%filename = 'C:\Documents and Settings\akim2\Desktop\IMFILL\new_parsed_xml.csv';
filename = '\\ailab03\NEW_LIDC\LIDC\extract LIDC\new_parsed_xml.csv';
excelSheet = csvread_mod(filename);
count= 1;
for t= 1:size(excelSheet,1)
R = rem(t,18);
if R==0
coor(count,1)=excelSheet(t,1) ;
count = count+1;
end
end
counter =1;
%Take imageSOPs and put them in imageSOP_UIDs matrix
%for t=1:size(excelSheet, 1)
fid = fopen ('nodules_features.txt', 'wt');
fid2 = fopen ('nodules_many_regions.txt', 'wt');
fprintf(fid,'%s,%s,%s,%s,%s,%s,%s,%s,%s \n','ImageName','Circularity','Roughness','Elongation','Compactness','Eccentricity','Solidity','Extent','RadialDistanceSD');
fprintf(fid2,'%s,%s,%s,%s,%s,%s,%s,%s,%s \n','ImageName','Circularity','Roughness','Elongation','Compactness','Eccentricity','Solidity','Extent','RadialDistanceSD');
for i=2:size(excelSheet, 1)
s1 = coor{i,:};
s2 = regexp(s1, '\|', 'split');
s2(:);
for j= 1:size(s2,2)
s3(j) = regexp(s2(j), '\;', 'split');
end
for k=1:size(s2,2)-1
c=s3{k};
y=c{1};
x=c{2};
x=str2num(x);
y=str2num(y);
b(k)=x;
d(k)=y;
a(1,k)=x-287;
a(2,k)=y-87;
end
maxx=0;
maxy=0;
minx=0;
miny=0;
%for c= 1:k
maxx= max(b);
maxy= max(d);
minx= min(b);
miny= min(d);
I= zeros((maxx-minx)+1,(maxy-miny)+1);
%end
for f=1:size(s2,2)-1
x=b(f)-minx;
y=d(f)-miny;
I(x+1,y+1)= 1;
end
I2=im2bw(I);
I3=imfill(I2,'holes');
%figure;
%imshow(I3);
%HERE GOES FEATURE CALCULATION STUFF
count = 0;
%for i = 1:size(files,1)
% fn = sprintf('%s/%s', 'segmented_nodules', char(files(i).name));
% fn2 = sprintf('%s/%s', 'segmented_background', char(files(i).name));
% info = dicominfo(fn);
% info4 = dicominfo(fn2);
% I = dicomread(info);
% I4 = dicomread(info4);
% BW = (I~=-2000);
L = bwlabel(I3);
a = L;
flag = 0;
if (nanmax(a(:)) > 1)
flag = 1;
end
a(a~=0 & a~=1) = 0;
STATS = regionprops (a,'Area', 'Extent', 'Perimeter', 'Centroid', 'ConvexArea', 'ConvexImage', 'Solidity', 'Eccentricity', 'MajorAxisLength', 'EquivDiameter', 'MinorAxisLength');
STATS.ConvexPerimeter = 2 * (sqrt(pi*STATS.ConvexArea));
STATS.Circularity = (4*pi*STATS.Area)/(STATS.ConvexPerimeter .^ 2);
STATS.Elongation = STATS.MajorAxisLength/STATS.MinorAxisLength;
% b = double(I);
% b(b==-2000) = NaN;
% b4 = double(I4);
% b4(b4==-2000) = NaN;
% STATS.MinIntensity = nanmin(b(:));
% STATS.MaxIntensity = nanmax(b(:));
% STATS.MeanIntensity = nanmean(b(:));
% STATS.SDIntensity = nanstd(b(:));
% STATS.MinIntensityBG = nanmin(b4(:));
% STATS.MaxIntensityBG = nanmax(b4(:));
% STATS.SDIntensityBG = nanstd(b4(:));
% STATS.MeanIntensityBG = nanmean(b4(:));
% STATS.IntensityDifference = abs(STATS.MeanIntensity - STATS.MeanIntensityBG);
% fprintf(files(i).name,',',STATS(1).Area,',', STATS(1).Extent,',', STATS(1).Perimeter,'\n');
% count = count+1
% fn2 = sprintf('%s/%s', 'contours', char(files(i).name));
% info2 = dicominfo(fn2);
% I2 = dicomread(info2);
d = NaN;
j = 0;
for r = 1:size(I3,1)
for c = 1:size(I3,2)
if (I2(r,c) == 1)
j = j + 1;
d(j) = sqrt(double((c-(STATS.Centroid(1)^2))+(r-(STATS.Centroid(2)^2))));
end
end
end
STATS.RadialDistanceSD = nanstd(d(:));
STATS.Perimeter = j;
I3 = bwperim(STATS.ConvexImage);
j = 0;
for r = 1:size(I3,1)
for c = 1:size(I3,2)
if (I3(r,c) == 1)
j = j + 1;
end
end
end
STATS.ConvexPerimeter = j;
STATS.Roughness = STATS.ConvexPerimeter/STATS.Perimeter;
STATS.Compactness = STATS.Perimeter^2/(4*pi*STATS.Area);
if (flag == 0)
fprintf(fid,'%s,%f,%f,%f,%f,%f,%f,%f,%f \n',files(i).name,STATS.Circularity,STATS.Roughness,STATS.Elongation,STATS.Compactness,STATS.Eccentricity,STATS.Solidity,STATS.Extent,STATS.RadialDistanceSD);
else
fprintf(fid2,'%s,%f,%f,%f,%f,%f,%f,%f,%f \n',files(i).name,STATS.Circularity,STATS.Roughness,STATS.Elongation,STATS.Compactness,STATS.Eccentricity,STATS.Solidity,STATS.Extent,STATS.RadialDistanceSD);
end
%end
clear I;
clear d;
clear b;
clear I3;
end
fclose(fid);
fclose(fid2);
%end
