September 19, 2016

Stroke Segmentation in CT Scans

Brain Image

Hemorrhage img

Shiny Application shiny orig

flow

flow

(A Lot of) Software Choices for fMRI Analysis

flow

From Carp, Joshua. "The secret lives of experiments: methods reporting in the fMRI literature." Neuroimage 63.1 (2012): 289-300.

Number of Downloads (My CRAN packages)

From the cranlogs R package:

R Package Interfaces with Imaging Software

flow

flow Neuroconductor:
A Neuroimaging R Repository
Hosted on GitHub and
Checked by Travis CI

flow

Neuroconductor Goals (Similar to BioC)

  1. Lower the bar to entry - just have to know R
    • more packages to interface with other programs
  2. All the things a repository has (checks/rules/stability)
    • additional checks with imaging software installed (e.g. examples)
  3. Data packages to test packages with real data (images can be big)
  4. Detailed vignettes/tutorials on how to actually perform an analysis
  5. Image analyses use all the "things R has to offer" (packaging system/reproducibility/etc.)

Under Development R Packages

  1. nitrc - download data from the NITRC repository
  2. MNITemplate\(^*\) - data of a population-level "template" image
  3. EveTemplate\(^*\) - data of a different template image
  4. kirby21 - data package with 2 subjects, 2 visits with multimodal imaging
  5. rcamino - interface to analyze DTI data
  6. msseg - MS lesion segmentation
  7. extrantsr - pipelines for structural imaging analysis

Not started yet

  1. hcp - interface with Human Connectome Project
  2. afnir - R port of AFNI software (No. 2 on the chart)

\(^*\) - working with Jean-Philippe Fortin on these

Neuroconductor flow

Hopeful Solutions

  1. Help improve reproducibility in imaging
  2. Standardize the syntax / R objects for imaging a bit (I'm realistic)
  3. Easily-accessible content

Problems

  1. More control over the workflow = more work (for us!)
  2. Users need external software (versions/installation)
  3. No control over external software
  4. Need the content (buy-in from the community)

Thanks

Current R Capabilities in Imaging

  1. Image I/O
  2. Plotting
  3. Image registration (linear/non-linear)
  4. Image smoothing
  5. Tissue-class segmentation (white/gray/CSF)
  6. Inhomogeneity correction

  7. Intensity normalization

  8. Temporal filtering
  9. Large linear models
  10. Analysis of fMRI task data

  11. Diffusion tensor models