Inference scripts

.gitignore

@@ -65,3 +65,6 @@ stdout.txt
 
 # network weights
 #CheckpointKi67/
+
+# Inference folder
+Inference/

README.md

@@ -68,7 +68,22 @@ python eval_opt_thresholds.py --inputPath test256 --configPath configs/eval_Ki67
 - Results are saved in the table specified by `dataname`.
 - To optimize the threshold associated with cells detected as negative at the marker, specify `--channel 1`.
 
-
+## Step 5: Inference
+- The `infer.py` script is used to run the model in inference mode.
+- Command line
+```
+python infer.py --inputPath KI67_Tiles_256_256_40x/ --configPath configs/eval_Ki67_LNEN.json --outputPath Inference --save_numpy --visualization
+``` 
+- The `inputPath` folder must have the following structure:
+    - `inputPath`
+        - patient_ID
+            - accept
+                - patient_id_tiles_x_y.jpg
+- The inference script saves the inference results for each tile in three formats:
+    - `json` file (default)
+    - `npy` numpy matrix if the `--save_numpy` argument is specified 
+    - `jpg` annotated image if the `--visualization` argument is specified 
+- The `outputPath` folder will have the same organization as the `inputPath` folder.
 
 ## TO DO LIST
 

infer.py

@@ -0,0 +1,73 @@
+from pipeline import Pipeline
+from config import Config
+import os  
+import glob
+import imageio
+import cv2
+import argparse
+import numpy as np
+import json
+import warnings
+import pandas as pd
+warnings.filterwarnings("ignore", category=DeprecationWarning)
+def get_parser():
+
+    parser = argparse.ArgumentParser('Inference')
+    parser.add_argument('--inputPath', '-i', required=True, help="Path to the test set")
+    parser.add_argument('--outputPath', '-o', required=True, help="Path to the folder where the results will be saved")
+    parser.add_argument('--configPath', '-c', required=True, help="Path to the json config file")
+    parser.add_argument('--save_numpy', action='store_true', help="Save the inference results for each tile in npy format")
+    parser.add_argument('--visualization', action='store_true', help="Save the tile annotated in jpg format for visualization")
+    return parser
+
+def visualizer(img,points):
+    r=1
+    colors=[
+            (255,0,0),
+            (0,255,0),
+            (0,0,255)
+            ]
+    image=np.copy(img)
+    for p in points:
+        x,y,c=p[0],p[1],p[2]
+        cv2.circle(image, (int(x), int(y)), int(r), colors[int(c)], 2)
+    return image
+
+
+def infer(args=None):
+    parser = get_parser()
+    args = parser.parse_args(args)
+    conf=Config()
+    conf.load(args.configPath)
+    pipeline=Pipeline(conf)
+
+    if os.path.isdir(args.inputPath):
+        #print('args.inputPath  = ', args.inputPath, os.listdir(args.inputPath))
+        patient_ids_list = os.listdir(args.inputPath)
+        for patient_id in patient_ids_list:
+            print('patient_id' , patient_id)
+            os.makedirs(os.path.join(args.outputPath, patient_id), exist_ok = True)
+            os.makedirs(os.path.join(args.outputPath, patient_id, 'accept'), exist_ok = True)
+            data = [ os.path.join(args.inputPath, patient_id, 'accept', f)  for f in os.listdir(os.path.join(args.inputPath, patient_id, 'accept')) if '.jpg' in f]
+            for d in data:
+                img=imageio.imread(d)
+                pred_cell, pred_neg, pred_pos =pipeline.predict(img, exhaustive=True)
+                if args.save_numpy:
+                    np.save(os.path.join(args.outputPath, patient_id, 'accept',d.split("/")[-1][:-4] + "_pred_neg_mask.npy"), pred_neg)
+                    np.save(os.path.join(args.outputPath, patient_id, 'accept',d.split("/")[-1][:-4] + "_pred_pos_mask.npy"), pred_pos)
+                if args.visualization:
+                    output=visualizer(img,pred_cell)
+                    imageio.imwrite(os.path.join(args.outputPath, patient_id, 'accept',d.split("/")[-1][:-4] + "viz.jpg"),output)
+                list_cells_for_json = []
+                for cell in pred_cell:
+                    c_cell= {}
+                    c_cell['x'] = cell[0]
+                    c_cell['y'] = cell[1]
+                    c_cell['label_id'] = int(cell[2]) + 1
+                    list_cells_for_json.append(c_cell)
+                json_pred_fname = os.path.join(args.outputPath, patient_id, 'accept', d.split("/")[-1][:-3]+'json')
+                with open(json_pred_fname, 'w') as f:
+                    json.dump(list_cells_for_json, f)
+
+if __name__ == "__main__":
+   infer()