Getting started
Welcome to medigan: medical generative (adversarial) networks
Let’s install medigan and generate a few synthetic images.
pip install medigan
from medigan import Generators
# model 1 is "00001_DCGAN_MMG_CALC_ROI"
Generators().generate(model_id=1, install_dependencies=True)
Overview
Overview of medigan users, library, and models on the example of mammography cancer image dataset generation.
Note
Check out our article on medigan to get all the details.
Models
- Generative Models
- 00001_DCGAN_MMG_CALC_ROI
- 00002_DCGAN_MMG_MASS_ROI
- 00003_CYCLEGAN_MMG_DENSITY_FULL
- 00004_PIX2PIX_MMG_MASSES_W_MASKS
- 00005_DCGAN_MMG_MASS_ROI
- 00006_WGANGP_MMG_MASS_ROI
- 00007_INPAINT_BRAIN_MRI
- 00008_C-DCGAN_MMG_MASSES
- 00009_PGGAN_POLYP_PATCHES_W_MASKS
- 00010_FASTGAN_POLYP_PATCHES_W_MASKS
- 00011_SINGAN_POLYP_PATCHES_W_MASKS
- 00012_C-DCGAN_MMG_MASSES
- 00013_CYCLEGAN_MMG_DENSITY_OPTIMAM_MLO
- 00014_CYCLEGAN_MMG_DENSITY_OPTIMAM_CC
- 00015_CYCLEGAN_MMG_DENSITY_CSAW_MLO
- 00016_CYCLEGAN_MMG_DENSITY_CSAW_CC
- 00017_DCGAN_XRAY_LUNG_NODULES
- 00018_WGANGP_XRAY_LUNG_NODULES
- 00019_PGGAN_CHEST_XRAY
- 00020_PGGAN_CHEST_XRAY
- 00021_CYCLEGAN_BRAIN_MRI_T1_T2
Getting Started
API Reference
Tests
- Tests
- Setup medigan for running tests
- Test 1: test_medigan_imports
- Test 2: test_init_generators
- Test 3: test_generate_methods
- Test 4: test_generate_methods_with_additional_args
- Test 5: test_get_generate_method
- Test 6: test_search_for_models_method
- Test 7: test_find_model_and_generate_method
- Test 8: test_rank_models_by_performance
- Test 9: test_find_and_rank_models_by_performance
- Test 10: test_find_and_rank_models_then_generate_method
- Test 11: test_get_models_by_key_value_pair