masilab/alarm

alarm's Issues

Why does CT slice change from single in ROI calculation?

Hi there,

I was looking at your code in get_liver_measurement_JC2_V2.py, and am wondering why when you start calculating the peripheral ROI's, you are changing the slice number that you are interested in from your erosion calculation? Particularly this segment of code:

start=np.inf
for i in range(seg_iso.shape[0]):
    #figure out the edge of liver in this particular direction
    if(seg_iso[i][single[1]][single[-1]]==1):
        start=i
        break
#fetch the 1/3 point away from edge
midptr=(start+int((single[0]-start)/3),single[1],single[-1])
#mask
newmask=np.zeros(seg_iso.shape)
for i in range(15):
    for j in range(15):
        for k in range(5):
            #draw a circle around that point (based on distance)
            a,b,c=midptr[0]+7-i,midptr[1]+7-j,midptr[-1]+7-k
            if((a-midptr[0])**2+(b-midptr[1])**2<=7**2):
                newmask[a][b][c]=1 #marked 1

Why is the value of k changing when we had established that single contains the i, j, k of the center cut? Thanks for any insight!