pet_plot_brainlv_3v.m

绝对经典,老外制作的功能强大的matlab实现PLS_TOOBOX

function pet_plot_brainlv_3v(ViewBootstrapRatio, PLSresultFile, ...
	grp_idx, lv_idx, new_fig, behav_idx, cluster_info, update, isstruct)

    if ~exist('isstruct','var')
       isstruct = 0;
    end

    if (new_fig)
        bg_img = getappdata(gcbf,'BackgroundImg');
        rot_amount = getappdata(gcbf,'RotateAmount');

        if ViewBootstrapRatio == 1
           blv = getappdata(gcbf,'BSRatio');
        elseif ViewBootstrapRatio == 2
           blv = getappdata(gcbf,'BLVData');
           bs = getappdata(gcbf,'BSRatio');
	elseif ViewBootstrapRatio == 0
           blv = getappdata(gcbf,'BLVData');
	end
    else
        bg_img = getappdata(gcf,'BackgroundImg');
        rot_amount = getappdata(gcf,'RotateAmount');

        if ViewBootstrapRatio == 1
           blv = getappdata(gcf,'BSRatio');
        elseif ViewBootstrapRatio == 2
           blv = getappdata(gcf,'BLVData');
           bs = getappdata(gcf,'BSRatio');
	elseif ViewBootstrapRatio == 0
           blv = getappdata(gcf,'BLVData');
	end
    end

    if ~isempty(behav_idx)			% called from datamatcorrs plot
        blv = blv{grp_idx, behav_idx};
    end

    if isempty(bg_img)
       no_background_image = 1;
    else
       no_background_image = 0;
    end

%    load('pet_map');				% newcolor: just commented

    if ~isempty(behav_idx)			% called from datamatcorrs plot

        load(PLSresultFile,'num_cond_lst','dims','newcoords','s','voxel_size','origin','behavname');
        num_behav = length(behavname);

    else
        load(PLSresultFile,'num_cond_lst','dims','newcoords','s','voxel_size','origin');
    end

    num_conditions = num_cond_lst(1);

    num_slices = dims(4);
    slice_idx = [1:num_slices];

    if grp_idx
       num_lv = num_conditions;
    else
       num_lv = size(blv,2);
    end

    brainlv = blv(:,lv_idx);

    if ViewBootstrapRatio == 1
       h = findobj(gcf,'Tag','BSThreshold'); thresh = str2num(get(h,'String'));
       h = findobj(gcf,'Tag','MaxRatio'); max_blv = str2num(get(h,'String'));
       h = findobj(gcf,'Tag','MinRatio'); min_blv = str2num(get(h,'String'));
    elseif ViewBootstrapRatio == 2
       h = findobj(gcf,'Tag','Threshold'); thresh = str2num(get(h,'String'));
       h = findobj(gcf,'Tag','MaxValue'); max_blv = str2num(get(h,'String'));
       h = findobj(gcf,'Tag','MinValue'); min_blv = str2num(get(h,'String'));

       h = findobj(gcf,'Tag','BSLVIndexEdit'); bs_lv_idx = str2num(get(h,'String'));
       h = findobj(gcf,'Tag','BSThreshold'); bs_thresh = str2num(get(h,'String'));

       bs = bs(:, bs_lv_idx);
       bs_strong = zeros(size(bs));
       bs_idx = [find(bs <=- bs_thresh); find(bs >= bs_thresh)];
       bs_strong(bs_idx) = 1;
       brainlv = brainlv .* bs_strong;
    elseif ViewBootstrapRatio == 0
       h = findobj(gcf,'Tag','Threshold'); thresh = str2num(get(h,'String'));
       h = findobj(gcf,'Tag','MaxValue'); max_blv = str2num(get(h,'String'));
       h = findobj(gcf,'Tag','MinValue'); min_blv = str2num(get(h,'String'));
    end

    too_large = find(brainlv > max_blv); brainlv(too_large) = max_blv;
    too_small = find(brainlv < min_blv); brainlv(too_small) = min_blv;

    if mod(rot_amount,2)
        img_height = dims(2);		% rows - after 90 or 270 rotation
        img_width  = dims(1);		% by default, 90 rotation
    else
        img_height = dims(1);		% rows
        img_width  = dims(2);
    end

    %  display the images
    %
    if (new_fig)
        [axes_hdl,colorbar_hdl] = pet_create_newblv_ui;
    else
%        axes_hdl = getappdata(gcf,'BlvAxes');
        colorbar_hdl = getappdata(gcf,'Colorbar');	% newcolor: was commented
    end

%    axes(axes_hdl);

    rows = dims(1);
    cols = dims(2);
    slices = dims(4);

    % create the appropriate colormap
    %
    % cmap = set_colormap(max_blv, min_blv, thresh);
   
    bg_values = [1 1 1];
    num_blv_colors = 25;
    brain_region_color_idx = 51;
    first_lower_color_idx = 101;
    first_upper_color_idx = 126;

    % set up the colormap for the background 
    %
    bg_brain_values = [0.54 0.54 0.54];
    if (no_background_image),
       bg_cmap = ones(100,1)*bg_brain_values;	% the brain regions
    else
       bg_cmap = bone(140);
       bg_cmap = bg_cmap(1:100,:);
    end;


    %  colormap entries
    %     	 1 - 100    : for the brain regions (background) image
    %           101 - 125    : for the negative blv values below threshold
    %           126 - 150    : for the positive blv values above threshold
    %     	  151       : for the non-brain regions

    cmap = zeros(151,3);
    jetmap = jet(64);
    cmap(1:100,:) = bg_cmap;			% the brain regions
    cmap(101:125,:) = jetmap([1:25],:);		% the negative blv values
    cmap(126:150,:) = jetmap([36:60],:);	% the positive blv values
    cmap(end,:) = bg_values;			% the nonbrain regions


    %  set up the colormap for the display colorbar
    %
    cbar_size = 100;
    cbar_map = ones(cbar_size,1) * bg_brain_values; 
    cbar_step = (max_blv - min_blv) / cbar_size;

    %  prevent_num_lower_color_0
    %
    if (abs(min_blv) - thresh) < cbar_step & (abs(min_blv) - thresh) ~= 0
        cbar_size = ceil((max_blv - min_blv) / (abs(min_blv) - thresh));
        cbar_map = ones(cbar_size,1) * bg_brain_values;
        cbar_step = (max_blv - min_blv) / cbar_size;
    end
    if (abs(max_blv) - thresh) < cbar_step & (abs(max_blv) - thresh) ~= 0
        cbar_size = ceil((max_blv - min_blv) / (abs(max_blv) - thresh));
        cbar_map = ones(cbar_size,1) * bg_brain_values;
        cbar_step = (max_blv - min_blv) / cbar_size;
    end

  if cbar_step ~= 0
%    num_lower_color = round((abs(min_blv) - thresh) / cbar_step);

      if max_blv > -abs(thresh)
         num_lower_color = round((abs(min_blv) - thresh) / cbar_step);
      else
         num_lower_color = round((abs(min_blv) - abs(max_blv)) / cbar_step);
      end

    if round(64 / 25 * num_lower_color) > 0
       jetmap = jet(round(64 / 25 * num_lower_color));
       cbar_map(1:num_lower_color,:) = jetmap(1:num_lower_color,:);	
    end

%    num_upper_color = round((max_blv - thresh) / cbar_step);

      if min_blv < abs(thresh)
         num_upper_color = round((max_blv - thresh) / cbar_step);
      else
         num_upper_color = round((max_blv - min_blv) / cbar_step);
      end

    if round(64 / 25 * num_upper_color) > 0
       jetmap = jet(round(64 / 25 * num_upper_color));
       first_jet_color = round((36 / 64) * size(jetmap,1));
       jet_range = [first_jet_color:first_jet_color+num_upper_color-1];
       cbar_map(end-num_upper_color+1:end,:) = jetmap(jet_range,:);
    end

    % Create the image slices in which voxels are set to be within certain range
    %
%    lower_interval = (abs(min_blv) - thresh) / (num_blv_colors-1);
 %   upper_interval = (max_blv - thresh) / (num_blv_colors-1);

      if max_blv > -abs(thresh)
         lower_interval = (abs(min_blv) - thresh) / (num_blv_colors-1);
      else
         lower_interval = (abs(min_blv) - abs(max_blv)) / (num_blv_colors-1);
      end

      if min_blv < abs(thresh)
         upper_interval = (max_blv - thresh) / (num_blv_colors-1);
      else
         upper_interval = (max_blv - min_blv) / (num_blv_colors-1);
      end

    disp_blv = zeros(1,length(newcoords)) + brain_region_color_idx;
    lower_idx = find(brainlv <= -thresh);
    blv_offset = brainlv(lower_idx) - min_blv; 

    if lower_interval ~=0
       lower_color_idx = round(blv_offset/lower_interval)+first_lower_color_idx;
    else
       lower_color_idx = ones(size(blv_offset)) * first_lower_color_idx;
    end

    disp_blv(lower_idx) = lower_color_idx;

    upper_idx = find(brainlv >= thresh);
    blv_offset = max_blv - brainlv(upper_idx); 

    if upper_interval ~=0
       upper_color_idx = num_blv_colors - round(blv_offset/upper_interval);
    else
       upper_color_idx = num_blv_colors * ones(size(blv_offset));
    end

    upper_color_idx = upper_color_idx + first_upper_color_idx - 1;
    disp_blv(upper_idx) = upper_color_idx;
  else
      disp_blv = zeros(1,length(newcoords)) + brain_region_color_idx;

      if abs(min_blv) < 1e-6
         max_blv = min_blv + eps;
      else
         max_blv = min_blv + abs(min_blv)*1e-9;
      end
  end

    % get non_cluster_coords
    %
    if isempty(cluster_info)
       cluster_idx = newcoords;
    else
       cluster_idx = cluster_info.data{1}.idx;
    end

    if isequal(newcoords, cluster_idx)
       non_cluster_coords = [];
    else
       [tmp cluster_coords] = intersect(newcoords,cluster_idx);
       non_cluster_coords = ones(1,length(newcoords));
       non_cluster_coords(cluster_coords) = 0;
       non_cluster_coords = find(non_cluster_coords);
    end

    if (no_background_image),
       non_brain_region_color_idx = size(cmap,1);
       img = zeros(1,rows*cols*slices) + non_brain_region_color_idx;

       disp_blv(non_cluster_coords) = brain_region_color_idx;

       img(newcoords) = disp_blv;
       img = reshape(img,[rows cols 1 slices]); 
    else
       max_bg = max(bg_img(:));
       min_bg = min(bg_img(:));
       img = (bg_img - min_bg) / (max_bg - min_bg) * 100;

       disp_blv(non_cluster_coords) = img(newcoords(non_cluster_coords));

       if exist('lower_idx','var') & ~isempty(lower_idx)
          img(newcoords(lower_idx)) = disp_blv(lower_idx); 
       end

       if exist('upper_idx','var') & ~isempty(upper_idx)
          img(newcoords(upper_idx)) = disp_blv(upper_idx);
       end
    end;

if 0
    blv = reshape(img,[rows*cols,slices]);

    %  rotate image
    %
    for i=1:num_slices
        tmp=reshape(blv(:,i),img_width,img_height);
        tmp=rot90(tmp,mod(rot_amount,4));
        blv(:,i)=tmp(:);
    end

    % save a cornor of the last slice as the background intensity
    %
    bg_intensity = blv(1, num_slices);

    %  calculate how many slices to display for each row and column
    %  it's an algorithm from montage which will layout the slice
    %  in near square
    %
    if dims(1) > dims(2)
        siz = [dims(1), dims(2), dims(4)];
    else
        siz = [dims(2), dims(1), dims(4)];
    end

    cols_disp = sqrt(prod(siz))/siz(2);
    rows_disp = siz(3)/cols_disp;
    if (ceil(cols_disp)-cols_disp) < (ceil(rows_disp)-rows_disp),
        cols_disp = ceil(cols_disp); rows_disp = ceil(siz(3)/cols_disp);
    else
        rows_disp = ceil(rows_disp); cols_disp = ceil(siz(3)/rows_disp);
    end

    max_slice = rows_disp * cols_disp;
    rest_slices = max_slice - num_slices;
    empty_slices = -1 * ones(1,rest_slices);
    slice_idx = [slice_idx, empty_slices];

    % for empty slice (27-30), filled them with background intensity
    %
    blv_filled = [blv, ...
	bg_intensity * ...
	ones(img_height*img_width,rows_disp*cols_disp-num_slices)];

    blv_disp = [];
    for(row = 0:rows_disp-1)

        % take 'cols_disp' amount of slices from blv_filled
        % and put into blv_row
        %
        blv_row = blv_filled(:,[row*cols_disp+1:row*cols_disp+cols_disp]);

        % reshape the slice to integrate the whole row together
        %
        blv_row = reshape(blv_row, [img_height, img_width*cols_disp]);

        blv_disp = [blv_disp; blv_row];

    end

    blv_disp = reshape(blv_disp, [rows_disp*img_height, cols_disp*img_width]);