CS 112

Assignment 8

Due: Friday, April 27, at the beginning of class

You can turn in your assignment up until 5:00pm on 4/27/07 without penalty, but it is best to hand in the assignment at the beginning of class. Your hardcopy submission should include a cover sheet and a printout of mriGUI.m. Your electronic submission is described in the section Uploading your saved work

Reading

The following material is especially useful to review for this assignment: notes and examples from Lectures #13-14, 19 and 21, and Lab #8 and 11. There is also extensive online documentation of the 3D visualization of graphics objects.

Getting Started: Create an assign8_programs folder

There are no initial programs needed for this assignment, as you will be creating your GUI-based program from scratch. Create a folder for your programs that includes assign8_programs in the name, e.g. sohie_assign8_programs. In MATLAB, set the Current Directory to your assign8_programs folder.

Uploading your saved work

Use Fetch or WinSCP to upload your saved work, and connect to the CS file server using your personal user account name and password. After logging in to your account:

When you are done with this assignment, you should have five (5) files stored in your assign8_programs folder: mriGUI.m, mriGUI.fig, mri.mat, transformView.m and loadinData.m.

Exercise: Create a GUI for your MRI program

Recall the text-based interactive program you wrote for Assignment 5 that allowed the user to choose various MRI slice views. In this assignment, you will create a GUI for that program, and add some more features to your program.

The demo below shows one possible interaction in MATLAB's command window of how the Assignment 5 program may have worked: 

 >> mri
    Welcome to my program!  
    Enter 1 to show all images at once 
          2 to display a horizontal slice
          3 to show all frames as a movie
          4 to display a sagittal slice
          5 to display a coronal slice 
          e to exit
   ==> 1
     Enter 1(all), 2(horiz.), 3(movie), 4(sag.), 5(cor.) or [e]xit
   ==> 2
    Select a horizontal slice bottom->top [1-27]: 15

     Enter 1(all), 2(horiz.), 3(movie), 4(sag.), 5(cor.) or [e]xit
   ==> 3
     Enter 1(all), 2(horiz.), 3(movie), 4(sag.), 5(cor.) or [e]xit
   ==> 4
     Select a sagittal slice left->right [1-128]: 80

     Enter 1(all), 2(horiz.), 3(movie), 4(sag.), 5(cor.) or [e]xit
   ==> 5
     Select a coronal slice front->back [1-128]: 75
   ==> e
>>
>>

Your mriGUI for this assignment might look something like this when complete:
Figure 1

The diagram above is meant only as a guideline. You can choose layout and colors as you see fit. You must, however, adhere to these requirements:

mriGUI requirements:

  1. The GUI must allow for the following displays:
  2. Each of the displays listed above in Requirement #1 must have a title (e.g. Saggital slice 25)
  3. If the slice number chosen for a particular view is out of range, the GUI must insist on correct input rather than attempting an erroneous display. (click here for sample screenshot)
  4. The GUI must have a colormap menu with at least five different colormap options. Upon selection of a colormap, the current display reflects the newly chosen colormap. If you would like to, you can create your own colormap and use that as one of the five choices (but this is not required).
  5. The GUI must allow the user to choose 4 horizontal slices from a listbox that are displayed as contour maps simultaneously. (click here for sample screenshot)
  6. The GUI must create a 3D head view where the user can specify the azimuth and elevation of the viewpoint with sliding bars. (click here for sample screenshots)
  7. The GUI must contain an exit or close button.
  8. Some of the code for your mriGUI may be copied directly from your Assignment 5 solutions (you can use your own code, or the solutions distributed in lab).

More information on the new features

Viewing four contour slices at the same time

Viewing four contour slices at the same time requires that the user can select multiple options from a listbox. In MATLAB, this is achieved by setting the min and max values such that (max-min)>1 . The min and max values can be set with the Property Inspector when creating the GUI with GUIDE. Contour slices are generated using the following MATLAB code:

newD = squeeze(D);
plothandle = contourslice(newD,[], [], [slice1, slice2, slice3, slice4], 8);
view(3);
set(plothandle,'LineWidth',2);

First, the 128x128x1x27 matrix D is squeezed into a 128x128x27 matrix newD. The single dimension of the original D matrix that is typically used for color data is dropped out. We'll use colormaps to include color in our plots. Next, the built-in MATLAB function contourslice is called to show the four slices slice1,slice2,slice3 and slice4 of the newD matrix. The last argument to contourslice, 8, specifies the number of contour lines per plane. The plot is shown using MATLAB's default 3D view, which is azimuth = -37.5 and elevation = 30. Finally, the linewidth is set to be thicker so that the contour lines are more easily seen.
In your mriGUI, you will read slice1, slice2, slice3 and slice4 from the listbox and then use the code above to display the slices simultaneously.

Viewing the 3D head with user-specified viewpoint

The mri data is squeezed, as it was in the contour plots above:

newD = squeeze(D);

and then the data are smoothed with smooth3 and then isosurface calculates the isodata. Patch displays this data as a graphics object.

Ds = smooth3(newD);
hiso = patch(isosurface(Ds,5),...
   'FaceColor',[1, 0.75, 0.65],'EdgeColor','none');

To show a sliced-away top of the head, we use isocaps to calculate the data for another patch that is shown at the same isovalue (5) as the surface. We use the original unsmoothed (yet still squeezed) data D to see the details of the interior.

hcap = patch(isocaps(newD,5),...
  'FaceColor','interp','EdgeColor','none');

We do some fine-tuning by setting the view, aspect ratio and lighting:

daspect([1,1,0.4]);
lightangle(45,30);
lighting phong;
isonormals(Ds, hiso);
set(hcap,'AmbientStrength', 0.6);
set(hiso,'SpecularColorReflectance',0,'SpecularExponent',50);

User-specified viewpoint

MATLAB tracks view using azimuth and elevation. azimuth is a polar angle in the x-y plane, with positive angles indicating counterclockwise rotation of the viewpoint (see diagram below). Elevation is the angle above (positive angle) or below (negative angle) the x-y plane.

Here is an example of viewing an object from directly overhead:

az = 0;
el = 90;
view(az, el);

In MATLAB, when creating the azimuth and elevation slider components, you can specify the min and max values using the Property Inspector in GUIDE. A reasonable range might be [0,360] for azimuth and [-90,90] for elevation. Then, your program can read the values from the azimuth and elevation sliders and use those values to update your view. This means that azimuth and elevation need to be in the handles structure of your GUI program so that they can be accessed by other components.

Notes:

  1. You may find panels and alignment useful in GUIDE.
  2. Remember to modify your mriGUI_OpeningFcn function to construct the fields of the handles structure that stores all information that is shared across GUI components.
  3. You can run loadinData.m from MATLAB's command window to create the mri.mat file, then you include load mri; in the mriGUI_OpeningFcn.
  4. Recall that every function that modifies the value of a field of the handles structure must include the following statement at the end:

        guidata(hObject,handles)
  5. You can also include printing statements by omitting the semi-colon at the end of any statement that produces a value, or adding calls to the disp function. These values will be printed in the Command Window.

Electronic submission

Be sure to include both the mriGUI.m and mriGUI.fig files in the assign8_programs folder that you submit electronically.