Matlab course Program

1
Matlab course - Program

• Today
• Basic matrix operations
• Basic visualization. Images and plots
• Thursday
• Programming, scripts, more visualization
• Friday
• Graphical user interfaces and advanced
  visualizaton
• Any suggestions wellcome, only today is fairly set

2
Matlab course - introduction

• Goals
• Visualize data
• Develop programs
• GUIs
• Read other peoples programs (eg. SPM)
• Hands on

3
Matlab as a language

• Vector/matrix based
• Many built in functions
• Few lines of code will get you far
• Extensive visualization capabilities
• Might run slower depending on task

4
Lets start

• Manipulating matrices and vectors
• Basically needed for all programming tasks
  everything
• Basic operations
• Add noise to a simulated signal an example
  (hemodynamic response function)

5
(No Transcript)
6
Visualizing data
7
Plotting data

• The plot function
• plot(y)
• plot(x,y)
• X and Y can be matrices
• plot() will plot the columns
• Use hold on to keep the old plot

8
Simulating the HRF
Buxton Introduction to FMRI
hrtvectortor hrt(timevector, tau, t0, n)
9
Excercise

• Part 1
• Generate a timevector, L10 s, interval0.1 s
• Use the hrf() function to generate a hemodynamic
  response
• Use plot to verify the output
• Part 2
• Imagine that you want to display 3 different hrf
  functions in a plot. Generate 3 functions and put
  them in a matrix columnwise. Use a single call to
  plot to view them
• Part 3
• Use randn() to add noise to the curves. Use a std
  of XX
• Plot again

Example hrf parameters timevec from 0 to 20
secs tau 1.2 t0 4 s n 3
Example noise parameters Use a std of 0.1 0.3
- 0.5
10
(No Transcript)
11
Imagesmatrices
12
Imagescaling
Min value
Max value
Full range
Partial range
13
imagesc() colorbar
imagesc(data)
imagesc(data,0 5000)
14
Loading a dicom file

• Matrixdicomread(filename)
• Matrixdouble(Matrix)
• imagesc(Matrix)

15
colormap(cmap_name)
16
hist(data(),bins)
17
Logical comparisons

• Largevaluesmatrix gt 500
• Can be true or false, 0 or 1

18
Roipoly will return a mask
19
Exercise

• Load a dicom into a matrix
• Use hist()( to identify noise and see if there is
  ghosting using imagesc()
• Cut away the noise
• Hint use a comparison to make a noise mask and
  then .
• Use roipoly to mask away 1 hemisphere

20
Birgitte how do we do stats on a mask?
21
Exercise

• Use mean() to get a mean of an ROI selected using
  roipoly
• Plot a histogram of your ROI

22
Saving work

• save workspacename
• save filename X saves only X.
• Save filenename X Y Z saves X, Y, and Z.

23
The online docs

• Very usefull
• Good links
• Good newsgroup
• Tomorrow structures the dir command