'FileDataset' object has no attribute 'InstanceNumber'

[randall heidorn]

Hello,

I'm just getting into image processing and using pydicom.  I've got a list of DICOM images that I'm pulling in for analysis.

However, when I run the code below I get the error :

/******************************************************************************************************/

ERROR:

  File "/anaconda3/lib/python3.6/site-packages/pydicom/dataset.py", line 480, in __getattr__

    return super(Dataset, self).__getattribute__(name)

AttributeError: 'FileDataset' object has no attribute 'InstanceNumber'

/******************************************************************************************************/

Here's the code I'm running, which was borrowed from 

"https://www.raddq.com/dicom-processing-segmentation-visualization-in-python/"

/******************************************************************************************************/

CODE:

def load_scan(path):

    slices = [dicom.read_file(path + '/' + s,force = True) for s in os.listdir(path)]

    slices.sort(key = lambda x: int(x.InstanceNumber))

    try:

        slice_thickness = np.abs(slices[0].ImagePositionPatient[2] - slices[1].ImagePositionPatient[2])

    except:

        slice_thickness = np.abs(slices[0].SliceLocation - slices[1].SliceLocation)

        

    for s in slices:

        s.SliceThickness = slice_thickness

        

    return slices

def get_pixels_hu(scans):

    image = np.stack([s.pixel_array for s in scans])

    # Convert to int16 (from sometimes int16), 

    # should be possible as values should always be low enough (<32k)

    image = image.astype(np.int16)

    # Set outside-of-scan pixels to 1

    # The intercept is usually -1024, so air is approximately 0

    image[image == -2000] = 0

    

    # Convert to Hounsfield units (HU)

    intercept = scans[0].RescaleIntercept

    slope = scans[0].RescaleSlope

    

    if slope != 1:

        image = slope * image.astype(np.float64)

        image = image.astype(np.int16)

        

    image += np.int16(intercept)

    

    return np.array(image, dtype=np.int16)

id=0

patient = load_scan(filepath)

)

imgs = get_pixels_hu(patient)

/******************************************************************************************************/

Anybody have an idea what's causing this or if there's a modification I can make in dataset.py to get around this? Is InstanceNumber unique to my dataset?

Thanks

[Darcy Mason]

I'd suggest putting a print statement in the loop to see which file the error occurs on. It may be that one of the files in your folder is not actually a dicom file, or at least not an image.

[Aidos Sarsembayev]

I faced the same issue (but with the ImagePositionPatient attribute) while following the further tutorial from Kaggle 

The exception was telling me that there's no such attribute and indeed there wasn't.

for s in os.listdir(path):

    f = dicom.read_file(path +'\\'+ s)

    print(s)

    print(f.ImagePositionPatient)

I simply printed out each file name and tracked which one threw me the exception.

Then I read that file and printed out all meta. There was no ImagePositionPatient in it.

So, it seems like you don't have such an attribute in one of the files.

воскресенье, 4 марта 2018 г., 21:59:24 UTC+6 пользователь randall heidorn написал: