OrgaQuant/orgaquant.py at master · TKassis/OrgaQuant · GitHub

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
# import dependencies
import tensorflow as tf
from tensorflow import keras
import streamlit as st
from keras_retinanet import models
from keras_retinanet.utils.image import read_image_bgr, preprocess_image, resize_image, adjust_contrast
from keras_retinanet.utils.visualization import draw_box
import cv2
import os
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

st.title('OrgaQuant: Organoid Quantification')

st.sidebar.header('Settings')

model_path = st.sidebar.text_input("Saved model to use (stored in train_models folder):", value='orgaquant_intestinal_v3')

st.cache(allow_output_mutation=True)
def load_orga_model():
    return models.load_model(os.path.join('trained_models', model_path + '.h5'), backbone_name='resnet50')

model = load_orga_model()


folder_path = st.sidebar.text_input("Path to folder containing images:", value='test_folder')

imagelist=[]

for root, directories, filenames in os.walk(folder_path):
    imagelist = imagelist + [os.path.join(root,x) for x in filenames if x.endswith(('.jpg','.tif','.TIF', '.png', '.jpeg', '.tiff'))]

sample_image = st.sidebar.slider("Sample Image:", min_value=0, max_value=len(imagelist), step=1, value=0)
min_side = st.sidebar.slider("Image Size:", min_value=800, max_value=2000, step=100, value=1200)
st.sidebar.text('Larger "Image Size" allows you to detect smaller orgaoids at the cost of computational demand.')
contrast = st.sidebar.slider("Contrast:", min_value=1.0, max_value=3.0, step=0.25, value=1.5)
st.sidebar.text('Larger "Contrast" can improve detection sometimes.')
threshold = st.sidebar.slider("Confidence Threshold:", min_value=0.0, max_value=1.0, step=0.05, value=0.85)
st.sidebar.text('Use larger "Threshold" to eliminate false positives.')

# load image
#image = read_image_bgr(os.path.join(folder_path, imagelist[sample_image]))
image = read_image_bgr(imagelist[sample_image])

# copy to draw on
draw = image.copy()
draw = cv2.cvtColor(draw, cv2.COLOR_BGR2RGB)

# preprocess image for network
image = adjust_contrast(image,contrast)
image = preprocess_image(image)
image, scale = resize_image(image, min_side=min_side, max_side=2048)

# process image
boxes, scores, labels = model.predict_on_batch(np.expand_dims(image, axis=0))

# correct for image scale
boxes /= scale

num_org = 0
# visualize detections
for box, score, label in zip(boxes[0], scores[0], labels[0]):
    # scores are sorted so we can break
    if score < threshold:
        break
    num_org= num_org + 1
    b = box.astype(int)
    draw_box(draw, b, color=(255, 0, 255))

st.image(draw,use_column_width=True)
st.write("Image name:", imagelist[sample_image])
st.write("Number of organoids detected:", num_org)

# Batch process from here on
st.sidebar.subheader('Batch Processing')

if st.sidebar.button("Process All"):
    progress_bar = st.sidebar.progress(0)
    for i, filename in enumerate(imagelist):
        try:
            #IMAGE_PATH = os.path.join(root,filename)
            IMAGE_PATH = filename
            # load image
            image = read_image_bgr(IMAGE_PATH)

            # copy to draw on
            draw = image.copy()
            draw = cv2.cvtColor(draw, cv2.COLOR_BGR2RGB)

            # preprocess image for network
            image = adjust_contrast(image,contrast)
            image = preprocess_image(image)
            image, scale = resize_image(image, min_side=min_side, max_side=2048)

            # process image
            boxes, scores, labels = model.predict_on_batch(np.expand_dims(image, axis=0))

            # correct for image scale
            boxes /= scale

            out = np.empty((0,4), dtype=np.float32)

            # visualize detections
            for box, score, label in zip(boxes[0], scores[0], labels[0]):
                # scores are sorted so we can break
                if score < threshold:
                    break
                out = np.append(out, box.reshape(1,4), axis=0)

                b = box.astype(int)
                draw_box(draw, b, color=(255, 0, 255))

            output = pd.DataFrame(out,columns=['x1', 'y1', 'x2', 'y2'], dtype=np.int16)
            output['Diameter 1 (Pixels)'] = output['x2'] - output['x1']
            output['Diameter 2 (Pixels)'] = output['y2'] - output['y1']
            output.to_csv(IMAGE_PATH + '.csv', index=False)
            plt.imsave(IMAGE_PATH + '_detected.png', draw)
            progress_bar.progress((i+1)/len(imagelist))
        except:
            pass

    st.success('Analysis complete!')

else:
    st.sidebar.text("Click above to process all images.")

st.sidebar.markdown('If you find this helpful for your research please cite: \
_[OrgaQuant: Human Intestinal Organoid Localization and Quantification Using Deep Convolutional Neural Networks](https://www.nature.com/articles/s41598-019-48874-y)_.')

st.sidebar.markdown('For assistance or to report bugs, please raise an issue on [GitHub](https://github.com/TKassis/OrgaQuant/issues)')