File size: 5,170 Bytes
9ee6a3d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
#!/usr/bin/env python3
"""
Fat filtering script - replace voxels with HU < -20 with label 10
Processes both manual labels and model predictions separately
"""

import os
from pathlib import Path
import numpy as np # type: ignore
import nibabel as nib # type: ignore

# Input directories
root_100_120 = Path("../100-120")
manual_labels_dir = root_100_120 / "labels_9_muscles"
model_pred_labels_dir = root_100_120 / "label_9_muscles_model_pred"
images_dir = root_100_120 / "images_100-120"

# Output directories
output_base = Path("../fat_filtered_100-120")
manual_output_dir = output_base / "labels_9_muscles_fat_filtered"
model_pred_output_dir = output_base / "label_9_muscles_model_pred_fat_filtered"

# Fat threshold
fat_hu_thresh = -20
fat_label = 10

def load_image_and_label(case_id: int, label_dir: Path):
    """Load CT image and label file for a specific case."""
    
    img_path = images_dir / f"{case_id}_0000.nii.gz"
    if not img_path.exists():
        print(f"Image not found: {img_path}")
        return None, None, None

    lab_path = label_dir / f"{case_id}.nii.gz"
    if not lab_path.exists():
        print(f"Label not found: {lab_path}")
        return None, None, None
    
    try:
        img_nib = nib.load(str(img_path))  # type: ignore
        lab_nib = nib.load(str(lab_path))  # type: ignore
        
        img_arr = img_nib.get_fdata()
        lab_arr = lab_nib.get_fdata()
        
        return img_arr, lab_arr, lab_nib
    except Exception as e:
        print(f"Error loading case {case_id}: {e}")
        return None, None, None

def apply_fat_filter(img_arr: np.ndarray, lab_arr: np.ndarray):
    """Apply fat filtering - replace muscle voxels with HU < -20 with label 10."""
    filtered_lab = lab_arr.copy()
    fat_mask = (img_arr <= fat_hu_thresh) & (lab_arr > 0)
    filtered_lab[fat_mask] = fat_label
    return filtered_lab

def process_labels(label_dir: Path, output_dir: Path, label_type: str):
    """Process all labels in a directory and save fat-filtered versions."""
    
    print(f"\nProcessing {label_type} labels...")
    print(f"Input directory: {label_dir}")
    print(f"Output directory: {output_dir}")
    
    output_dir.mkdir(parents=True, exist_ok=True)
    
    processed_count = 0
    failed_count = 0
    
    for case_id in range(100, 121):
        print(f"Processing case {case_id}...")
        
        img_arr, lab_arr, lab_nib = load_image_and_label(case_id, label_dir)
        
        if img_arr is not None and lab_arr is not None and lab_nib is not None:
            filtered_lab = apply_fat_filter(img_arr, lab_arr)
            
            filtered_nib = nib.Nifti1Image(
                filtered_lab.astype(np.uint8),  # type: ignore
                lab_nib.affine,
                lab_nib.header
            )
            
            output_path = output_dir / f"{case_id}.nii.gz"
            nib.save(filtered_nib, str(output_path))  # type: ignore
            
            original_muscle_voxels = np.count_nonzero((lab_arr > 0) & (lab_arr <= 9))  # type: ignore
            fat_voxels = np.count_nonzero(filtered_lab == fat_label)  # type: ignore
            total_muscle_voxels = np.count_nonzero(lab_arr > 0)  # type: ignore
            
            print(f"  Case {case_id}: {fat_voxels} fat voxels added (original muscle voxels: {original_muscle_voxels})")
            
            processed_count += 1
        else:
            print(f"  Case {case_id}: Failed to load")
            failed_count += 1
    
    print(f"\n{label_type} processing complete:")
    print(f"  Successfully processed: {processed_count} cases")
    print(f"  Failed: {failed_count} cases")
    print(f"  Output saved to: {output_dir}")

def main():
    """Main function to process both manual and model prediction labels."""
    
    print("="*80)
    print("FAT FILTERING SCRIPT")
    print("Replace voxels with HU < -20 with label 10")
    print("="*80)
    
    if not manual_labels_dir.exists():
        print(f"Error: Manual labels directory not found: {manual_labels_dir}")
        return
    
    if not model_pred_labels_dir.exists():
        print(f"Error: Model prediction labels directory not found: {model_pred_labels_dir}")
        return
    
    if not images_dir.exists():
        print(f"Error: Images directory not found: {images_dir}")
        return
    
    process_labels(
        label_dir=manual_labels_dir,
        output_dir=manual_output_dir,
        label_type="Manual"
    )
    
    process_labels(
        label_dir=model_pred_labels_dir,
        output_dir=model_pred_output_dir,
        label_type="Model Prediction"
    )
    
    print("\n" + "="*80)
    print("FAT FILTERING COMPLETE")
    print("="*80)
    print(f"Results saved to:")
    print(f"  - {manual_output_dir} (manual labels with fat filtering)")
    print(f"  - {model_pred_output_dir} (model predictions with fat filtering)")
    print(f"\nFat filtering details:")
    print(f"  - Threshold: HU < {fat_hu_thresh}")
    print(f"  - Fat label: {fat_label}")
    print(f"  - Original muscle labels preserved (1-9)")
    print(f"  - Fat voxels labeled as {fat_label}")

if __name__ == "__main__":
    main()