Using Contiguous Bi-Clustering for data driven temporal analysis of fMRI based functional connectivity.

Slides:



Advertisements
Similar presentations
FMRI Methods Lecture 10 – Using natural stimuli. Reductionism Reducing complex things into simpler components Explaining the whole as a sum of its parts.
Advertisements

Hierarchical Clustering. Produces a set of nested clusters organized as a hierarchical tree Can be visualized as a dendrogram – A tree-like diagram that.
Comp 122, Spring 2004 Greedy Algorithms. greedy - 2 Lin / Devi Comp 122, Fall 2003 Overview  Like dynamic programming, used to solve optimization problems.
Greedy Algorithms Greed is good. (Some of the time)
1 EE5900 Advanced Embedded System For Smart Infrastructure Static Scheduling.
Visualization of dynamic power and synchrony changes in high density EEG A. Alba 1, T. Harmony2, J.L. Marroquín 2, E. Arce 1 1 Facultad de Ciencias, UASLP.
Segmentácia farebného obrazu
Clustering & image segmentation Goal::Identify groups of pixels that go together Segmentation.
Automatic Identification of ROIs (Regions of interest) in fMRI data.
Clustering II.
Thresholding Otsu’s Thresholding Method Threshold Detection Methods Optimal Thresholding Multi-Spectral Thresholding 6.2. Edge-based.
Mutual Information Mathematical Biology Seminar
International Workshop on Computer Vision - Institute for Studies in Theoretical Physics and Mathematics, April , Tehran 1 IV COMPUTING SIZE.
Aki Hecht Seminar in Databases (236826) January 2009
© University of Minnesota Data Mining for the Discovery of Ocean Climate Indices 1 CSci 8980: Data Mining (Fall 2002) Vipin Kumar Army High Performance.
Advanced Topics in Data Mining Special focus: Social Networks.
Complex Model Construction Mortenson Chapter 11 Geometric Modeling
Systems Biology, April 25 th 2007Thomas Skøt Jensen Technical University of Denmark Networks and Network Topology Thomas Skøt Jensen Center for Biological.
Ensemble Learning (2), Tree and Forest
Chapter 3: Cluster Analysis  3.1 Basic Concepts of Clustering  3.2 Partitioning Methods  3.3 Hierarchical Methods The Principle Agglomerative.
Clustering Unsupervised learning Generating “classes”
Escaping local optimas Accept nonimproving neighbors – Tabu search and simulated annealing Iterating with different initial solutions – Multistart local.
Physical Mapping of DNA Shanna Terry March 2, 2004.
MATISSE - Modular Analysis for Topology of Interactions and Similarity SEts Igor Ulitsky and Ron Shamir Identification.
Efficient Gathering of Correlated Data in Sensor Networks
1Ellen L. Walker Segmentation Separating “content” from background Separating image into parts corresponding to “real” objects Complete segmentation Each.
Report on Intrusion Detection and Data Fusion By Ganesh Godavari.
A Clustering Algorithm based on Graph Connectivity Balakrishna Thiagarajan Computer Science and Engineering State University of New York at Buffalo.
Using Graph Theory to Study Neural Networks (Watrous, Tandon, Conner, Pieters & Ekstrom, 2012)
2004 All Hands Meeting Analysis of a Multi-Site fMRI Study Using Parametric Response Surface Models Seyoung Kim Padhraic Smyth Hal Stern (University of.
Course 13 Curves and Surfaces. Course 13 Curves and Surface Surface Representation Representation Interpolation Approximation Surface Segmentation.
Chapter 10, Part II Edge Linking and Boundary Detection The methods discussed in the previous section yield pixels lying only on edges. This section.
Data Extraction using Image Similarity CIS 601 Image Processing Ajay Kumar Yadav.
CS654: Digital Image Analysis
CS 8751 ML & KDDData Clustering1 Clustering Unsupervised learning Generating “classes” Distance/similarity measures Agglomerative methods Divisive methods.
Category Independent Region Proposals Ian Endres and Derek Hoiem University of Illinois at Urbana-Champaign.
Course14 Dynamic Vision. Biological vision can cope with changing world Moving and changing objects Change illumination Change View-point.
Network Theory: Community Detection Dr. Henry Hexmoor Department of Computer Science Southern Illinois University Carbondale.
Image Segmentation Nitin Rane. Image Segmentation Introduction Thresholding Region Splitting Region Labeling Statistical Region Description Application.
Thresholding Foundation:. Thresholding In A: light objects in dark background To extract the objects: –Select a T that separates the objects from the.
Network Partition –Finding modules of the network. Graph Clustering –Partition graphs according to the connectivity. –Nodes within a cluster is highly.
Selected Topics in Data Networking Explore Social Networks:
Visualization of Biological Information with Circular Drawings.
Meeting 8: Features for Object Classification Ullman et al.
Color Image Segmentation Mentor : Dr. Rajeev Srivastava Students: Achit Kumar Ojha Aseem Kumar Akshay Tyagi.
Asstt. Professor Adeel Akram. Other Novel Routing Approaches Link reversal Aimed for highly dynamic networks Goal: to identify some path, as opposed.
Machine Vision ENT 273 Lecture 4 Hema C.R.
Unsupervised Learning: Clustering
Unsupervised Learning: Clustering
Chapter 5. Greedy Algorithms
Computer Vision Lecture 13: Image Segmentation III
Data Mining K-means Algorithm
Mean Shift Segmentation
Computer Vision Lecture 12: Image Segmentation II
Community detection in graphs
3.1 Clustering Finding a good clustering of the points is a fundamental issue in computing a representative simplicial complex. Mapper does not place any.
Davide Nardo, Valerio Santangelo, Emiliano Macaluso  Neuron 
Chao Zhang1, Yu Zheng2, Xiuli Ma3, Jiawei Han1
Decoding Wakefulness Levels from Typical fMRI Resting-State Data Reveals Reliable Drifts between Wakefulness and Sleep  Enzo Tagliazucchi, Helmut Laufs 
Kruskal’s Minimum Spanning Tree Algorithm
Graph Theoretic Analysis of Resting State Functional MR Imaging
The Development of Human Functional Brain Networks
Digital Image Processing
Volume 74, Issue 4, Pages (May 2012)
Intrinsic and Task-Evoked Network Architectures of the Human Brain
Between Thoughts and Actions: Motivationally Salient Cues Invigorate Mental Action in the Human Brain  Avi Mendelsohn, Alex Pine, Daniela Schiller  Neuron 
Figure 1 Lesion locations causing cervical dystonia
Hierarchical Clustering
The Development of Human Functional Brain Networks
Presentation transcript:

Using Contiguous Bi-Clustering for data driven temporal analysis of fMRI based functional connectivity

Background The brain works in a hierarchical manner Neurons-> (voxels->)regions->network-> function Who plays with whom when ?

Existing methods Data driven: PCA/ICA – requires large number of time points, performs a partition (finds disjoint components) over all time points, hierarchical clustering. Hypothesis driven seed based/ generate network using anatomically defined regions. Problems: may not be accurate, can miss new information (e.g. sub-regional activity).

Advantages: There may be a partial overlap between two networks. Can reveal connections that appear in a subset of the TRs. Bi-clustering methods tend to yield many results, from which it is hard to extract the biologically significant ones. Disadvantages:

Suggested Model Based on the prior knowledge that we have on the way that the brain works, we can assume: A relatively low number of regions A relatively low number of consecutive time point series.

Suggested method - heuristic Perform analysis separately for each hemisphere Use a sliding time window (currently 10 time pts). For each window detect highly homogenous regions (neighboring voxel groups over size s). Merge regions/groups with high spatial overlap (time intervals either overlapping or disjoint). Filter regions Merge spatially disjoint regions/groups with highly similar signal (and high temporal overlap).

Identifying homogenous regions Input: Graph G=(V,E) where V is the voxel set and E is the adjacency structure between them. General goal: we want to find sets V’  V and for each set an accompanying interval I= [i,j] such that G[V’] is connected and possibly of small diameter The elements in V’ are similar when comparing their subvectors in the time interval I

Identifying homogenous regions For a given time interval I: Identify a set of seed nodes v* 1 ….v* n For each v* search for group U by scanning the neighbors of v* in BFS and add to U each vertex v that satisfies sim(v,v*)≥α or sim(v,sig U ) in interval I. Stop when no vertex can be added to U or when the BFS depth is > d Accept the set U if it is sufficiently large. Addition: merge U 1 and U 2 if they overlap and signal is similar enough (sim(sig U1,sig U2 ) ≥ α )

Selecting seed nodes Seed nodes are selected using a variation of the “Furthest point” heuristic combined with the requirement of a seed to be “interesting” enough. I.E. a list of “interesting” seed candidates is generated. At least k (=2) extreme values of mean±2stds. At each step the candidate which is furthest from all previously selected seeds is chosen.

Evaluating regions * Automated Anatomical Labeling (AAL) is a digital atlas of the human brain (partition). It is developed by a French research group based in Caen. Caen Size Shape/clustering coefficient Repetition (appearance in multiple time windows) OR TR span. Homogeneity (mean pairwise similarity) Anatomic mapping homogeneity the fraction of the region that is mapped(*) to the same anatomic label. Tracing dynamics Comparing with paradigm (if exists)

Validations Comparing hemispheres Analysis on scrambled data (circular shift maintains signal properties). Signal dynamics analysis on moved groups

Data 4 foot motor datasets, 65 TRs (TR=2.5sec) Rest 10TR Right 10TR Rest 5TR Left 10TR Rest 5TR Left 10TR Right 10TR Rest 5TR Rest 5TR 5 palm motor datasets 117TRs, 108TRs

Core regions SubjectCorrelation Thresh #time windows #groups (size>=10) #group with area rep>=0.5 Relevant motor areas BEKI / /931+/+ BEKI /393351/306+/+ MAWE / /549+/+ MAWE /368210/+/ OFPE / /885+/+ OFPE /507296/348+/+ SHEL /785635/519+/+ SHEL /359296/265+/+ SHRO / /707+/+* SHRO /433241/359+*/+

Merging regions based on spatial overlap: 1) Find all group pairs with spatial jaccard above a threshold 2) Select the pair with the highest jac (greedy) 3) Extract the spatial intersection 4) Expand the group using BFS (under time intervals union) 5) Accept new group if it is large enough 6) Repeat steps 1-5 until no pairs are found

Intermediate Results: Correlation threshold Jaccard threshold Initial #groups Resulting #groups % of input %of groups with split times Mean homogeneity %~6-16% %~1-5% %~1-2% %~0-0.6% %~4-14% %~1-4.4% %~0-3% %~0-1%0.79 * Cases in which one group contains another are missed.

Intermediate Results – tracing group dynamics Subject = BEKI ; Hemisphere = Left; Homogeneity = 0.706; size = 61 voxels; time intervals=42:51,84:97,105:114; area=motor; spatial center=(9,36,32) Signal Correlation Right hand Left hand

Intermediate Results – tracing group dynamics Signal Correlation Right hand Left hand Subject = BEKI ; Hemisphere = Left; Homogeneity = 0.737; size = 18 voxels; time intervals=84:97,105:114; area=motor; spatial center=(8,36,32)

Intermediate Results – tracing group dynamics Signal Correlation Right hand Left hand Subject = SHEL ; Hemisphere = Left; Homogeneity = 0.737; size = 18 voxels; time intervals=84:97,105:114; area=motor; spatial center=(8,36,32)

Comparing group dynamics to moved group dynamics Real group Moved group

Comparing group dynamics to moved group dynamics Real group Moved group

Real group Moved group Comparing group dynamics to moved group dynamics

Real group Moved group Comparing group dynamics to moved group dynamics

Defining region groups based on signal similarity 1) Find all group pairs with TR jaccard>thresh 2) Select the pair (g1,g2) with the highest signal correlation under TR intersection (#TRs>thresh, cor>thresh) 3) Extract a group that is composed of (g1Ug2) voxels under the intersection of time points 4) Expand new group time intervals while maintaining homogeneity above a threshold 5) Accept new group if it is large enough 6) Repeat steps 1-5 until no group pairs are found

Visualizations

Visualization – 3D

Open issues: How to merge separate regions Jaccard threshold When and how to filter regions Problem with our initial assumption