heatmap_cfgs.yaml

# CUDA_VISIBLE_DEVICES=0,1 python create_heatmaps.py --config config_template.yaml
exp_arguments:
  # number of classes
  n_classes: 2
  # name tag for saving generated figures and assets
  save_exp_code: HEATMAP_OUTPUT_scorer
  # where to save raw asset files
  raw_save_dir: /home/cyyan/Projects/HER2proj/results/Yale_4Heatmaps_PhiHER2_yhat/heatmap_raw
  # where to save final heatmaps
  production_save_dir: /home/cyyan/Projects/HER2proj/results/Yale_4Heatmaps_PhiHER2_yhat/heatmap_production_hd
  batch_size: 384
data_arguments: 
  # where is data stored; can be a single str path or a dictionary of key, data_dir mapping
  # data_dir: "/home/cyyan/Projects/HER2proj/data_ModPath_HER2_v3/pkg_v3/Yale_HER2_cohort/SVS"
  data_dir: /home/cyyan/Projects/HER2proj/data_ModPath_HER2_v3/pkg_v3/Yale_HER2_cohort/SVS
  spec_slide: /home/cyyan/Projects/HER2proj/results/Yale_3CaseSplits/her2status_KFoldsCV/splits_time0_fold1.csv #['Her2Neg_Case_09.svs', 'Her2Pos_Case_06.svs']
  # spec_slide: /mnt/DATA/HEROHE_challenge/HEROHE_TestGTinfo.csv
  # spec_slide: ['73.mrxs', '122.mrxs', '125.mrxs', '253.mrxs', '260.mrxs', '262.mrxs']
  # column name for key in data_dir (if a dict mapping is used)
  data_dir_key: source
  # csv list containing slide_ids (can additionally have seg/patch paramters, class labels, etc.)
  process_list: 
  # preset file for segmentation/patching
  preset: 
  # file extention for slides
  slide_ext: .svs
  # label dictionary for str: interger mapping (optional)
  label_dict: {"Negative": 0, "Positive": 1}
patching_arguments:
  # arguments for patching
  patch_size: [256, 256]
  overlap: 0.5 # heatmap overlapping rate
  patch_level: 0
  custom_downsample: 1
model_arguments: 
  # arguments for initializing model from checkpoint
  # ckpt_path: /home/cyyan/Projects/HER2proj/models/Yale/HER2status/ABMIL_insttop2000_0dropout_lossCE_1WSampling/time0/s_3_checkpoint.pt
  # ckpt_path: "/home/cyyan/Projects/HER2proj/models/HEROHE/HER2status/new_ABMIL_tile5k_all/time0/s_0_checkpoint.pt"
  ckpt_path: "/home/cyyan/Projects/HER2proj/models/Yale/HER2status/new_PhiHER2_tileAll_twice1000_0abmil_0sel/time0/s_1_checkpoint.pt"
  cluster_path: "/home/cyyan/Projects/HER2proj/results/Yale_4APCluster_rmPos73/time_0_fold_1_prototypes.pt"
  model_type:  PhiHER2 #ABMIL see utils/eval_utils/
  initiate_fn: initiate_model # see utils/eval_utils/
  model_size: ccl2048
  drop_out: false
  retccl_filepath: "/home/cyyan/Projects/HER2proj/models/CCL_best_ckpt.pth"

heatmap_arguments:
  # downsample at which to visualize heatmap (-1 refers to downsample closest to 32x downsample)
  vis_level: 1
  # transparency for overlaying heatmap on background (0: background only, 1: foreground only)
  alpha: 0.4
  # whether to use a blank canvas instead of original slide
  blank_canvas: false
  # whether to also save the original H&E image
  save_orig: false
  # file extension for saving heatmap/original image
  save_ext: png
  # whether to calculate percentile scores in reference to the set of non-overlapping patches
  use_ref_scores: false
  # whether to use gaussian blur for further smoothing
  blur: false
  # whether to shift the 4 default corner points for checking if a patch is inside a foreground contour
  use_center_shift: true
  # whether to only compute heatmap for ROI specified by x1, x2, y1, y2
  use_roi: false 
  # whether to calculate heatmap with specified overlap (by default, coarse heatmap without overlap is always calculated)
  calc_heatmap: true
  # whether to binarize attention scores
  binarize: false
  # binarization threshold: (0, 1)
  binary_thresh: -1
  # factor for downscaling the heatmap before final dispaly
  custom_downsample: 1
  cmap: jet

sample_arguments:
  samples:
    - name: "topk_high_attention"
      sample: true
      seed: 1
      k: 15 # save top-k patches
      mode: topk


s7_clusterPatternVis arguments:
  # path of the visualization outputs
  interactive_vis_dir: /home/cyyan/Projects/HER2proj/results/Yale_4Heatmaps_PhiHER2_yhat/Interactive_vis_tiles
  # See tiles corresponding to a particular value of the target variable. Coincide with the names of the folders in path/hightiles/'
  vis_target_val:
  # when tagged, use the decision embeddings (related to the target variable).
  vis_decision_emb: true
  # If tagged, tiles will be clustered. If not, color of the points corresponds to the decision score(tiles logits)
  vis_cluster: true