File size: 7,656 Bytes
8312ddd
 
 
 
 
d28c822
4d94df5
 
3cf3c0d
4d94df5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b5ba092
4d94df5
 
b5ba092
 
4d94df5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cfb7ff1
4d94df5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
67b1d32
4d94df5
 
 
 
 
 
 
 
 
 
 
 
 
 
8312ddd
4d94df5
 
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
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
import cv2
import numpy as np
import pandas as pd
import gradio as gr
import matplotlib.pyplot as plt
from datetime import datetime
from sklearn.cluster import DBSCAN
from scipy import ndimage

class BloodCellAnalyzer:
    def __init__(self):
        self.min_cell_area = 100
        self.max_cell_area = 5000
        self.min_circularity = 0.7
        
    def preprocess_image(self, image):
        """Enhanced image preprocessing with multiple color spaces and filtering."""
        if len(image.shape) == 2:
            image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
        
        # Convert to multiple color spaces for robust detection
        hsv = cv2.cvtColor(image, cv2.COLOR_RGB2HSV)
        lab = cv2.cvtColor(image, cv2.COLOR_RGB2LAB)
        gray = cv2.cvtColor(image, cv2.COLOR_RGB2GRAY)
        
        # Create masks for different color ranges
        hsv_mask = cv2.inRange(hsv, np.array([0, 20, 20]), np.array([180, 255, 255]))
        lab_mask = cv2.inRange(lab, np.array([20, 120, 120]), np.array([200, 140, 140]))
        
        # Combine masks
        combined_mask = cv2.bitwise_or(hsv_mask, lab_mask)
        
        # Apply advanced morphological operations
        kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (5, 5))
        clean_mask = cv2.morphologyEx(combined_mask, cv2.MORPH_CLOSE, kernel, iterations=2)
        clean_mask = cv2.morphologyEx(clean_mask, cv2.MORPH_OPEN, kernel, iterations=1)
        
        # Apply distance transform to separate touching cells
        dist_transform = cv2.distanceTransform(clean_mask, cv2.DIST_L2, 5)
        _, sure_fg = cv2.threshold(dist_transform, 0.5 * dist_transform.max(), 255, 0)
        
        return clean_mask.astype(np.uint8), sure_fg.astype(np.uint8)

    def extract_cell_features(self, contour):
        """Extract comprehensive features for each detected cell."""
        area = cv2.contourArea(contour)
        perimeter = cv2.arcLength(contour, True)
        circularity = 4 * np.pi * area / (perimeter ** 2) if perimeter > 0 else 0
        
        # Calculate additional shape features
        hull = cv2.convexHull(contour)
        hull_area = cv2.contourArea(hull)
        solidity = float(area) / hull_area if hull_area > 0 else 0
        
        # Calculate moments and orientation
        moments = cv2.moments(contour)
        cx = int(moments['m10'] / moments['m00']) if moments['m00'] != 0 else 0
        cy = int(moments['m01'] / moments['m00']) if moments['m00'] != 0 else 0
        
        # Calculate eccentricity using ellipse fitting
        if len(contour) >= 5:
            (x, y), (MA, ma), angle = cv2.fitEllipse(contour)
            eccentricity = np.sqrt(1 - (ma / MA) ** 2) if MA > 0 else 0
        else:
            eccentricity = 0
            angle = 0
        
        return {
            'area': area,
            'perimeter': perimeter,
            'circularity': circularity,
            'solidity': solidity,
            'eccentricity': eccentricity,
            'orientation': angle,
            'centroid_x': cx,
            'centroid_y': cy
        }

    def detect_cells(self, image):
        """Detect and analyze blood cells with advanced filtering."""
        mask, sure_fg = self.preprocess_image(image)
        
        # Find contours
        contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
        
        # Extract features and filter cells
        cells = []
        valid_contours = []
        
        for i, contour in enumerate(contours):
            features = self.extract_cell_features(contour)
            
            # Apply multiple criteria for cell validation
            if (self.min_cell_area < features['area'] < self.max_cell_area and 
                features['circularity'] > self.min_circularity and 
                features['solidity'] > 0.8):
                
                features['label'] = i + 1
                cells.append(features)
                valid_contours.append(contour)
        
        return valid_contours, cells, mask

    def analyze_image(self, image):
        """Perform comprehensive image analysis and generate visualizations."""
        if image is None:
            return None, None, None, None
        
        # Detect cells and extract features
        contours, cells, mask = self.detect_cells(image)
        vis_img = image.copy()
        
        # Draw detections and labels
        for cell in cells:
            contour = contours[cell['label'] - 1]
            cv2.drawContours(vis_img, [contour], -1, (0, 255, 0), 2)
            cv2.putText(vis_img, str(cell['label']), 
                       (cell['centroid_x'], cell['centroid_y']),
                       cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1)
        
        # Create DataFrame and calculate summary statistics
        df = pd.DataFrame(cells)
        if not df.empty:
            summary_stats = {
                'total_cells': len(cells),
                'avg_cell_size': df['area'].mean(),
                'std_cell_size': df['area'].std(),
                'avg_circularity': df['circularity'].mean(),
                'cell_density': len(cells) / (image.shape[0] * image.shape[1])
            }
            df = df.assign(**{k: [v] * len(df) for k, v in summary_stats.items()})
        
        # Generate visualizations
        fig = self.generate_analysis_plots(df)
        
        return vis_img, mask, fig, df

    def generate_analysis_plots(self, df):
        """Generate comprehensive analysis plots."""
        if df.empty:
            return None
            
        plt.style.use('dark_background')
        fig = plt.figure(figsize=(15, 10))
        
        # Create subplot grid
        gs = plt.GridSpec(2, 3, figure=fig)
        ax1 = fig.add_subplot(gs[0, 0])
        ax2 = fig.add_subplot(gs[0, 1])
        ax3 = fig.add_subplot(gs[0, 2])
        ax4 = fig.add_subplot(gs[1, :])
        
        # Cell size distribution
        ax1.hist(df['area'], bins=20, color='cyan', edgecolor='black')
        ax1.set_title('Cell Size Distribution')
        ax1.set_xlabel('Area')
        ax1.set_ylabel('Count')
        
        # Area vs Circularity
        scatter = ax2.scatter(df['area'], df['circularity'], 
                            c=df['solidity'], cmap='viridis', alpha=0.6)
        ax2.set_title('Area vs Circularity')
        ax2.set_xlabel('Area')
        ax2.set_ylabel('Circularity')
        plt.colorbar(scatter, ax=ax2, label='Solidity')
        
        # Eccentricity distribution
        ax3.hist(df['eccentricity'], bins=15, color='magenta', edgecolor='black')
        ax3.set_title('Eccentricity Distribution')
        ax3.set_xlabel('Eccentricity')
        ax3.set_ylabel('Count')
        
        # Cell position scatter plot
        scatter = ax4.scatter(df['centroid_x'], df['centroid_y'], 
                            c=df['area'], cmap='plasma', alpha=0.6, s=100)
        ax4.set_title('Cell Positions and Sizes')
        ax4.set_xlabel('X Position')
        ax4.set_ylabel('Y Position')
        plt.colorbar(scatter, ax=ax4, label='Cell Area')
        
        plt.tight_layout()
        return fig

# Create Gradio interface
analyzer = BloodCellAnalyzer()
demo = gr.Interface(
    fn=analyzer.analyze_image,
    inputs=gr.Image(type="numpy"),
    outputs=[
        gr.Image(label="Detected Cells"),
        gr.Image(label="Segmentation Mask"),
        gr.Plot(label="Analysis Plots"),
        gr.DataFrame(label="Cell Data")
    ],
    title="Blood Cell Analysis Tool",
    description="Upload an image to analyze blood cells and extract features."
)

if __name__ == "__main__":
    demo.launch()