tutorial update

.github/workflows/python-app.yml

@@ -0,0 +1,33 @@
+# This workflow will install Python dependencies, lint with a single version of Python
+# For more information see: https://help.github.com/actions/language-and-framework-guides/using-python-with-github-actions
+
+# pylint: disable=<warning>
+
+name: Python application
+
+on:
+  push:
+    branches: [ main,mammo,optho ]
+  pull_request:
+    branches: [ main,mammo,optho ]
+
+jobs:
+  build:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: Set up Python 3.8
+      uses: actions/setup-python@v2
+      with:
+        python-version: 3.8
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        python -m pip install toml==0.10.2 yapf==0.31.0 isort==5.9.3 pylint==2.11.0
+    - name: Lint with pylint, format with yapf and isort
+      run: |
+        pylint * --verbose -r y
+        isort . --check-only --diff
+        yapf --recursive --verbose --in-place .

.gitignore

@@ -0,0 +1,129 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+pip-wheel-metadata/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+.python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2023 Rajpurkar Lab / Harvard Medical AI
+Copyright (c) 2021 Alex Tamkin
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

conf/dataset/HAM10000.yaml → conf/dataset/ham10000.yaml

@@ -2,6 +2,6 @@
 loss: cross_entropy
 metric: auroc
 
-name: HAM10000
+name: ham10000
 batch_size: 64
 num_workers: 8

requirements.txt

@@ -26,3 +26,4 @@ h5py
 pylidc
 bs4
 mne
+SimpleITK

src/datasets/catalog.py

@@ -33,7 +33,7 @@
     'isic2019': isic2019.ISIC2019,
     'isic2020': isic2020.ISIC2020,
     'ddi': ddi.ddi,
-    'HAM10000': HAM10000.HAM10000,
+    'ham10000': HAM10000.HAM10000,
     'pad_ufes_20': pad_ufes_20.pad_ufes_20,
 
     # ECG
@@ -63,7 +63,7 @@
     'mimic-cxr',
     #dermatology.
     'ddi',
-    'HAM10000',
+    'ham10000',
     'isic2020',
     'pad_ufes_20',
     'isic2019',

src/datasets/ct/lndb_utils.py

@@ -0,0 +1,93 @@
+import csv
+
+import numpy as np
+import SimpleITK as sitk
+from scipy.ndimage import zoom
+
+
+def readCsv(csvfname):
+    # read csv to list of lists
+    with open(csvfname, 'r') as csvf:
+        reader = csv.reader(csvf)
+        csvlines = list(reader)
+    return csvlines
+
+
+def writeCsv(csfname, rows):
+    # write csv from list of lists
+    with open(csfname, 'w', newline='') as csvf:
+        filewriter = csv.writer(csvf)
+        filewriter.writerows(rows)
+
+
+def readMhd(filename):
+    # read mhd/raw image
+    itkimage = sitk.ReadImage(filename)
+    scan = sitk.GetArrayFromImage(itkimage)  #3D image
+    #  print(scan.shape)
+
+    spacing = itkimage.GetSpacing()  #voxelsize
+    origin = itkimage.GetOrigin()  #world coordinates of origin
+    transfmat = itkimage.GetDirection()  #3D rotation matrix
+    #  print(dir(itkimage))
+    return scan, spacing, origin, transfmat
+
+
+def writeMhd(filename, scan, spacing, origin, transfmat):
+    # write mhd/raw image
+    itkim = sitk.GetImageFromArray(scan, isVector=False)  #3D image
+    itkim.SetSpacing(spacing)  #voxelsize
+    itkim.SetOrigin(origin)  #world coordinates of origin
+    itkim.SetDirection(transfmat)  #3D rotation matrix
+    sitk.WriteImage(itkim, filename, False)
+
+
+def getImgWorldTransfMats(spacing, transfmat):
+    # calc image to world to image transformation matrixes
+    transfmat = np.array([transfmat[0:3], transfmat[3:6], transfmat[6:9]])
+    for d in range(3):
+        transfmat[0:3, d] = transfmat[0:3, d] * spacing[d]
+    transfmat_toworld = transfmat  #image to world coordinates conversion matrix
+    transfmat_toimg = np.linalg.inv(transfmat)  #world to image coordinates conversion matrix
+
+    return transfmat_toimg, transfmat_toworld
+
+
+def convertToImgCoord(xyz, origin, transfmat_toimg):
+    # convert world to image coordinates
+    xyz = xyz - origin
+    xyz = np.round(np.matmul(transfmat_toimg, xyz))
+    return xyz
+
+
+def convertToWorldCoord(xyz, origin, transfmat_toworld):
+    # convert image to world coordinates
+    xyz = np.matmul(transfmat_toworld, xyz)
+    xyz = xyz + origin
+    return xyz
+
+
+def extractCube(scan, spacing, xyz, cube_size=80, cube_size_mm=51):
+    # Extract cube of cube_size^3 voxels and world dimensions of cube_size_mm^3 mm from scan at image coordinates xyz
+    xyz = np.array([xyz[i] for i in [2, 1, 0]], np.int)
+    spacing = np.array([spacing[i] for i in [2, 1, 0]])
+    scan_halfcube_size = np.array(cube_size_mm / spacing / 2, np.int)
+    if np.any(xyz < scan_halfcube_size) or np.any(xyz + scan_halfcube_size > scan.shape):  # check if padding is necessary
+        maxsize = max(scan_halfcube_size)
+        scan = np.pad(
+            scan, ((
+                maxsize,
+                maxsize,
+            )), 'constant', constant_values=0
+        )
+        xyz = xyz + maxsize
+
+    scancube = scan[xyz[0] - scan_halfcube_size[0]:xyz[0] + scan_halfcube_size[0],  # extract cube from scan at xyz
+                    xyz[1] - scan_halfcube_size[1]:xyz[1] + scan_halfcube_size[1],
+                    xyz[2] - scan_halfcube_size[2]:xyz[2] + scan_halfcube_size[2]]
+
+    sh = scancube.shape
+
+    scancube = zoom(scancube, (cube_size / sh[0], cube_size / sh[1], cube_size / sh[2]), order=2)  #resample for cube_size
+
+    return scancube