{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":61446,"databundleVersionId":6962461,"sourceType":"competition"},{"sourceId":7178241,"sourceType":"datasetVersion","datasetId":4148557}],"dockerImageVersionId":30616,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"for dicussion refer to\nhttps://www.kaggle.com/competitions/blood-vessel-segmentation/discussion/456118#2557787\n![Selection_999(4369).png](attachment:24c13f32-7376-4b48-832a-ce12bc80ecd9.png)","metadata":{},"attachments":{"24c13f32-7376-4b48-832a-ce12bc80ecd9.png":{"image/png":"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"}}},{"cell_type":"code","source":"import torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nimport numpy as np\nimport cv2\nimport math\n\nimport timm\nprint('timm', timm.__version__)\nfrom timm.models.resnet import *\n\nimport matplotlib.pyplot as plt\nfrom IPython.display import Image, display\n\n###########################################33\n#helper\ndef image_show_norm(name, image, min=None, max=None, type='bgr', resize=1):\n\tif max is None: max = image.max()\n\tif min is None: min = image.min()\n\tif type == 'rgb': image = np.ascontiguousarray(image[:, :, ::-1])\n\n\tH, W = image.shape[0:2]\n\tcv2.namedWindow(name, cv2.WINDOW_GUI_NORMAL)  # WINDOW_NORMAL\n\tcv2.imshow(name, (np.clip((image - min) / (max - min), 0, 1) * 255).astype(np.uint8))\n\tcv2.resizeWindow(name, round(resize * W), round(resize * H))\n\n\n\ndef np_sigmoid(x):\n\treturn 1 / (1 + np.exp(-x))\n\nprint('IMPORT OK')","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2023-12-11T19:32:02.045595Z","iopub.execute_input":"2023-12-11T19:32:02.045988Z","iopub.status.idle":"2023-12-11T19:32:06.890363Z","shell.execute_reply.started":"2023-12-11T19:32:02.045958Z","shell.execute_reply":"2023-12-11T19:32:06.889494Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"root_dir = \\\n\t'/kaggle/input/blood-vessel-segmentation'\n\t#'/home/user/share1/kaggle/2023/blood-vessel-segmentation'\n\n\nIMAGE_SIZE  = 640\nIMAGE_DEPTH = 32\n\n# Net\nclass MyDecoderBlock3d(nn.Module):\n\tdef __init__(\n\t\tself,\n\t\tin_channel,\n\t\tskip_channel,\n\t\tout_channel,\n\t):\n\t\tsuper().__init__()\n\t\tself.conv1 = nn.Sequential(\n\t\t\tnn.Conv3d(in_channel + skip_channel, out_channel, kernel_size=3, padding=1, bias=False),\n\t\t\tnn.BatchNorm3d(out_channel, eps=1e-4),\n\t\t\tnn.ReLU(inplace=True),\n\t\t)\n\t\tself.attention1 = nn.Identity()\n\t\tself.conv2 = nn.Sequential(\n\t\t\tnn.Conv3d(out_channel,out_channel,kernel_size=3, padding=1, bias=False),\n\t\t\tnn.BatchNorm3d(out_channel, eps=1e-4),\n\t\t\tnn.ReLU(inplace=True),\n\t\t)\n\t\tself.attention2 = nn.Identity()\n\n\tdef forward(self, x, skip=None):\n\t\tx = F.interpolate(x, scale_factor=2, mode='nearest')\n\t\tif skip is not None:\n\t\t\tx = torch.cat([x, skip], dim=1)\n\t\t\tx = self.attention1(x)\n\t\tx = self.conv1(x)\n\t\tx = self.conv2(x)\n\t\tx = self.attention2(x)\n\t\treturn x\n\nclass MyUnetDecoder3d(nn.Module):\n\tdef __init__(self,\n\t\t\t\t in_channel,\n\t\t\t\t skip_channel,\n\t\t\t\t out_channel,\n\t\t\t\t ):\n\t\tsuper().__init__()\n\t\tself.center = nn.Identity()\n\n\t\ti_channel = [in_channel, ] + out_channel[:-1]\n\t\ts_channel = skip_channel\n\t\to_channel = out_channel\n\t\tblock = [\n\t\t\tMyDecoderBlock3d(i, s, o)\n\t\t\tfor i, s, o in zip(i_channel, s_channel, o_channel)\n\t\t]\n\t\tself.block = nn.ModuleList(block)\n\n\tdef forward(self, feature, skip):\n\t\td = self.center(feature)\n\t\tdecode = []\n\t\tfor i, block in enumerate(self.block):\n\t\t\t#print(i, d.shape, skip[i].shape if skip[i] is not None else 'none')\n\t\t\t#print(block.conv1[0])\n\t\t\t#print('')\n\n\t\t\ts = skip[i]\n\t\t\td = block(d, s)\n\t\t\tdecode.append(d)\n\t\tlast = d\n\t\treturn last, decode\n\n#--------------------------------------\n\nclass Net(nn.Module):\n\tdef __init__(self, ):\n\t\tsuper().__init__()\n\t\tencoder_dim = [64, 256, 512, 1024, 2048]\n\t\tdecoder_dim = [256, 128, 128, 64, 32 ]\n\n\t\t#self.encoder = seresnext26d_32x4d(pretrained=True, in_chans=3)\n\t\tself.encoder = resnet50d(pretrained=False, in_chans=3)\n\t\tself.add_conv1 = nn.Sequential(\n\t\t\tnn.Conv3d(1, 32, kernel_size=3, padding=1, bias=False),\n\t\t\tnn.BatchNorm3d(32, eps=1e-4),\n\t\t\tnn.ReLU(inplace=True),\n\t\t\tnn.Conv3d(32, 32, kernel_size=3, padding=1, bias=False),\n\t\t\tnn.BatchNorm3d(32, eps=1e-4),\n\t\t\tnn.ReLU(inplace=True),\n\t\t\tnn.Conv3d(32, 32, kernel_size=3, padding=1, bias=False),\n\t\t\tnn.BatchNorm3d(32, eps=1e-4),\n\t\t\tnn.ReLU(inplace=True),\n\t\t)\n\n\n\t\tself.decoder = MyUnetDecoder3d(\n\t\t\tin_channel  = encoder_dim[-1],\n\t\t\tskip_channel= encoder_dim[:-1][::-1]+[32],\n\t\t\tout_channel = decoder_dim,\n\t\t)\n\t\tself.vessel = nn.Conv3d(decoder_dim[-1], 1, kernel_size=1)\n\n\t\t#just a simple demo. please improve this\n\t\tself.convert_3d = torch.nn.ModuleList([\n\t\t\t\tnn.Identity(),\n\t\t\t\tnn.Sequential(nn.Conv3d(  64,  64,kernel_size=( 2,1,1),stride=( 2,1,1),padding=(0,0,0)),  nn.BatchNorm3d(  64, eps=1e-4), nn.ReLU(inplace=True), ),\n\t\t\t\tnn.Sequential(nn.Conv3d( 256, 256,kernel_size=( 4,1,1),stride=( 4,1,1),padding=(0,0,0)),  nn.BatchNorm3d( 256, eps=1e-4), nn.ReLU(inplace=True), ),\n\t\t\t\tnn.Sequential(nn.Conv3d( 512, 512,kernel_size=( 8,1,1),stride=( 8,1,1),padding=(0,0,0)),  nn.BatchNorm3d( 512, eps=1e-4), nn.ReLU(inplace=True), ),\n\t\t\t\tnn.Sequential(nn.Conv3d(1024,1024,kernel_size=(16,1,1),stride=(16,1,1),padding=(0,0,0)),  nn.BatchNorm3d(1024, eps=1e-4), nn.ReLU(inplace=True), ),\n\t\t\t\tnn.Sequential(nn.Conv3d(2048,2048,kernel_size=(32,1,1),stride=(32,1,1),padding=(0,0,0)),  nn.BatchNorm3d(2048, eps=1e-4), nn.ReLU(inplace=True), ),\n\t\t])\n\n\tdef forward(self, subvolume):\n\t\txx = subvolume\n\t\tB, D, H, W = xx.shape\n\n\t\tx = xx.reshape(B * D, 1, H, W)\n\t\tx = x.expand(-1, 3, -1, -1)\n\n\t\tencode = []\n\t\txx = self.add_conv1(xx.unsqueeze(1)); encode.append(xx)\n\n\t\te = self.encoder\n\t\tx = e.conv1(x)\n\t\tx = e.bn1(x)\n\t\tx = e.act1(x);   encode.append(x)\n\t\tx = F.avg_pool2d(x, kernel_size=2, stride=2)\n\t\tx = e.layer1(x); encode.append(x)\n\t\tx = e.layer2(x); encode.append(x)\n\t\tx = e.layer3(x); encode.append(x)\n\t\tx = e.layer4(x); encode.append(x)\n\n\t\tfor i, x in enumerate(encode):\n\t\t\tif i == 0: continue\n\t\t\tx = encode[i]\n\t\t\tBD, c, h, w = x.shape\n\t\t\tx = x.reshape(B, D, c, h, w)\n\t\t\tx = x.transpose(1, 2)\n\t\t\tx = self.convert_3d[i](x)\n\t\t\tencode[i] = x\n\n\t\tlast, decode = self.decoder(\n\t\t\tfeature=encode[-1], skip=encode[:-1][::-1]\n\t\t)\n\n\t\tvessel = self.vessel(last).squeeze(1)\n\t\tvessel = torch.sigmoid(vessel.float())\n\t\treturn vessel\n\n#############################################################################33\n\ndef run_check_net():\n\theight, width = 480, 480\n\tdepth = 32\n\tbatch_size = 2\n\n\tsubvolume = torch.from_numpy(np.random.uniform(0, 1, (batch_size, depth, height, width))).float().cuda()\n\tnet = Net().cuda()\n\n\n\twith torch.no_grad():\n\t\twith torch.cuda.amp.autocast(enabled=True):\n\t\t\tvessel = net(subvolume)\n\tprint(subvolume.shape)\n\tprint(vessel.shape)\n\nrun_check_net()\nprint('NET OK!!!')","metadata":{"execution":{"iopub.status.busy":"2023-12-11T19:32:06.892118Z","iopub.execute_input":"2023-12-11T19:32:06.892400Z","iopub.status.idle":"2023-12-11T19:32:16.793835Z","shell.execute_reply.started":"2023-12-11T19:32:06.892376Z","shell.execute_reply":"2023-12-11T19:32:16.792923Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Data\ndef norm_by_min_max(x, xmin, xmax, alpha=0.01):\n\txmin = float(xmin)\n\txmax = float(xmax)\n\tx = (x-xmin)/(xmax-xmin)\n\tif 1:\n\t\tx[x>1]=(x[x>1]-1)*alpha +1\n\t\tx[x<0]=(x[x<0])*alpha\n\t#x = np.clip(x,0,1)\n\treturn x\n\ndef load_dummy_data():\n\t#x,y,z = 740,470, 389\n\tname = 'kidney_3_sparse'\n\tx,y,z = 420, 150, 356\n\n\tsubvolume = []\n\ttruth = []\n\tfor i in range(z,z+IMAGE_DEPTH):\n\t\timage_file = f'{root_dir}/train/{name}/images/{i:04d}.tif'\n\t\tm = cv2.imread(image_file, cv2.IMREAD_UNCHANGED)\n\t\tsubvolume.append(m)\n\n\t\timage_file = f'{root_dir}/train/{name}/labels/{i:04d}.tif'\n\t\tv = cv2.imread(image_file, cv2.IMREAD_GRAYSCALE)\n\t\ttruth.append(v)\n\n\tsubvolume = np.stack(subvolume)\n\tsubvolume = subvolume[:,y:y+IMAGE_SIZE,x:x+IMAGE_SIZE]\n\txmin, xmax = (18806, 21903) #precompute low and high percentile\n\tsubvolume = norm_by_min_max(subvolume, xmin, xmax)\n\n\ttruth = np.stack(truth)\n\ttruth = truth[:,y:y+IMAGE_SIZE,x:x+IMAGE_SIZE]\n\ttruth = truth//255\n\n\treturn subvolume, truth\n\n\ndef run_demo():\n\n\tsubvolume, truth = load_dummy_data()\n\n\tnet = Net()\n\tcheckpoint_file ='/kaggle/input/2d-to-3d-demo-data/00000828.pth'\n\tstate_dict = torch.load(checkpoint_file, map_location=lambda storage, loc: storage)['state_dict']\n\tprint(net.load_state_dict(state_dict, strict=True))  # True\n\tnet = net.eval()\n\tnet = net.cuda()\n\n\ttensor = torch.from_numpy(subvolume).float().cuda()\n\twith torch.no_grad():\n\t\twith torch.cuda.amp.autocast(enabled=True):\n\t\t\tvessel = net(tensor.unsqueeze(0)).squeeze(0)\n\tvessel = vessel.float().data.cpu().numpy()\n\n\n\t#--------------------------------------------------------------\n\tif 0: #visualisation (run offline. does not work in kaggle\n\n\t\tD,H,W = subvolume.shape\n\t\tsubvolume = np.clip(subvolume,0,1)\n\t\tx_mean = subvolume.mean(0)\n\t\tp_mean = vessel.mean(0)\n\t\ty_mean = truth.mean(0)\n\n\t\t#---\n\t\tdef add_contrast_for_x(x_mean):\n\t\t\tx_mean = np_sigmoid((1-x_mean)*5)\n\t\t\tx_mean = 1- (x_mean-x_mean.min())/(x_mean.max()-x_mean.min())\n\t\t\treturn x_mean\n\n\t\tdef add_contrast_for_py(p_mean, y_mean):\n\t\t\tzero = np.zeros((H,W))\n\t\t\tpy_mean = np.dstack([zero,p_mean,y_mean])\n\t\t\tpy_mean = ((py_mean-py_mean.min())/(py_mean.max()-py_mean.min()))**0.25\n\t\t\treturn py_mean\n\t\t#---\n\n\t\tx_mean  = add_contrast_for_x(x_mean)\n\t\tpy_mean = add_contrast_for_py(p_mean, y_mean)\n\t\timage_show_norm('x_mean',x_mean, min=0,max=1,resize=1)\n\t\timage_show_norm('py_mean',py_mean, min=0,max=1,resize=1)\n\t\tcv2.waitKey(0)\n\telse:\n\t\tdisplay(Image(filename='../input/2d-to-3d-demo-data/Selection_999(4367).png'))\n\t\tdisplay(Image(filename='../input/2d-to-3d-demo-data/Selection_999(4368).png'))\n \n\n\t#--------------------------------------------------------------\n\tif 0: #3d visualisation (run offline. does not work in kaggle)\n\t\timport pyvista as pv\n\t\ty = truth>0\n\t\tp = vessel>0.5\n\t\thit = y*p\n\t\tfp = (1-y)*p\n\t\tmiss = y*(1-p)\n\n\t\tpl = pv.Plotter()\n\n\t\tmhit  = pv.PolyData(np.stack(np.where(hit > 0.1)).T).glyph(geom=pv.Cube())\n\t\tmfp   = pv.PolyData(np.stack(np.where(fp  > 0.1)).T).glyph(geom=pv.Cube())\n\t\tmmiss = pv.PolyData(np.stack(np.where(miss > 0.1)).T).glyph(geom=pv.Cube())\n\t\tpl.add_mesh(mhit,  color='yellow' )\n\t\tpl.add_mesh(mfp,   color='green' )\n\t\tpl.add_mesh(mmiss, color='red' )\n\t\tpl.show()\n\telse:\n\t\tprint('why cannot play GIF ???????')        \n\t\tdisplay(Image(filename='../input/2d-to-3d-demo-data/Peek 2023-12-12 02-38.gif', format='png'))\n\t\tdisplay(Image(filename='../input/2d-to-3d-demo-data/Peek 2023-12-12 02-36.gif', format='png'))\n    \n\nrun_demo()\nprint('DEMO OK!!!')","metadata":{"execution":{"iopub.status.busy":"2023-12-11T19:33:21.776675Z","iopub.execute_input":"2023-12-11T19:33:21.777278Z","iopub.status.idle":"2023-12-11T19:33:37.109962Z","shell.execute_reply.started":"2023-12-11T19:33:21.777231Z","shell.execute_reply":"2023-12-11T19:33:37.109091Z"},"trusted":true},"execution_count":null,"outputs":[]}]}