A data-driven approach for stratifying psychotic and mood disorders subjects using structural magnitude resonance imaging data

2020 ◽

Vol 29 (06) ◽

pp. 2030001

Author(s):

Abeer M. Mahmoud ◽

Hanen Karamti ◽

Fadwa Alrowais

Keyword(s):

Feature Space ◽

Machine Learning Algorithms ◽

Data Driven ◽

Functional Magnetic Resonance ◽

Resonance Imaging ◽

Brain Disorder ◽

Medical Problems ◽

Discriminative Approach ◽

Large Brain ◽

Data Driven Approach

Functional Magnetic Resonance Imaging (fMRI), for many decades acts as a potential aiding method for diagnosing medical problems. Several successful machine learning algorithms have been proposed in literature to extract valuable knowledge from fMRI. One of these algorithms is the convolutional neural network (CNN) that competent with high capabilities for learning optimal abstractions of fMRI. This is because the CNN learns features similarly to human brain where it preserves local structure and avoids distortion of the global feature space. Focusing on the achievements of using the CNN for the fMRI, and accordingly, the Deep Convolutional Auto-Encoder (DCAE) benefits from the data-driven approach with CNN’s optimal features to strengthen the fMRI classification. In this paper, a new two consequent multi-layers DCAE deep discriminative approach for classifying fMRI Images is proposed. The first DCAE is unsupervised sub-model that is composed of four CNN. It focuses on learning weights to utilize discriminative characteristics of the extracted features for robust reconstruction of fMRI with lower dimensional considering tiny details and refining by its deep multiple layers. Then the second DCAE is a supervised sub-model that focuses on training labels to reach an outperformed results. The proposed approach proved its effectiveness and improved literately reported results on a large brain disorder fMRI dataset.

Download Full-text

Regional Brain Activity Associated with Visual Backward Masking

Journal of Cognitive Neuroscience ◽

10.1162/0898929052880011 ◽

2005 ◽

Vol 17 (1) ◽

pp. 13-23 ◽

Cited By ~ 34

Author(s):

Michael F. Green ◽

David Glahn ◽

Stephen A. Engel ◽

Keith H. Nuechterlein ◽

Fred Sabb ◽

...

Keyword(s):

Visual Processing ◽

Visual Masking ◽

Anterior Cingulate ◽

Data Driven ◽

Motion Processing ◽

Backward Masking ◽

Imaging Data ◽

Neural Basis ◽

Data Driven Approach ◽

Visual Backward Masking

In visual backward masking, the visibility of a briefly presented visual target is disrupted by a mask that is presented shortly thereafter. The goal of the current study was to identify regions in the human cortex that may provide the neural basis of visual masking. We searched for areas whose activity correlated with perception as we systematically varied the strength of masking. A total of 13 subjects performed a backward masking task during functional magnetic resonance imaging. Target and mask were presented at three delay intervals (34, 68, and 102 msec) and behavioral measures confirmed that the targets were more visible at longer masking intervals. Two sets of regions of interest were identified: Distinct regions in the visual cortex (V1/V2, LO, hMT+) were segregated using scans to localize visual processing drawn from the existing literature. Additional cortical regions were selected in a data-driven approach based on their activity during the backward masking task. For each set, we determined the regions whose magnitude of activation increased at longer masking intervals. Nine of the subjects provided valid behavioral performance data on the visual masking task and imaging data from these subjects were used for subsequent analysis. The scans of visual processing areas identified four regions, including: early visual areas (V1 and V2), the motion-sensitive regions in the lateral occipital (LO) lobe (hMT+), and two components (dorsal and ventral) of the object-sensitive region, LO. Of these, the ventral and dorsal LO regions were sensitive to the strength of the mask. For the data-driven approach, six regions were identified on the basis of a difference map in which all masking intervals were contrasted with rest. These included the inferior parietal, anterior cingulate, precentral, insula, thalamic, and occipital areas. The predicted effects of more activity with weaker masking were seen in the thalamus, inferior parietal, and anterior cingulate. This study isolated three types of visual processing areas. The first included regions that subserve key stages of vision (including object and motion processing). The second type responded to the presentation of briefly presented visual stimuli, regardless of masking interval. The third type (selected from the first two) included regions sensitive to the interval between the target and mask. These latter regions (including ventral LO, inferior parietal, anterior cingulate, and thalamus) may form the neural substrate of backward masking.

Download Full-text

A data-driven approach to extract connectivity structures from diffusion tensor imaging data

2015 IEEE International Conference on Big Data (Big Data) ◽

10.1109/bigdata.2015.7363843 ◽

2015 ◽

Cited By ~ 2

Author(s):

Yu Jin ◽

Joseph F. JaJa ◽

Rong Chen ◽

Edward H. Herskovits

Keyword(s):

Diffusion Tensor Imaging ◽

Diffusion Tensor ◽

Data Driven ◽

Imaging Data ◽

Data Driven Approach ◽

Diffusion Tensor Imaging Data

Download Full-text

An image-based data-driven analysis of cellular architecture in a developing tissue

eLife ◽

10.7554/elife.55913 ◽

2020 ◽

Vol 9 ◽

Cited By ~ 4

Author(s):

Jonas Hartmann ◽

Mie Wong ◽

Elisa Gallo ◽

Darren Gilmour

Keyword(s):

Data Science ◽

Image Data ◽

Data Driven ◽

Numerical Representation ◽

Contextual Knowledge ◽

Imaging Data ◽

Cellular Architecture ◽

Posterior Lateral Line ◽

Data Driven Approach ◽

The Rich

Quantitative microscopy is becoming increasingly crucial in efforts to disentangle the complexity of organogenesis, yet adoption of the potent new toolbox provided by modern data science has been slow, primarily because it is often not directly applicable to developmental imaging data. We tackle this issue with a newly developed algorithm that uses point cloud-based morphometry to unpack the rich information encoded in 3D image data into a straightforward numerical representation. This enabled us to employ data science tools, including machine learning, to analyze and integrate cell morphology, intracellular organization, gene expression and annotated contextual knowledge. We apply these techniques to construct and explore a quantitative atlas of cellular architecture for the zebrafish posterior lateral line primordium, an experimentally tractable model of complex self-organized organogenesis. In doing so, we are able to retrieve both previously established and novel biologically relevant patterns, demonstrating the potential of our data-driven approach.

Download Full-text

A Novel Data-Driven Approach to Preoperative Mapping of Functional Cortex Using Resting-State Functional Magnetic Resonance Imaging

Neurosurgery ◽

10.1227/neu.0000000000000141 ◽

2013 ◽

Vol 73 (6) ◽

pp. 969-983 ◽

Cited By ~ 77

Author(s):

Timothy J. Mitchell ◽

Carl D. Hacker ◽

Jonathan D. Breshears ◽

Nick P. Szrama ◽

Mohit Sharma ◽

...

Keyword(s):

Magnetic Resonance Imaging ◽

Magnetic Resonance ◽

Functional Magnetic Resonance Imaging ◽

Resting State ◽

Tumor Resection ◽

Data Driven ◽

Functional Magnetic Resonance ◽

Resonance Imaging ◽

Preoperative Mapping ◽

Data Driven Approach

Abstract BACKGROUND: Recent findings associated with resting-state cortical networks have provided insight into the brain's organizational structure. In addition to their neuroscientific implications, the networks identified by resting-state functional magnetic resonance imaging (rs-fMRI) may prove useful for clinical brain mapping. OBJECTIVE: To demonstrate that a data-driven approach to analyze resting-state networks (RSNs) is useful in identifying regions classically understood to be eloquent cortex as well as other functional networks. METHODS: This study included 6 patients undergoing surgical treatment for intractable epilepsy and 7 patients undergoing tumor resection. rs-fMRI data were obtained before surgery and 7 canonical RSNs were identified by an artificial neural network algorithm. Of these 7, the motor and language networks were then compared with electrocortical stimulation (ECS) as the gold standard in the epilepsy patients. The sensitivity and specificity for identifying these eloquent sites were calculated at varying thresholds, which yielded receiver-operating characteristic (ROC) curves and their associated area under the curve (AUC). RSNs were plotted in the tumor patients to observe RSN distortions in altered anatomy. RESULTS: The algorithm robustly identified all networks in all patients, including those with distorted anatomy. When all ECS-positive sites were considered for motor and language, rs-fMRI had AUCs of 0.80 and 0.64, respectively. When the ECS-positive sites were analyzed pairwise, rs-fMRI had AUCs of 0.89 and 0.76 for motor and language, respectively. CONCLUSION: A data-driven approach to rs-fMRI may be a new and efficient method for preoperative localization of numerous functional brain regions.

Download Full-text

Latent subtypes of manic or irritable episode symptoms in two population-based cohorts

10.1101/2021.04.14.21255394 ◽

2021 ◽

Author(s):

Ryan Arathimos ◽

Chiara Fabbri ◽

Evangelos Vassos ◽

Katrina A S Davis ◽

Oliver Pain ◽

...

Keyword(s):

Mood Disorders ◽

Latent Class ◽

Epidemiological Studies ◽

Data Driven ◽

Risk Scores ◽

Latent Classes ◽

Uk Biobank ◽

Data Driven Approach ◽

Latent Subgroups ◽

Irritable Mood

Background Episodic changes in mood characterise disorders such as bipolar disorder, which includes distinct periods of manic excitability or irritability, along with additional symptoms experienced during these periods. Common clinical understanding informs diagnostic criteria and epidemiological studies reflect clinical thresholds. Aims To use a data-driven approach to defining groupings of symptoms experienced during periods of manic or irritable mood, which could inform understanding of mood disorders and guide case classification by identifying subgroups with homogeneous clinical/functional outcomes. Methods We used latent class analysis (LCA) to conduct an exploration of the latent structure in symptom responses in the UK Biobank and PROTECT studies, by investigating how symptoms, experienced during periods of manic or irritable mood, formed latent subgroups. We tested associations of latent subgroups with sociodemographic characteristics, diagnoses of psychiatric disorders and polygenic risk scores (PRS). Results Five latent classes were identified that captured patterns of symptoms experienced during periods of manic or irritable mood (N=42,183) in UK Biobank. We identified one class that experienced disruptive episodes of mostly irritable mood that was largely comprised of cases of depression/anxiety, and a class of individuals with increased confidence/creativity that reported lower disruptiveness and lower functional impairment. The five latent classes were replicated in an independent cohort, the PROTECT study (N=4,445), with similar distinctions between classes. Conclusion Our data-driven approach to grouping individuals identified distinct latent classes. A dimensional classification of mood disorders informed by our findings will be able to better assess or subtype these disorders in future studies.

Download Full-text