def custom_crop_last(slide_, y, batch_to_predict, valid_bit_list, x_max,
crop_size):
for x in range(0, x_max - 1):
image_crop = slide.read_slide_crop(
slide_, x * crop_size,
slide.get_slide_size(slide_)[1] - crop_size,
crop_size).convert('RGB')
resized_crop = slide.resize_image_a(image_crop, 224, 224)
resized_crop_np = np.asarray(resized_crop, np.float32)
if filter.check_valid(resized_crop_np):
valid_bit_list[y * x_max + x] = 1
else:
valid_bit_list[y * x_max + x] = 0
batch_to_predict[y * x_max + x] = resized_crop_np
image_crop = slide.read_slide_crop(
slide_,
slide.get_slide_size(slide_)[0] - crop_size,
slide.get_slide_size(slide_)[1] - crop_size, crop_size).convert('RGB')
resized_crop = slide.resize_image_a(image_crop, 224, 224)
resized_crop_np = np.asarray(resized_crop, np.float32)
if filter.check_valid(resized_crop_np):
valid_bit_list[y * x_max + x_max - 1] = 1
else:
valid_bit_list[y * x_max + x_max - 1] = 0
batch_to_predict[y * x_max + x_max - 1] = resized_crop_np
def custom_crop_last(slide_, y, batch_to_predict, valid_bit_matrix, x_max):
batch_to_predict[y] = []
valid_bit_matrix[y] = []
for x in range(0, x_max - 1):
image_crop = slide.read_slide_crop(
slide_, x * CROP_SIZE,
slide.get_slide_size(slide_)[1] * CROP_SIZE).convert('RGB')
resized_crop = slide.resize_image_a(image_crop, 224, 224)
resized_crop_np = np.asarray(resized_crop, np.float32)
if filter.check_valid(resized_crop_np):
valid_bit_matrix[y].append(1)
else:
valid_bit_matrix[y].append(0)
batch_to_predict[y].append(resized_crop_np)
image_crop = slide.read_slide_crop(
slide_,
slide.get_slide_size(slide_)[0] - CROP_SIZE,
slide.get_slide_size(slide_)[1] * CROP_SIZE).convert('RGB')
resized_crop = slide.resize_image_a(image_crop, 224, 224)
resized_crop_np = np.asarray(resized_crop, np.float32)
if filter.check_valid(resized_crop_np):
valid_bit_matrix[y].append(1)
else:
valid_bit_matrix[y].append(0)
batch_to_predict[y].append(resized_crop_np)
def mc_predict_from_path(iterations, file_path, pred_folder, drop_rate,
crop_size):
mc_predictions = []
slidename = str(path.basename(file_path))
flattened_image_name = slidename.replace(".svs",
"_" + str(crop_size) + ".bin", 1)
path_ = path.join(MAP_FOLDER, flattened_image_name)
if os.path.isfile(path_):
batch_to_predict_np = read_blob(flattened_image_name, MAP_FOLDER)
else:
slide_ = slide.open_slide(file_path)
slide_size = slide.get_slide_size(slide_)
save_blob(slide_size, slidename.replace(".svs", ".info", 1),
MAP_FOLDER)
log.print_info("Slide size : " + str(slide_size))
pool = []
x_max = ceil(slide_size[0] / crop_size)
y_max = ceil(slide_size[1] / crop_size)
log.print_info("Matrix size : " + str(y_max) + " " + str(x_max))
batch_to_predict = [None] * x_max * y_max
valid_bit_list = [None] * x_max * y_max
for y in range(0, y_max - 1):
pool.append(
Thread(target=custom_crop,
args=(slide_, y, batch_to_predict, valid_bit_list,
x_max, crop_size)))
pool[-1].start()
t = Thread(target=custom_crop_last,
args=(slide_, y_max - 1, batch_to_predict, valid_bit_list,
x_max, crop_size))
t.start()
for p in pool:
p.join()
t.join()
valid_bit_np = np.asarray(valid_bit_list)
save_blob(
valid_bit_np,
slidename.replace(".svs", "") + "_" + str(crop_size) + ".vbit",
MAP_FOLDER)
batch_to_predict_np = np.asarray(batch_to_predict)
save_blob(batch_to_predict_np, flattened_image_name, MAP_FOLDER)
del valid_bit_list
del valid_bit_np
del batch_to_predict
for i in range(0, iterations):
log.print_info("Step " + str(i))
prediction_list = cnn.predict_from_model_multithread(
batch_to_predict_np, drop_rate)
mc_predictions.append(np.asarray(prediction_list))
save_blob(
mc_predictions,
slidename.replace(".svs", "") + "_" + str(crop_size) + "_" +
str(iterations) + ".pred", pred_folder)
del prediction_list
gc.collect()
del batch_to_predict_np
def get_prediction_matrix_multithread(slidename):
slide_ = slide.open_slide(os.path.join(MAP_FOLDER, slidename))
slide_size = slide.get_slide_size(slide_)
log.print_info("Slide size : " + str(slide_size))
pool = []
x_max = ceil(slide_size[0] / CROP_SIZE)
y_max = ceil(slide_size[1] / CROP_SIZE)
log.print_info("Matrix size : " + str(y_max) + " " + str(x_max))
batch_to_predict = [None] * y_max
valid_bit_matrix = [None] * y_max
for y in range(0, y_max - 1):
pool.append(
Thread(target=custom_crop,
args=(slide_, y, batch_to_predict, valid_bit_matrix,
x_max)))
pool[-1].start()
t = Thread(target=custom_crop_last,
args=(slide_, y_max - 1, batch_to_predict, valid_bit_matrix,
x_max))
t.start()
for p in pool:
p.join()
t.join()
batch_to_predict_np = np.asarray(batch_to_predict)
print(batch_to_predict_np.shape)
prediction_list = [[]] * batch_to_predict_np.shape[0]
for i in range(0, batch_to_predict_np.shape[0]):
cnn.predict_from_model_multithread(batch_to_predict_np[i][:][:][:],
prediction_list, i)
prediction_matrix = np.asarray(prediction_list)
print(prediction_matrix.shape)
return utils.list_np_to_pil(
batch_to_predict_np, utils.COLOR), prediction_matrix, valid_bit_matrix