jgamper / compay-syntax Goto Github PK

:microscope: Syntax - the arrangement of whole-slide-images and their image tiles to create well-formed computational pathology pipelines.

License: MIT License

Python 100.00%

computational-pathology medical-imaging wsis openslide slide-images pathology-image whole-slide-image whole-slide-imaging

compay-syntax's Introduction

Help building our code and community!

All contributions are welcome! Please raise an issue for a bug, feature or pull request!
Tweet about this repo!
Give this repo a star! ⭐

Quick Start

Tissue mask and tiling pipeline

from syntax.slide import Slide
from syntax.transformers.tissue_mask import OtsuTissueMask
from syntax.transformers.tiling import SimpleTiling
from syntax.transformers import Pipeline, visualize_pipeline_results

slide = Slide(slide_path=slide_path)
pipeline = Pipeline([OtsuTissueMask(), SimpleTiling(magnification=40,
                                                  tile_size=224,
                                                  max_per_class=10)])
slide = pipeline.fit_transform(slide)
visualize_pipeline_results(slide=slide,
                           transformer_list=pipeline.transformers,
                           title_list=['Tissue Mask', 'Random Tile Sampling'])

Install

pip install compay-syntax==0.4.0

You will need to install openslide. This is included in the dependencies so may be automatically installed. If not I found this to help:

#!/bin/bash

sudo apt-get install openslide-tools
pip install openslide-python

compay-syntax's People

Contributors

Stargazers

Watchers

Forkers

yshtii

compay-syntax's Issues

development / test OpenJP2 versus Openslide objects

Gotta start documenting the development somehow? Is there a way to keep goals that we would like to reach?

Either way, it would be good to write unittest applied to one WSI that would be read by OpenJP2 versus Openslide, and compare the outputs (just to be sure), and also develop something like thumbnail for OpenJP2.

Annotations

Record standard annotation formats in the lab, and work from that

Set up Travis CI

use wget within to get data for testing.

Edit contribution.md

https://github.com/Peter554/WholeSlideImageSampler/blob/0c8d9eae1f84378e8b42aa8de025a2ae5188d233/CONTRIBUTING.md#L1

[BUG]

Describe the bug
I have tried to run the code snippet in the Readme.MD file and I am getting a few errors. The WSIs is from the Panda Challenge hosted on Kaggle.

Code snippet or Error Message
The code that I have used:

slide = Slide(slide_path=dataset[0])

pipeline = Pipeline([OtsuTissueMask(), SimpleTiling(magnification=40,
                                                  tile_size=224,
                                                  max_per_class=10)])
slide = pipeline.fit_transform(slide)
visualize_pipeline_results(slide=slide,
                           transformer_list=pipeline.transformers,
                           title_list=['Tissue Mask', 'Random Tile Sampling'])

The error I have got is:

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
/opt/conda/lib/python3.6/site-packages/syntax/slide/slide.py in __init__(self, slide_path, level0, verbose)
     44             try:
---> 45                 self.level0 = float(self.properties['openslide.objective-power'])
     46                 if self.verbose:

/opt/conda/lib/python3.6/site-packages/openslide/__init__.py in __getitem__(self, key)
    250         if v is None:
--> 251             raise KeyError()
    252         return v

KeyError: 

During handling of the above exception, another exception occurred:

Exception                                 Traceback (most recent call last)
<ipython-input-7-d399eadbf4ad> in <module>
      4 from syntax.transformers import Pipeline, visualize_pipeline_results
      5 
----> 6 slide = Slide(slide_path=dataset[0])
      7 
      8 pipeline = Pipeline([OtsuTissueMask(), SimpleTiling(magnification=40,

/opt/conda/lib/python3.6/site-packages/typeguard/__init__.py in wrapper(*args, **kwargs)
    838         memo = _CallMemo(python_func, _localns, args=args, kwargs=kwargs)
    839         check_argument_types(memo)
--> 840         retval = func(*args, **kwargs)
    841         check_return_type(retval, memo)
    842 

/opt/conda/lib/python3.6/site-packages/syntax/slide/slide.py in __init__(self, slide_path, level0, verbose)
     47                     print('Level 0 found @ {}X'.format(self.level0))
     48             except:
---> 49                 raise Exception('Slide does not have property objective-power.')
     50         else:
     51             self.level0 = float(level0)

Exception: Slide does not have property objective-power.

Patch Coordinates?

Hi,

This is a brill repo!

I was wondering whether it is possible to retrieve the coordinates of the patches. For example, are the patches randomly retrieved?

Taran