Machine Learning SIG: Difference between revisions

From MEG Core
Jump to navigation Jump to search
Content added Content deleted
Line 19: Line 19:
==Analysis Types==
==Analysis Types==
#Decoding
#Decoding
##Multivariate time series classification between conditions <ref>[https://hal.archives-ouvertes.fr/hal-01848442/document Intro1] </ref><ref>[https://direct.mit.edu/jocn/article-abstract/29/4/677/28605/Decoding-Dynamic-Brain-Patterns-from-Evoked?redirectedFrom=fulltext Intro2] </ref>
##Multivariate time series classification between conditions <ref>[https://hal.archives-ouvertes.fr/hal-01848442/document ] Encoding and Decoding Neuronal Dynamics:
Methodological Framework to Uncover the Algorithms of Cognition </ref><ref>[https://direct.mit.edu/jocn/article-abstract/29/4/677/28605/Decoding-Dynamic-Brain-Patterns-from-Evoked?redirectedFrom=fulltext ] Decoding Dynamic Brain Patterns from Evoked Responses: A Tutorial on Multivariate Pattern Analysis Applied to Time Series Neuroimaging Data </ref>
###Filtering Issues<ref>[https://www.sciencedirect.com/science/article/pii/S0165027021000157] High-pass filtering artifacts in multivariate classification of neural time
###Filtering Issues<ref>[https://www.sciencedirect.com/science/article/pii/S0165027021000157] High-pass filtering artifacts in multivariate classification of neural time
series data</ref>
series data</ref>
Line 26: Line 27:
##eg. Healthy Control vs Major Depressive Disorder
##eg. Healthy Control vs Major Depressive Disorder
##What are the significant features (brain regions, Hz)
##What are the significant features (brain regions, Hz)
##Automated diagnosis of TBI from MEG low frequency signals <ref>[https://www.sciencedirect.com/science/article/pii/S1053811912004259]</ref>
##Automated diagnosis of TBI from MEG low frequency signals <ref>[https://www.sciencedirect.com/science/article/pii/S1053811912004259] An automatic MEG low-frequency source imaging approach for detecting injuries in mild and moderate TBI patients with blast and non-blast causes</ref>
#Prediction of future condition / Biomarkers
#Prediction of future condition / Biomarkers
##What signals predict conversion from mild cognitive impairment to Alzheimer
##What signals predict conversion from mild cognitive impairment to Alzheimer

Revision as of 16:44, 8 November 2021

Objectives

Advance general knowledge of machine learning techniques within the MEG community. Discuss journal articles, replicate techniques on NIH data, develop new ML techniques at NIH.

Format

  1. Specific Projects (Weekly)
    1. Code Review
    2. Project updates
    3. Questions and Answers Clinic for users
  2. General (Monthly)
    1. Journal Club
    2. Hackathons – implement novel technique from JC with provided data
    3. Tutorial Workshops - instruct worked out examples with provided code/data
    4. General ML training
      1. Parameter tuning
      2. Model optimization
      3. Techniques
      4. Toolbox tutorials (Scikit-learn / keras)

Analysis Types

  1. Decoding
    1. Multivariate time series classification between conditions [1][2]
      1. Filtering Issues[3]
    2. Realtime - Brain computer interface / neurofeedback
  2. Subject classification –
    1. eg. Healthy Control vs Major Depressive Disorder
    2. What are the significant features (brain regions, Hz)
    3. Automated diagnosis of TBI from MEG low frequency signals [4]
  3. Prediction of future condition / Biomarkers
    1. What signals predict conversion from mild cognitive impairment to Alzheimer
    2. What signals predict recovery from traumatic brain injury
    3. What signals predict poor outcome for epilepsy surgery
  4. Multimodal Integration
    1. Timing derived from MEG / localization from fMRI – joint ICA (??)
    2. Signal comparison between naturalistic viewing data MEG/fMRI
  5. Signal classification
    1. Artifact
    2. Signal of interest – Spike detection for epilepsy
    3. Signal/sequence predicts correct response
  6. Temporal learning models – RNN DL / markov model
    1. Inferences from deep learning models
    2. CNN on images vs. CNN of brain responses to same images

Preliminary Resources

MNE Python Decoding MNE Python Decoding at Source
SciKit Learn
Keras

Relevant Papers

Intro1

Intro2

Language Decoding

  1. ↑ [1] Encoding and Decoding Neuronal Dynamics: Methodological Framework to Uncover the Algorithms of Cognition
  2. ↑ [2] Decoding Dynamic Brain Patterns from Evoked Responses: A Tutorial on Multivariate Pattern Analysis Applied to Time Series Neuroimaging Data
  3. ↑ [3] High-pass filtering artifacts in multivariate classification of neural time series data
  4. ↑ [4] An automatic MEG low-frequency source imaging approach for detecting injuries in mild and moderate TBI patients with blast and non-blast causes
Retrieved from "https://megcore.nih.gov/index.php?title=Machine_Learning_SIG&oldid=4709"