Update README.md

mask_optimization/README.md

@@ -4,7 +4,7 @@ This subdirectory provides scripts to mask optimization.
 ## scripts
 
 ### `1.apply_segmentation.py` 
-This script runs the segmentation models trained on the current dataset and save the results.
+This script runs the segmentation models trained on the current dataset and save the results (requires MLFlow).
 
 usage:
 ```
@@ -18,6 +18,11 @@ Input Variable | Description
 --data_dir | input directory
 --out_dir | output directory
 
+Note: Please modify the source code so that the segmentation model can be loaded from MLFlow server (l.92-l.116).
+
+output:
+- segmentation mask predicted by the segmentation model, which ends with `.npy`
+
 ### `2.ridge_regression.py` 
 This script applies ridge regression between the IF-mask intensity and the output probabiities.
 
@@ -26,6 +31,9 @@ usage:
 python 2.ridge_regression.py antibody 
 ```
 
+output:
+- dataframe containing the number of pixels predicted as positive by the segmentation model and coefficient of ridge regression, which ends with `.npy`
+
 ### `3.IF_cutoff_optimization.py` 
 This script optimizes IF intensity cufoff. 
 
@@ -34,6 +42,9 @@ usage:
 python 3.IF_cutoff_optimization.py antibody 
 ```
 
+output:
+- dataframe containing the updated IF threshold, which ends with `_otsu.pkl`
+
 ### `4_CELL.MCC_calculation.py` 
 This script calculates Matthew’s correlation coefficient (MCC) between the prediction and mask for leucocytes, myeloid cells, lymphocytes, plasma cells, and endothelial cells.
 
@@ -42,9 +53,17 @@ usage:
 python 4_CELL.MCC_calculation.py antibody 
 ```
 
+output:
+- dictionary containing MCC, which ends with `.pkl`
+
+
 ### `5_CELL.overlap_ratio_optimization.py` 
 This script optimizes the nucleus overlap cut-off based on the MCC.
 usage:
 ```
 python 5_CELL.overlap_ratio_optimization.py antibody 
 ```
+
+output:
+- updated mask files, which ends with `_IHC_cellpose_mask.png` or `_IHC_nonrigid_mask2.png`
+