxidized) by brain activity is captured as a signal. This is called the blood oxygen level dependent (BOLD) effect; fMRI measures the change in the MR signal effected by BOLD to identify active parts of the brain. fMRI is spatiotemporal (4D) data: 3D brain images are measured over time. One voxel is 3 mm3, and 64 × 64 × 49 (voxel) 100 (time points) is a standard setting. The example in Figure 2.2 is called block design: the subject is periodically given a task, and the attendant brain activity is recorded. The voxel value time series related to the task correlates with the problem time series taking binary values of on and off. Sen et al. (2010) and Lewis et al. (2011) examined brain functions when research participants engaged in finger tapping with different types and speeds. There are other event-related designs proposed by Buckner et al. (1996).
A Neuroergonomics Approach to Human Performance in Aviation
Published in Michael A. Vidulich, Pamela S. Tsang, Improving Aviation Performance through Applying Engineering Psychology, 2019
Frédéric Dehais, Daniel Callan
A challenge of importance for neuroergonomics is to succeed in reproducing ecological conditions in well-controlled laboratory protocols and to determine solutions for application of portable devices to measure human performance in realistic settings. In laboratory settings, expensive high-resolution devices such as functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG) can be employed. Functional magnetic resonance imaging indirectly measures brain activity by primarily looking at blood oxygenation level dependent changes between various experimental conditions (Ogawa, Lee, Kay, & Tank, 1990). Excellent spatial resolution of brain activity, both cortical and subcortical, is provided by fMRI. However, one of the limitations of fMRI is that it lacks good temporal resolution (on the order of several seconds, limited by the slow rise of the hemodynamic response). MEG directly measures magnetic fields generated by simultaneous local field potentials in large groups of similarly oriented neurons (Baillet, 2017). The temporal resolution for MEG is good (on the order of 1 msec), and the transparency of magnetic fields with respect to various tissues (skin, bone, cerebral spinal fluid) provides advantages over electroencephalography (EEG) with respect to source localization. Using an individual specific anatomical MRI to model the brain spatial resolution lower than one cm can be achieved (Sato, Yoshioka, Tkajihara, Toyama, Goda, & Kawato, 2004).
Privacy and Ethics in Brain–Computer Interface Research
Published in Chang S. Nam, Anton Nijholt, Fabien Lotte, Brain–Computer Interfaces Handbook, 2018
Advances in neuroimaging in recent decades helped pave the way for BCI work on decoding brain states. While computed tomography and magnetic resonance imaging have provided increasingly detailed diagnostic information about structural features of the brain, the development of functional magnetic resonance imaging (fMRI) is a critical advance in understanding mental activity. fMRI allows for measurement of brain activity (indirectly through measures of blood flow) during certain mental processes or in conjunction with the experience of certain mental states. This has allowed researchers to determine, for instance, which visual image someone is viewing (even being able to reconstruct the image (Schoenmakers et al. 2013) or what implicit attitudes correlate with moral decision-making (Greene et al. 2001). It has also allowed for measurement of specific intentions (Haynes et al. 2007).
Classification of autism spectrum disorders individuals and controls using phase and envelope features from resting-state fMRI data
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2022
Mahshid Naghashzadeh, Mehran Yazdi, Alireza Zolghadrasli
The traditional procedure for diagnosing Autism is primarily based on narrative interactions between individuals and clinical professionals. These methods, lacking biological evidence, are prone to generate a high disagreement during the diagnosis and require a long period to recognise abnormalities. Functional magnetic resonance imaging (fMRI) has been widely employed to investigate the brain’s functional characteristics to complement the current behaviour-based diagnoses (Guo et al. 2017). fMRI is a neuroimaging procedure utilising magnetic resonance imaging (MRI) technology that measures brain activity by detecting changes associated with a blood oxygen level-dependent (BOLD) signal. The fMRI study’s main approaches are task-related fMRI and resting-state fMRI (Seraj et al. 2019). Resting-state fMRI studies are focused on measuring the correlation between spontaneous activation patterns of brain areas. During a resting-state experiment, subjects are placed into the scanner and asked to close their eyes and to think about nothing in particular, without falling asleep (Van Den Heuvel and Pol 2010). Resting-state fMRI has provided a convenient tool to examine the changes in the intrinsic connectivity of specific regions and networks in Autism and control subjects (Hull et al. 2017). There remains substantial controversy regarding the nature of connectivity impairment in Autism, with researchers arguing in favour of under-connectivity, over-connectivity, or unique patterns of both under-connectivity and over-connectivity depending on the brain region (Ecker et al. 2015).
Working memory network plasticity after exercise intervention detected by task and resting-state functional MRI
Published in Journal of Sports Sciences, 2021
Lina Zhu, Xuan Xiong, Xiaoxiao Dong, Yi Zhao, Adam Kawczyński, Aiguo Chen, Wei Wang
Functional magnetic resonance imaging (fMRI) is a neuroimaging technique that allows the assessment of brain activity correlates by measuring changes in blood oxygenation levels that occur in response to the metabolic requirements of activated neurons. In addition to the identification of brain regions that are involved in specific cognitive processes, fMRI also allows the assessment of the integrity of functional connections between brain regions. The characterization of brain function depends not only on the activation patterns in areas during specific cognitive tasks but also on understanding the connectivity between them. Increasing evidence has indicated that assessment of both task-fMRI and resting-state fMRI (rs-fMRI) is likely to enable a more fine-tuned analysis of brain function by correlating spontaneous neural signals with brain responses to specific stimuli (García-García et al., 2015; Kannurpatti & Biswal, 2008; Kannurpatti et al., 2012; Mastrovito, 2013). While previous observations have attempted to explain the neural basis of exercise on the brain using single MRI methods, the mechanisms assessed by combining task-fMRI and rs-fMRI have not been established.
Multimodal data fusion for systems improvement: A review
Published in IISE Transactions, 2022
Nathan Gaw, Safoora Yousefi, Mostafa Reisi Gahrooei
Let us also consider another data fusion example in healthcare. In a research hospital, there is a cohort of patients with a brain disease along with healthy controls for which three different types of neuroimaging were collected: (i) structural magnetic resonance images (sMRI), (ii) functional magnetic resonance images (fMRI), and (iii) magnetoencephalography (MEG). sMRI conveys brain structure and provides the highest spatial resolution, but has no temporal resolution. fMRI indicates blood oxygen level and provides an acceptable spatial resolution along with a lower temporal resolution. MEG records magnetic fields generated by brain electrical activity, and provides a higher temporal resolution at the cost of a lower spatial resolution. The task is to build a statistical model capable of optimally combining the information available in images A, B, and C to accurately quantify patient disease severity and highlight particular brain locations or functions that may be causing impairment. Early fusion (see Figure 1(a)) can use an algorithm, such as Independent Component Analysis (ICA), to identify spatially independent signals that convey underlying brain networks/structures (f), from which features can be extracted and used to predict disease severity (g). Late fusion (see Figure 1(b)) can develop separate machine learning models trained on each image type, fA, fB, and fC, to make predictions of disease severity, yA, yB, and yC, that can then be fused into an overall prediction score by g (ex., through averaging the scores). Intermediate fusion (see Figure 1(c)) performs neuroimaging fusion during the model training process (e.g., via fused group lasso) and can simultaneously fuse information from images A, B, and C while building an accurate model that can predict the patient disease severity.