The algorithm failed on one or more cases.

Dear,

Can you provide me some error information, I use try out to run successfully, but submit run shows error.

Thanks Wang

Dear Wang,

unfortunately, I do not have access to the logs either. Here's some recommendations, however:

• If you click on your algorithm and on "Results", you should get a list of all the images that have been run trough your container, together with the status. If you click on the status (e.g., on "Succeeded with warnings") you will get the log of the container. At least all other participants have told me this worked for them.
• You might want to try out the algorithm with a plain white image (which is part of the preliminary test set). We incorporated this image in order to perform a technical check for algorithms that found no candidates. Several participants' containers had issues with these on the final test set last time.

Should you still run into issues, I might recommend writing a polite email to grand-challenge support and ask about the logs. For some cases, these were not available on the web interface in the past, but the guys there are really nice and helpful if you ask politely :-)

Best,

Marc

I've just got the same problem.Have you solve it? Thanks a lot!

Apparently, they were able to, at least they successfully submitted after this discussion :-)

Hello, I am having the same problem. My container worked correctly during the try-out phase. And it gave an output like the example below.

Sample output below for 007.tiff data. (try-out was also successful)

{'type': 'Multiple points', 'points': [{'point': [0.10917296464773094, 0.03707331924495863, 0], 'probability': 0.6642420196533203, 'name': 'mitotic figure'}, ..., {'point': [0.14601884021634012, 0.15773218954833626, 0], 'probability': 0.9289340209960938, 'name': 'non-mitotic figure'}], 'version': {'major': 1, 'minor': 0}}

Sample output below for a white tiff. (try-out also output as successful)

{'type': 'Multiple points', 'points': [], 'version': {'major': 1, 'minor': 0}}

I couldn't find my mistake. Could you take a look at the little code block below?

''' import glob import cv2 from PIL import Image from inference import MitosisAnalysis from util import loadmodule import json mitosisanalysis = MitosisAnalysis()

class Mitosisdetection():

def init(self, ):
    self.inputpath = "/input/images/histopathology-roi-cropout/.tif"
    self.outputfilepath = "/output/mitotic-figures.json"

def process(self):
    imagepath=glob.glob(self.inputpath)[0]
    img = cv2.imread(imagepath, cv2.IMREADUNCHANGED)
    dpi=Image.open(image_path).info["dpi"]
    result = {}

    result["type"]= "Multiple points"
    result["points"] = []
    result["version"]= {
        "major": 1,
        "minor": 0
    }

    resulti = mitosisanalysis.run(img,dpi)
    result["points"].extend(resulti)
    print(result)

    with open(str(self.outputfilepath), "w") as f:
        json.dump(result, f)

if name == "main": # loads the image(s), applies DL detection model & saves the result Mitosisdetection().process()

''''

Thank you, Engin

Dear Engin,

several cases have failed because of an error related to cv2.resize, for example 019.tiff:

2022-08-23T20:54:41.088000+00:00 Traceback (most recent call last): 2022-08-23T20:54:41.088000+00:00 File "runner.py", line 63, in 2022-08-23T20:54:41.088000+00:00 Mitosisdetection().process() 2022-08-23T20:54:41.088000+00:00 File "runner.py", line 54, in process 2022-08-23T20:54:41.088000+00:00 real_y=y) 2022-08-23T20:54:41.088000+00:00 File "/opt/algorithm/inference.py", line 159, in run 2022-08-23T20:54:41.088000+00:00 regionImage, (mitosisResizeParameterRow, mitosisResizeParameterCol), cv2.INTER_LINEAR 2022-08-23T20:54:41.088000+00:00 cv2.error: OpenCV(4.5.5) /io/opencv/modules/imgproc/src/resize.cpp:4052: error: (-215:Assertion failed) !ssize.empty() in function 'resize' 2022-08-23T20:54:41.088000+00:00

Best,

Marc