talk

docs/lab-2024-04/lab-2024-04.md

@@ -5,13 +5,15 @@ size: 16:9
 paginate: true
 ---
 
+<!-- _paginate: false -->
+
 <style scoped>
 p {
     font-size: 24px;
 }
 </style>
 
-# Functional Connectivity
+# Functional MRI
 
 ## Jaewon Chung
 
@@ -26,55 +28,234 @@ _Johns Hopkins University - Biomedical Engineering_
 
 ---
 
-# Outline
+# What is Magnetic Resonance Imaging (MRI)?
+
+<br>
+
+![center w:1000](./images/mri-basics.jpeg)
+
+<footer>
+
+[The development of brain magnetic resonance approaches in large animal models for preclinical research
+](https://academic.oup.com/af/article/9/3/44/5522880)
+
+</footer>
+
+---
+
+# MRI Parameters
+
+- _Repetition Time (TR)_: Time between successive RF pulses
+- _Echo Time (TE)_: Time between RF pulse and signal acquisition
+- _Flip Angle_: Angle from initial alignment
+
+  - Controlled by amplitude and duration of RF pulse
+
+- One machine, different sequences, different 'modalities' of images
+
+---
+
+# Examples of Sequences
+
+1. T1-weighted (T1w): Excellent for anatomical detail.
+   - brain, internal organs
+2. T2-weighted (T2w): Sensitive to fluid and inflammation.
+   - cysts, tumors
+3. Gradient Echo (GRE): Sensitive to blood products.
+   - blood flow, hemorrhage, iron deposition
+4. Diffusion-weighted Imaging (DWI): Sensitive to water diffusion.
+   - white matter tracts, early onset stroke
+
+---
+
+# What is Functional MRI (fMRI)?
+
+- Idea: Blood flow increases in active brain regions.
+- Measurement: Blood Oxygenation Level Dependent (BOLD) signal
+- Caveat: fMRI measures **relative** changes in blood flow!
+
+<div class="columns">
+<div>
+
+![center w:350](./images/fmri_axial.gif)
+
+</div>
+
+<div>
+
+![center w:350](./images/fmri.gif)
+
+</div>
+
+</div>
+
+---
+
+# Task vs. Resting-State fMRI
+
+<div class="columns">
+<div>
+
+- Resting-State fMRI: Subjects are at rest, no task
+- Task fMRI: Subjects perform a task (e.g., finger tapping, watching video)
+
+</div>
+
+<div>
+
+![center w:1000](./images/mri-mirror.jpeg)
+
+</div>
+
+</div>
+
+---
+
+# Relationship Between Neural Activity and Blood Flow
+
+<div class="columns">
+<div>
+
+Lots of evidence from animals and humans!
+
+- Local field potentials (LFPs): measures electrical activity of nearby neurons
+- Positive correlation between LFP and BOLD signal
+
+</div>
+
+<div>
+
+![center w:1000](./images/lfp-device.png)
+
+</div>
+</div>
+
+<footer>
+
+[Neurophysiological investigation of the basis of the fMRI signal](https://www.biorxiv.org/content/10.1101/2021.11.01.466686v2)
+[Coupling between neuronal firing rate, gamma LFP, and BOLD fMRI is related to interneuronal correlations](https://doi.org/10.1016/j.cub.2007.06.066)
+
+</footer>
+
+---
+
+# Processing fMRI Data
+
+<br>
+
+![center w:850](./images/fmri-processing.png)
+
+<footer>
+
+[A low-resource reliable pipeline to democratize multi-modal connectome estimation and analysis](https://www.biorxiv.org/content/10.1101/2021.11.01.466686v2)
 
-1. What is fMRI?
-2. How fMRI Works
-3. fMRI Applications
-4. Data processing
+</footer>
 
 ---
 
-# What is fMRI?
+# Preprocessing Steps
+
+- Slice time correction
+  - Corrects for differences in acquisition time between slices
+  - Typically interpolated to the middle slice
+
+<br>
+
+![center w:850](./images/slice-time.png)
 
 ---
 
-# How fMRI Works
+# Preprocessing Steps
+
+- Motion Correction: Aligns all volumes (timepoints) to a reference image
+  - 1st image: Reference image
+  - Motion parameters: translation and rotation
+  - Some volumes may be discarded.
 
 ---
 
-# fMRI Applications
+# Registration
+
+<div class="columns">
+<div>
+
+1. Coregistration: Aligns the fMRI to the subject's high-resolution anatomical scan.
+2. Normalization: Aligns subject's anatomical scan to standard space (e.g., MNI152 space) for group comparisons.
 
-fMRI is widely used to study:
-Cognitive processes (memory, language, decision-making)
-Sensory and motor functions
-Brain development
-Neurological and psychiatric conditions
+</div>
+
+<div>
+
+![center w:350](./images/fmri_axial.gif)
+
+</div>
+
+</div>
 
 ---
 
-# Resting-State fMRI
+# Nuisance Correction
 
-Subjects are asked to stay still and relaxed, often with their eyes closed.
-fMRI scans measure the brain activity that occurs naturally at rest.
-Functional connectivity analyses reveal intrinsic brain networks present even without a specific task.
+1. Bandpass filtering
+   - Filtering out low and high frequencies
+1. Cyclical detrending
+   - Removing linear and quadratic trends
+1. CompCor
+   - Removing physiological noise (e.g., cardiac and respiratory signals)
+   - Noise estimated from white matter and cerebrospinal fluid
 
 ---
 
 # What is Functional Connectivity?
 
-Functional connectivity examines how different brain regions are temporally correlated.
-Regions that fluctuate in activity together are considered functionally 'connected.'
-This reflects networks of brain areas working in concert.
+<div class="columns">
+<div>
+
+### Estimation
+
+- Apply a parcellation
+- Average BOLD signal within each region
+- Compute correlation between regions
+
+### Interpretation
+
+- Meaures how brain regions are temporally correlated.
+- Regions that fluctuate in together = functionally 'connected'
+
+</div>
+
+<div>
+
+![center w:350](./images/dkt.png)
+
+</div>
+</div>
 
 ---
 
-# Functional Connectivity Applications
+# Anatomy of Functional Connectome
+
+<div class="columns">
+<div>
+
+- Ipsilateral: within the same hemisphere
+- Contralateral: between hemispheres
+- Homotopic: between mirror areas
+- High correlation in homotopic connections
+
+</div>
+
+<div>
 
-Mapping healthy brain networks
-Investigating brain development across the lifespan
-Identifying network changes in neurological and psychiatric conditions
-Potential for clinical diagnosis and tracking treatment effectiveness
+![center w:1000](./images/functional-connectome.png)
+
+</div>
+
+</div>
+
+---
+
+# Default Mode
 
 ---
 
@@ -101,9 +282,18 @@ Functional Connectivity Analysis: Researchers calculated correlations in brain a
 
 ---
 
-# Limitations
+# Structure and Function
+
+---
+
+# Useful Links
+
+- [MRI-to-graphs (m2g)](https://github.com/neurodata/m2g) - Our fMRI and dMRI pipeline for estimating connectomes
+- [MRI Connectomes](https://neurodata.io/mri/) - Our processed connectomes using open-access datasets
+- [Slides on Diffusion MRI](https://neurodata.io/talks/ndmg.pdf) - A brief overview of diffusion MRI pipeline in m2g.
+
+---
 
-fMRI measures blood flow (an indirect measure of neural activity)
-Temporal resolution is relatively low (seconds) compared to the speed of brain processes (milliseconds).
+<br><br><br><br><br>
 
-- Participants have to lie completely still while in an fMRI.
+<p><h1 align="center">Questions?</p>