{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"
\n\n
\n\n

๐Ÿซ๐Ÿซ€ EDA โ€“ HuBMAP+HPA โ€“ ์žฅ๊ธฐ ๋ถ„ํ•  ๋ฐ์ดํ„ฐ EDA ๐Ÿซ€๐Ÿซ

\n
CREATED BY: DARIEN SCHETTLER

\nAll credit to belongs to the original author DARIEN SCHETTLER!
\n\n\n\n
\n\n---\n\n
\n\n
\n ๐Ÿ›‘   WARNING:

\n 1. EDA ํ˜„์žฌ์ง„ํ–‰์ค‘!!! (ํ˜„์žฌ ์ปค๋„)

\n 2. [๋ฒ ์ด์Šค๋ผ์ธ]_๋งํฌ ํ˜„์žฌ์ง„ํ–‰์ค‘!!! \n
\n\n\n
\n ๐Ÿ‘   ์ด ๊ธ€์ด ๋„์›€์ด ๋˜์…จ๋‹ค๋ฉด   ๐Ÿ‘

ํˆฌํ‘œ ๋ถ€ํƒ๋“œ๋ ค์š”!

ํˆฌํ‘œ ํ•ด์ฃผ์‹ ๋‹ค๋ฉด ์ œ๊ฐ€ ์ด๋Ÿฌํ•œ ์ž‘์—…์„ ํ•˜๋Š”๋ฐ ๋งŽ์€ ๋„์›€๊ณผ ์‘์›์ด ๋  ๊ฒƒ ๊ฐ™์Šต๋‹ˆ๋‹ค ๐Ÿ˜…\n
\n\n\n","metadata":{}},{"cell_type":"markdown","source":"

\n\n

\n\n

์ฝ˜ํ…์ธ  ๋ชฉ๋ก

\n\n---\n\n

0    ๋ถˆ๋Ÿฌ์˜ค๊ธฐ

\n\n---\n\n

1    ๋ฐฐ๊ฒฝ ์ง€์‹

\n\n---\n\n

2    ์ค€๋น„

\n\n---\n\n

3    ๋„์›€ ํ•จ์ˆ˜๋“ค

\n\n---\n\n

4    ๋ฐ์ดํ„ฐ ๋ถ„์„,ํƒํ—˜

\n\n---\n\n

5    ๋ชจ๋ธ๋ง

\n\n---\n\n

9    ๋ถ€๋ก

\n\n---","metadata":{}},{"cell_type":"markdown","source":"
\n\n\n\n

0  ๋ถˆ๋Ÿฌ์˜ค๊ธฐ    ⤒

","metadata":{}},{"cell_type":"code","source":"print(\"\\n... IMPORTS STARTING ...\\n\")\n\nprint(\"\\n\\tVERSION INFORMATION\")\n# Machine Learning and Data Science Imports\nimport tensorflow as tf; print(f\"\\t\\tโ€“ TENSORFLOW VERSION: {tf.__version__}\");\nimport tensorflow_hub as tfhub; print(f\"\\t\\tโ€“ TENSORFLOW HUB VERSION: {tfhub.__version__}\");\nimport tensorflow_addons as tfa; print(f\"\\t\\tโ€“ TENSORFLOW ADDONS VERSION: {tfa.__version__}\");\nimport tensorflow_io as tfio; print(f\"\\t\\tโ€“ TENSORFLOW I/O VERSION: {tfio.__version__}\");\nimport pandas as pd; pd.options.mode.chained_assignment = None;\nimport numpy as np; print(f\"\\t\\tโ€“ NUMPY VERSION: {np.__version__}\");\nimport sklearn; print(f\"\\t\\tโ€“ SKLEARN VERSION: {sklearn.__version__}\");\nfrom sklearn.preprocessing import RobustScaler, PolynomialFeatures\nfrom pandarallel import pandarallel; pandarallel.initialize();\nfrom sklearn.model_selection import GroupKFold, StratifiedKFold\nfrom scipy.spatial import cKDTree\n\n# # RAPIDS\n# import cudf, cupy, cuml\n# from cuml.neighbors import NearestNeighbors\n# from cuml.manifold import TSNE, UMAP\n\n# Built In Imports\nfrom kaggle_datasets import KaggleDatasets\nfrom collections import Counter\nfrom datetime import datetime\nfrom glob import glob\nimport warnings\nimport requests\nimport hashlib\nimport imageio\nimport IPython\nimport sklearn\nimport urllib\nimport zipfile\nimport pickle\nimport random\nimport shutil\nimport string\nimport json\nimport math\nimport time\nimport gzip\nimport ast\nimport sys\nimport io\nimport os\nimport gc\nimport re\n\n# Visualization Imports\nfrom matplotlib.colors import ListedColormap\nfrom matplotlib.patches import Rectangle\nimport matplotlib.patches as patches\nimport plotly.graph_objects as go\nimport matplotlib.pyplot as plt\nfrom tqdm.notebook import tqdm; tqdm.pandas();\nimport plotly.express as px\nimport seaborn as sns\nfrom PIL import Image, ImageEnhance\nimport matplotlib; print(f\"\\t\\tโ€“ MATPLOTLIB VERSION: {matplotlib.__version__}\");\nfrom matplotlib import animation, rc; rc('animation', html='jshtml')\nimport plotly\nimport PIL\nimport cv2\n\nimport plotly.io as pio\nprint(pio.renderers)\n\ndef seed_it_all(seed=7):\n \"\"\" Attempt to be Reproducible \"\"\"\n os.environ['PYTHONHASHSEED'] = str(seed)\n random.seed(seed)\n np.random.seed(seed)\n tf.random.set_seed(seed)\n\n \nprint(\"\\n\\n... IMPORTS COMPLETE ...\\n\")\n\n","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-06-28T16:16:13.543813Z","iopub.execute_input":"2022-06-28T16:16:13.545025Z","iopub.status.idle":"2022-06-28T16:16:25.527252Z","shell.execute_reply.started":"2022-06-28T16:16:13.544871Z","shell.execute_reply":"2022-06-28T16:16:25.526216Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n\n\n

1  ๋ฐฐ๊ฒฝ ์ง€์‹    ⤒

\n\n---\n","metadata":{}},{"cell_type":"markdown","source":"
\n\n

1.1 ๊ธฐ๋ณธ ๋Œ€ํšŒ ์ •๋ณด

\n\n---\n\n
์ฃผ์š” ๊ณผ์ œ ์„ค๋ช…\n\nIn this competition, youโ€™ll identify and segment functional tissue units **(FTUs)** across five human organs.\n* You'll build your model using a dataset of tissue section images, with the best submissions segmenting FTUs as accurately as possible.\n* The five human organs we are segmenting are:\n * **์ „๋ฆฝ**\n * **๋น„์žฅ**\n * **ํ**\n * **์‹ ์žฅ**\n * **๋Œ€์žฅ**\n\n
๊ธฐ๋ณธ ๋ฐฐ๊ฒฝ์ง€์‹\n\nโ€œlife hacks - 37์กฐ๊ฐœ์˜ ์„ธํฌ๋“ค์˜ ๊ด€๊ณ„์™€ ๊ทธ ๊ธฐ๋Šฅ๋“ค์„ ํŒŒ์•…ํ•˜๋Š” ๊ฒƒ. ( ์„ธํฌ๋“ค์˜ ๊ตฌ์„ฑ )\n\nHPA ์ „ ์บ๊ธ€๋Œ€ํšŒ - ์‹ ์žฅ์˜ ์ฃผ์š” ๊ธฐ๋Šฅ์„ ์ˆ˜ํ–‰ํ•˜๋Š” ์„ธํฌ ์กฐ์ง ๋‹จ์œ„(์ด์›ƒ)์„ ์„ธ๊ทธ๋ฉ˜ํ…Œ์ด์…˜ํ•˜๋Š” ๊ณผ์ œ. ์‹ ์žฅ์—๋Š” 1๋ฐฑ๋งŒ๊ฐœ์˜ ์‚ฌ๊ตฌ์ฒด FTU๊ฐ€ ์กด์žฌํ•œ๋‹ค.\n์‹ ์žฅ์—์„œ FTU๋ฅผ ์„ธ๊ทธ๋ฉ˜ํ…Œ์ด์…˜ํ–ˆ๋˜ ๋ฐฉ๋ฒ•๊ณผ ๊ธฐ์ˆ ๋“ค์„ ๋‹ค์–‘ํ•œ ์กฐ์ง(๋Œ€์ƒํ™•๋Œ€)๊ณผ ๋‹ค๋ฅธ ๋ฐ์ดํ„ฐ์…‹(๋‹ค๋ฅธ ํ˜•์‹)์—๋„ ์ ์šฉํ•ด๋ณด๋ ค๊ณ  ํ•œ๋‹ค.(์ผ๋ฐ˜ํ™”)\n\n* ์‚ฌ๊ตฌ์ฒด - \n\n์ฃผ์ตœ์‚ฌ์ธ HuBMAP์€ ๋ถ„์ž ๋‹จ์œ„์˜ ์ธ๊ฐ„ ์ƒ์ฒด ์ง€๋„์— ๊ด€์‹ฌ์ด ๋งŽ์€ ๊ธฐ๊ด€์ด๋‹ค.\n๋˜ ๋‹ค๋ฅธ ์ฃผ์ตœ์‚ฌ์ธ HPA๋Š” ์ธ๊ฐ„์˜ ์„ธํฌ, ์กฐ์ง ๋ฐ ๊ธฐ๊ด€์—์„œ ๋‹จ๋ฐฑ์งˆ์ด ๋ฐœํ˜„๋˜๋Š” ์›๋ฆฌ๋ฅผ ํƒ๊ตฌํ•˜๋Š” ๊ธฐ๊ด€์ด๋‹ค.\n\n

","metadata":{}},{"cell_type":"markdown","source":"
\n\n

1.2 COMPETITION EVALUATION

\n\n---\n\n
GENERAL EVALUATION INFORMATION\n\nDice Coefficient\n \n$$\n\\frac{2 * |X \\cap Y|}{|X| + |Y|}\n$$\n\n - ์ด๋ฏธ์ง€๋“ฑ์˜ Segmentation ์—์„œ ์“ฐ์ด๋Š” ์ง€ํ‘œ\n\n\n\n\n
\n\n
SUBMISSION FILE INFORMATION\n\n* '1 3' 1์—์„œ ์‹œ์ž‘ํ•ด์„œ 3๊นŒ์ง€์˜ ์˜์—ญ์„ ์˜๋ฏธํ•จ (1,2,3).\n* 0๊ณผ 1๋กœ ์ด๋ฃจ์–ด์ง„ ์˜์—ญ์˜ ๋ฐ”์ด๋„ˆ๋ฆฌ ์ฝ”๋”ฉ\n * ๋ชจ๋“  ์ด๋ฏธ์ง€๋Š” ์˜์—ญ์„ ํ‘œ์‹œํ•˜๋Š” ๋งˆ์Šคํฌ์™€ ๊ฐ™์ด ๊ตฌ์„ฑ์ด ๋˜์–ด์žˆ๋‹ค.\n * 0์€ ๋งˆํ‚น์ด ์•ˆ๋˜์–ด์žˆ๋Š” ์˜์—ญ\n * 1์€ ๋งˆํ‚น์ด ๋˜์–ด์žˆ๋Š” ์˜์—ญ\n\n\"1 3 10 5\" -> 1,2,3 + 10,11,12,13,14\n\n```html\nid,rle\n1,1 1 5 1\n2,1 1\n3,1 1\n4,1 5 2 17\netc.\n```\n\n
IS THIS A CODE COMPETITION?\n\nYES\n\n
","metadata":{}},{"cell_type":"markdown","source":"
\n\n

1.3 DATASET OVERVIEW

\n\n---\n\n
GENERAL INFORMATION\n\n๋Œ€ํšŒ์˜ ๋ชฉ์ ์€ ๋‹ค์–‘ํ•œ ๊ธฐ๊ด€์„ ์ดฌ์˜ํ•œ ์‚ฌ์ง„์—์„œ ๊ฐ๊ฐ FTU(functional tissue unit)์˜ ์œ„์น˜๋ฅผ ์•Œ์•„๋‚ด๋Š” ๊ฒƒ. \n\nํ›ˆ๋ จ **์–ด๋…ธํ…Œ์ด์…˜์€ RLE-์ธ์ฝ”๋”ฉ ๋งˆ์Šคํฌ๋กœ ์ œ๊ณต๋œ๋‹ค.**, ์ด๋ฏธ์ง€๋Š” **16-bit**, **ํšŒ์ƒ‰ ์Šค์ผ€์ผ**, **PNG ํฌ๋งท์œผ๋กœ ๊ตฌ์„ฑ**.\n\n๋‹ค์–‘ํ•œ ๋ฐ์ดํ„ฐ (2๊ฐ€์ง€ ์†Œ์Šค)๋กœ ๊ตฌ์„ฑ์ด ๋˜์–ด์žˆ๋Š” ๊ฒƒ์€ ๋‹ค์–‘ํ•œ ์†Œ์Šค์—์„œ ๋‚˜์˜ค๋Š” ๋‹ค์–‘ํ•œ ํ”„๋กœ์ฝœ๊ณผ ํ•ด์ƒ๋„์˜ ์ด๋ฏธ์ง€์— ๋Œ€์ฒ˜ํ•  ์ˆ˜ ์žˆ๋Š”์ง€๋ฅผ ํ™•์ธํ•˜๊ธฐ ์œ„ํ•œ ๊ฒƒ\n\nThe data is sourced as following:\n* ํ›ˆ๋ จ๋ฐ์ดํ„ฐ๋Š” ๊ฐœ๋ฐฉ๋œ HPA ๋ฐ์ดํ„ฐ\n* ๊ณต๊ฐœ๋œ ํ…Œ์ŠคํŠธ์…‹์€ ๋น„๊ณต๊ฐœ๋œ HPA data + HuBMAP data\n* ๋น„๊ณต๊ฐœ ํ…Œ์ŠคํŠธ์…‹์€ **HuBMAP data๋งŒ์œผ๋กœ ๊ตฌ์„ฑ**. \n\n
\n\n** ํ›ˆ๋ จ๋ฐ์ดํ„ฐ์™€ ํ…Œ์ŠคํŠธ๋ฐ์ดํ„ฐ์˜ ์›์ฒœ๊ณผ ํŠน์ง•์ด ๋‹ค๋ฅด๊ธฐ๋•Œ๋ฌธ์— ์ผ๋ฐ˜ํ™”๋œ ๋ชจ๋ธ์„ ๊ตฌ์„ฑํ•˜๋Š”๊ฒŒ ์ด๋ฒˆ ๋Œ€ํšŒ์˜ ํ‚ค ํฌ์ธํŠธ**\n\n
\n\n
FILE INFORMATION\n\n**[`train|test.csv`]**\n\n\n\n
\n\n**[`sample_submission.csv`]**\n\n\n\n
\n\n**[`train|test_images`]**\n \nThe images\n* **550๊ฐœ์˜ ์ด๋ฏธ์ง€**. \n* **์ด๋ฏธ์ง€๋‹น ์ ์–ด๋„ ํ•œ๊ฐœ์˜ FTU ์กด์žฌ**.\n* ๊ฑด๊ฐ•ํ•œ ๊ธฐ์ฆ์ž๊ฐ€ ์ œ๊ณตํ•œ ์‚ฌ์ง„์„ ๋ณ‘๋ฆฌํ•™์ž๋“ค์˜ FTU ๋ถ„์„ํ•ด์„œ ํ‘œ์‹œํ•œ๊ฒƒ\n* HPA details:\n * HPA ๋ชจ๋“  ์ด๋ฏธ์ง€์˜ ํฌ๊ธฐ๋Š” **3000 x 3000 pixels** ๊ทธ ๋‚ด๋ถ€์—์„œ ์„ธํฌ๊ฐ€ ํ‘œ์‹œ๋œ ์˜์—ญ์€ ๋Œ€๋žต 2500 x 2500 pixels. \n \n* HuBMAP details:\n * Hubmap ์ด๋ฏธ์ง€๋Š” ํฌ๊ธฐ๊ฐ€ ๋‹ค์–‘ํ•œ๋ฐ ํฐ๊ฑฐ๋Š” **4500x4500** ์ž‘์€๊ฑฐ๋Š” **160x160 pixels** ๊นŒ์ง€ ๋‹ค์–‘ํ•˜๊ฒŒ ์กด์žฌ. \n\n
\n\n**[`train_annotations`]**\n\nThe annotations \n* **the FTUs์˜ ์œ„์น˜(ํ…Œ๋‘๋ฆฌ)๋ฅผ ๋‹ค๊ฐํ˜•์œผ๋กœ ํ‘œ์‹œํ•œ ๋งˆ์Šคํฌ**\n\n
\n","metadata":{}},{"cell_type":"markdown","source":"
\n\n\n\n

2  ์„ธํŒ…    ⤒

\n\n---\n","metadata":{}},{"cell_type":"markdown","source":"
\n\n

2.1 ๊ฐ€์†๊ธฐ(GPU,TPU) ์ค€๋น„

\n\n---\n\n**TPU**๋ฅผ ์‚ฌ์šฉํ•˜๋ ค๋ฉด, **TPUClusterResolver**๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ดˆ๊ธฐํ™”๋ฅผ ํ•ด์•ผํ•ฉ๋‹ˆ๋‹ค. ์ดˆ๊ธฐํ™”๋Š” ์›๊ฒฉ ํด๋Ÿฌ์Šคํ„ฐ์— ์—ฐ๊ฒฐํ•˜์—ฌ ํด๋ผ์šฐ๋“œ TPU๋ฅผ ์ดˆ๊ธฐํ™”ํ•˜๋Š” ๊ฒƒ์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค.\n1. ์บ๊ธ€์—์„œ TPU๋ฅผ ์‚ฌ์šฉํ•  ๋•Œ๋Š” TPUClusterResolver์— ๋”ฐ๋กœ ์ธ์ˆ˜(๋งค๊ฐœ๋ณ€์ˆ˜)๋ฅผ ์ง€์ •์•ˆํ•ด๋„ ๋ฉ๋‹ˆ๋‹ค.\n2. ๊ทธ๋Ÿฌ๋‚˜ GCE์™€ ์ฝ”๋žฉ์—์„œ๋Š” ์ง€์ •ํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. \n\n\n
\n\n```python\n# The name you gave to the TPU to use\nTPU_WORKER = 'my-tpu-name'\n\n# or you can also specify the grpc path directly\n# TPU_WORKER = 'grpc://xxx.xxx.xxx.xxx:8470'\n\n# The zone you chose when you created the TPU to use on GCP.\nZONE = 'us-east1-b'\n\n# The name of the GCP project where you created the TPU to use on GCP.\nPROJECT = 'my-tpu-project'\n\ntpu = tf.distribute.cluster_resolver.TPUClusterResolver(tpu=TPU_WORKER, zone=ZONE, project=PROJECT)\n```\n\n
\n ๐Ÿ›‘   WARNING:

- Tensorflow ๋ฌธ์„œ์—๋Š” ๋งค๊ฐœ๋ณ€์ˆ˜(์ธ์ˆ˜)๊ฐ€ ํ”„๋กœ์ ํŠธ ์ด๋ฆ„์ด๋ผ๊ณ  ๋‚˜์™€์žˆ์ง€๋งŒ, ์‹ค์ œ๋กœ ๋ชจ๋ธ์— ์ œ๊ณตํ•ด์•ผ ํ•˜๋Š” ๋งค๊ฐœ๋ณ€์ˆ˜๋Š” GCP ๋Œ€์‹œ๋ณด๋“œ์— ๋‚˜์™€์žˆ๋Š” ํ”„๋กœ์ ํŠธ ID ์ž…๋‹ˆ๋‹ค.
\n
\n\n
\n ๐Ÿ“–   REFERENCES:

\n - Guide - Use TPUs
\n - ๋ฌธ์„œ - TPUClusterResolver
\n\n
","metadata":{}},{"cell_type":"code","source":"print(f\"\\n... ACCELERATOR SETUP STARTING ...\\n\")\n\n# Detect hardware, return appropriate distribution strategy\ntry:\n # TPU detection. No parameters necessary if TPU_NAME environment variable is set. On Kaggle this is always the case.\n TPU = tf.distribute.cluster_resolver.TPUClusterResolver() \nexcept ValueError:\n TPU = None\n\nif TPU:\n print(f\"\\n... RUNNING ON TPU - {TPU.master()}...\")\n tf.config.experimental_connect_to_cluster(TPU)\n tf.tpu.experimental.initialize_tpu_system(TPU)\n strategy = tf.distribute.experimental.TPUStrategy(TPU)\nelse:\n print(f\"\\n... RUNNING ON CPU/GPU ...\")\n # Yield the default distribution strategy in Tensorflow\n # --> Works on CPU and single GPU.\n strategy = tf.distribute.get_strategy() \n\n# What Is a Replica?\n# --> A single Cloud TPU device consists of FOUR chips, each of which has TWO TPU cores. \n# --> Therefore, for efficient utilization of Cloud TPU, a program should make use of each of the EIGHT (4x2) cores. \n# --> Each replica is essentially a copy of the training graph that is run on each core and \n# trains a mini-batch containing 1/8th of the overall batch size\nN_REPLICAS = strategy.num_replicas_in_sync\n \nprint(f\"... # OF REPLICAS: {N_REPLICAS} ...\\n\")\n\nprint(f\"\\n... ACCELERATOR SETUP COMPLTED ...\\n\")","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:25.529108Z","iopub.execute_input":"2022-06-28T16:16:25.529432Z","iopub.status.idle":"2022-06-28T16:16:25.545024Z","shell.execute_reply.started":"2022-06-28T16:16:25.529402Z","shell.execute_reply":"2022-06-28T16:16:25.543663Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

2.2 ๋Œ€ํšŒ ๋ฐ์ดํ„ฐ ๊ฐ€์ ธ์˜ค๊ธฐ

\n\n---\n\nTPU ๋ฐ์ดํ„ฐ๋Š” ๋ฐ˜๋“œ์‹œ **GCS(๊ตฌ๊ธ€ ํด๋ผ์šฐ๋“œ ์ €์žฅ์†Œ)**์—์„œ ์ง์ ‘ ๋ถˆ๋Ÿฌ์™€์•ผ ํ•˜์ง€๋งŒ, ์บ๊ธ€์€ **์บ๊ธ€๋ฐ์ดํ„ฐ์…‹**์ด๋ผ๋Š” ๋ผ์ด๋ธŒ๋Ÿฌ๋ฆฌ๋ฅผ ํ†ตํ•ด์„œ GCS์•ˆ์— ๋ฐ์ดํ„ฐ์— ์ ‘๊ทผํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

\n\n
\n ๐Ÿ“Œ   TIPS:

- If you have multiple datasets attached to the notebook, you should pass the name of a specific dataset to the `get_gcs_path()` function. In our case, the name of the dataset is the name of the directory the dataset is mounted within.

\n
","metadata":{}},{"cell_type":"code","source":"print(\"\\n... DATA ACCESS SETUP STARTED ...\\n\")\n\nif TPU:\n # Google Cloud Dataset path to training and validation images\n DATA_DIR = KaggleDatasets().get_gcs_path('hubmap-organ-segmentation')\n save_locally = tf.saved_model.SaveOptions(experimental_io_device='/job:localhost')\n load_locally = tf.saved_model.LoadOptions(experimental_io_device='/job:localhost')\nelse:\n # Local path to training and validation images\n DATA_DIR = \"/kaggle/input/hubmap-organ-segmentation\"\n save_locally = None\n load_locally = None\n\nprint(f\"\\n... DATA DIRECTORY PATH IS:\\n\\t--> {DATA_DIR}\")\n\nprint(f\"\\n... IMMEDIATE CONTENTS OF DATA DIRECTORY IS:\")\nfor file in tf.io.gfile.glob(os.path.join(DATA_DIR, \"*\")): print(f\"\\t--> {file}\")\n\nprint(\"\\n\\n... DATA ACCESS SETUP COMPLETED ...\\n\")","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:25.548063Z","iopub.execute_input":"2022-06-28T16:16:25.548454Z","iopub.status.idle":"2022-06-28T16:16:25.566054Z","shell.execute_reply.started":"2022-06-28T16:16:25.548422Z","shell.execute_reply":"2022-06-28T16:16:25.565026Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(f\"\\n... XLA OPTIMIZATIONS STARTING ...\\n\")\n\nprint(f\"\\n... CONFIGURE JIT (JUST IN TIME) COMPILATION ...\\n\")\n# enable XLA optmizations (10% speedup when using @tf.function calls)\ntf.config.optimizer.set_jit(True)\n\nprint(f\"\\n... XLA OPTIMIZATIONS COMPLETED ...\\n\")","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:25.568263Z","iopub.execute_input":"2022-06-28T16:16:25.568554Z","iopub.status.idle":"2022-06-28T16:16:25.574757Z","shell.execute_reply.started":"2022-06-28T16:16:25.568527Z","shell.execute_reply":"2022-06-28T16:16:25.57378Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

2.3 LEVERAGING XLA OPTIMIZATIONS

\n\n---\n\n\n**XLA**์€ ํ…์„œํ”Œ๋กœ์šฐ ํ™˜๊ฒฝ์•ˆ์—์„œ ๋„๋ฉ”์ธ ๋ณ„๋กœ ํŠนํ™”๋œ ์„ ํ˜•๋Œ€์ˆ˜ ๊ณ„์‚ฐ ์ปดํŒŒ์ผ๋Ÿฌ์ด๋ฉฐ, **์—ฐ์‚ฐ ์†๋„์™€ ๋ฉ”๋ชจ๋ฆฌ ์‚ฌ์šฉ๋Ÿ‰์„ ํ–ฅ์ƒ์‹œํ‚จ๋‹ค**.\n\n
\n\nTensorflow ์ฝ”๋“œ๊ฐ€ ์‹คํ–‰๋˜๋ฉด ๋ชจ๋“  ์ž‘์—…์€ ๊ฐœ๋ณ„์ ์œผ๋กœ ์‹คํ–‰๋ฉ๋‹ˆ๋‹ค. ๊ฐ ์ž‘์—…์—๋Š” ์‚ฌ์ „์— ์ปดํŒŒ์ผ ๋œ GPU/TPU ์ปค๋„ ๊ตฌํ˜„์ด ์ง„ํ–‰๋ฉ๋‹ˆ๋‹ค.

\nXLA๋Š” ์ด๋Ÿฌํ•œ ๋ชจ๋ธ์„ ์‹คํ–‰ํ•˜๋Š”๋ฐ ๋Œ€์ฒดํ•  ์ˆ˜ ์žˆ๋Š” ๋ชจ๋“œ๋ฅผ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค. ์ฆ‰, TensorFlow ๊ทธ๋ž˜ํ”„๋ฅผ ์ฃผ์–ด์ง„ ๋ชจ๋ธ์— ๋Œ€ํ•ด ํŠน๋ณ„ํžˆ ์ƒ์„ฑ๋œ ๊ณ„์‚ฐ ์ปค๋„ ์‹œํ€€์Šค๋กœ ์ปดํŒŒ์ผํ•ฉ๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ์ปค๋„์€ ๋ชจ๋ธ์— ๊ณ ์œ ํ•˜๊ธฐ ๋•Œ๋ฌธ์— ์ตœ์ ํ™”๋ฅผ ์œ„ํ•ด ๋ชจ๋ธ๋ณ„ ์ •๋ณด๋ฅผ ํ™œ์šฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

\n\n
\n ๐Ÿ›‘   WARNING:

- XLA can not currently compile functions where dimensions are not inferrable: that is, if it's not possible to infer the dimensions of all tensors without running the entire computation
\n
\n\n
\n ๐Ÿ“Œ   NOTE:

- XLA compilation is only applied to code that is compiled into a graph (in TF2 that's only a code inside tf.function).
- The jit_compile API has must-compile semantics, i.e. either the entire function is compiled with XLA, or an errors.InvalidArgumentError exception is thrown)\n
\n\n
\n ๐Ÿ“–   REFERENCE:

- XLA: Optimizing Compiler for Machine Learning
\n
","metadata":{}},{"cell_type":"markdown","source":"
\n\n

2.4 ๊ธฐ๋ณธ์ ์ธ ๋ฐ์ดํ„ฐ ํ™•์ธ ๋ฐ ์ดˆ๊ธฐํ™”

\n\n---\n","metadata":{}},{"cell_type":"code","source":"print(\"\\n... BASIC DATA SETUP STARTING ...\\n\\n\")\n\n# Open the training dataframe and display the initial dataframe\nTRAIN_IMAGES_DIR = os.path.join(DATA_DIR, \"train_images\")\nTRAIN_LABELS_DIR = os.path.join(DATA_DIR, \"train_annotations\")\nTRAIN_CSV = os.path.join(DATA_DIR, \"train.csv\")\ntrain_df = pd.read_csv(TRAIN_CSV)\n\n# Get all training images\nall_train_images = glob(os.path.join(TRAIN_IMAGES_DIR, \"*.tiff\"), recursive=True)\nall_train_labels = glob(os.path.join(TRAIN_LABELS_DIR, \"*.json\"), recursive=True)\n\nprint(\"\\n... ORIGINAL TRAINING DATAFRAME... \\n\")\ndisplay(train_df)\n\nTEST_IMAGES_DIR = os.path.join(DATA_DIR, \"test_images\")\nTEST_CSV = os.path.join(DATA_DIR, \"test.csv\")\ntest_df = pd.read_csv(TEST_CSV)\n\nSS_CSV = os.path.join(DATA_DIR, \"sample_submission.csv\")\nss_df = pd.read_csv(SS_CSV)\n\n# Get all testing images if there are any\nall_test_images = glob(os.path.join(TEST_IMAGES_DIR, \"*.tiff\"), recursive=True)\n\nprint(\"\\n\\n\\n... ORIGINAL SUBMISSION DATAFRAME... \\n\")\ndisplay(ss_df)\n\n# For debugging purposes when the test set hasn't been substituted we will know\nDEBUG=len(ss_df)==0\n\n# if DEBUG:\n# TEST_DIR = TRAIN_DIR\n# all_test_images = all_train_images\n# ss_df = train_df.iloc[:10]\n# ss_df = ss_df[[\"id\", \"class\"]]\n# ss_df[\"predicted\"] = \"\"\n \n# print(\"\\n\\n\\n... DEBUG SUBMISSION DATAFRAME... \\n\")\n# display(ss_df)\n\ndef rgb2hex(rgb_tuple):\n r,g,b = rgb_tuple\n def clamp(x): \n return max(0, min(x, 255))\n return \"#{0:02x}{1:02x}{2:02x}\".format(clamp(r), clamp(g), clamp(b))\nORGANS = ['kidney', 'largeintestine', 'lung', 'prostate', 'spleen']\n_COLOURS = [(230, 0, 73), (11, 180, 255), (80, 233, 145), (230, 216, 0), (155, 25, 245)]\nO2C_MAP = {_o:_c for _o,_c in zip(ORGANS, _COLOURS)}\nO2C_HEX_MAP = {_o:rgb2hex(_c) for _o,_c in O2C_MAP.items()}\n\nprint(f\"\\n\\n\\n... ARE WE DEBUGGING: {DEBUG}... \\n\")\n\nprint(\"\\n... BASIC DATA SETUP FINISHED ...\\n\\n\")","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:25.57599Z","iopub.execute_input":"2022-06-28T16:16:25.577056Z","iopub.status.idle":"2022-06-28T16:16:26.168351Z","shell.execute_reply.started":"2022-06-28T16:16:25.577003Z","shell.execute_reply":"2022-06-28T16:16:26.167308Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

2.5 ๊ธฐ๋ณธ ๋ฐ์ดํ„ฐ์— ์ถ”๊ฐ€ ๋ฐ์ดํ„ฐ ์ž…๋ ฅํ•˜๊ธฐ

\n\n---\n\n๊ฐ๊ฐ์˜ ๋‹จ๊ณ„๋ณ„๋กœ ๊ฒ€์ฆํ•  ์ˆ˜ ์žˆ๋„๋ก ๋‚˜๋ˆ„์–ด์„œ ์ฝ”๋“œ๋ฅผ ์ ์—ˆ์ง€๋งŒ ์ˆœ์ฐจ์ ์œผ๋กœ ๊ตฌ์„ฑ๋˜์–ด ์žˆ์œผ๋ฏ€๋กœ ์ญ‰ ๋Œ๋ฆฌ์…”๋„ ๋ฌด๋ฐฉํ•ฉ๋‹ˆ๋‹ค.\n\n**NOTE:** \n","metadata":{}},{"cell_type":"code","source":"sex_2_int = {\"Male\":0, \"Female\":1}\nint_2_sex = {v:k for k,v in sex_2_int.items()}\nage_divisor = 100.0\n\ntrain_df[\"sex\"] = train_df[\"sex\"].map(sex_2_int)\ntrain_df[\"age\"] = train_df[\"age\"]/age_divisor\ntrain_df","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:26.170009Z","iopub.execute_input":"2022-06-28T16:16:26.170443Z","iopub.status.idle":"2022-06-28T16:16:26.207214Z","shell.execute_reply.started":"2022-06-28T16:16:26.170399Z","shell.execute_reply":"2022-06-28T16:16:26.206161Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_img_map = {int(x[:-5].rsplit(\"/\", 1)[-1]):x for x in all_train_images}\ntrain_lbl_map = {int(x[:-5].rsplit(\"/\", 1)[-1]):x for x in all_train_labels}\ntest_img_map = {int(x[:-5].rsplit(\"/\", 1)[-1]):x for x in all_test_images}\n\ntrain_df.insert(3, \"img_path\", train_df[\"id\"].map(train_img_map))\ntrain_df.insert(4, \"lbl_path\", train_df[\"id\"].map(train_lbl_map))\ntest_df.insert(3, \"img_path\", test_df[\"id\"].map(test_img_map))\n\ntrain_df","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:26.208392Z","iopub.execute_input":"2022-06-28T16:16:26.208709Z","iopub.status.idle":"2022-06-28T16:16:26.24996Z","shell.execute_reply.started":"2022-06-28T16:16:26.208682Z","shell.execute_reply":"2022-06-28T16:16:26.248963Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n\n\n\n\n

\n 3  ๋ณด์ถฉ ํ•จ์ˆ˜ & ํด๋ž˜์Šค๋“ค     ⤒\n

\n\n---","metadata":{}},{"cell_type":"code","source":"# ref: https://www.kaggle.com/paulorzp/run-length-encode-and-decode\n# modified from: https://www.kaggle.com/inversion/run-length-decoding-quick-start\ndef rle_decode(mask_rle, shape, color=1):\n \"\"\" TBD\n \n Args:\n mask_rle (str): run-length as string formated (start length)\n shape (tuple of ints): (height,width) of array to return \n \n Returns: \n Mask (np.array)\n - 1 indicating mask\n - 0 indicating background\n\n \"\"\"\n # Split the string by space, then convert it into a integer array\n s = np.array(mask_rle.split(), dtype=int)\n\n # Every even value is the start, every odd value is the \"run\" length\n starts = s[0::2] - 1\n lengths = s[1::2]\n ends = starts + lengths\n\n # The image image is actually flattened since RLE is a 1D \"run\"\n if len(shape)==3:\n h, w, d = shape\n img = np.zeros((h * w, d), dtype=np.float32)\n else:\n h, w = shape\n img = np.zeros((h * w,), dtype=np.float32)\n\n # The color here is actually just any integer you want!\n for lo, hi in zip(starts, ends):\n img[lo : hi] = color\n \n # Don't forget to change the image back to the original shape\n return img.reshape(shape).T\n\n# https://www.kaggle.com/namgalielei/which-reshape-is-used-in-rle\ndef rle_decode_top_to_bot_first(mask_rle, shape):\n \"\"\" TBD\n \n Args:\n mask_rle (str): run-length as string formated (start length)\n shape (tuple of ints): (height,width) of array to return \n \n Returns:\n Mask (np.array)\n - 1 indicating mask\n - 0 indicating background\n\n \"\"\"\n s = mask_rle.split()\n starts, lengths = [np.asarray(x, dtype=int) for x in (s[0:][::2], s[1:][::2])]\n starts -= 1\n ends = starts + lengths\n img = np.zeros(shape[0]*shape[1], dtype=np.uint8)\n for lo, hi in zip(starts, ends):\n img[lo:hi] = 1\n return img.reshape((shape[1], shape[0]), order='F').T # Reshape from top -> bottom first\n\n# ref.: https://www.kaggle.com/stainsby/fast-tested-rle\ndef rle_encode(img):\n \"\"\" TBD\n \n Args:\n img (np.array): \n - 1 indicating mask\n - 0 indicating background\n \n Returns: \n run length as string formated\n \"\"\"\n pixels = img.T.flatten()\n pixels = np.concatenate([[0], pixels, [0]])\n runs = np.where(pixels[1:] != pixels[:-1])[0] + 1\n runs[1::2] -= runs[::2]\n return ' '.join(str(x) for x in runs)\n\ndef flatten_l_o_l(nested_list):\n \"\"\" Flatten a list of lists \"\"\"\n return [item for sublist in nested_list for item in sublist]\n\ndef load_json_to_dict(json_path):\n \"\"\" tbd \"\"\"\n with open(json_path) as json_file:\n data = json.load(json_file)\n return data\n\ndef tf_decode_tiff(img_path, to_numpy=False, to_rgb=False):\n img = tf.io.read_file(img_path)\n img = tfio.experimental.image.decode_tiff(img)\n \n # Optionals\n if to_rgb: img = tfio.experimental.color.rgba_to_rgb(img)\n if to_numpy: img = img.numpy()\n \n return img","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:26.251429Z","iopub.execute_input":"2022-06-28T16:16:26.251744Z","iopub.status.idle":"2022-06-28T16:16:26.271014Z","shell.execute_reply.started":"2022-06-28T16:16:26.251718Z","shell.execute_reply":"2022-06-28T16:16:26.26961Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n\n\n\n\n

\n 4  ๋ฐ์ดํ„ฐ์…‹ ํƒํ—˜!    ⤒\n

\n\n---","metadata":{}},{"cell_type":"markdown","source":"
\n\n

4.0 ํ•˜๋‚˜์˜ ์˜ˆ์‹œ๋ฅผ ์‚ดํŽด๋ณด์ž!

\n\n---\n\n๋ชจ๋“  ๋ฐ์ดํ„ฐ๋“ค์ด ๊ฐ ์œ„์น˜์— ์ž˜ ์žˆ๋Š”์ง€, ๋ฐ์ดํ„ฐ์˜ ๊ธฐ๋ณธ ์‚ฌํ•ญ์„ ํ™•์ธํ•˜๋Š” ๊ณผ์ •์ž…๋‹ˆ๋‹ค.\n","metadata":{}},{"cell_type":"code","source":"def get_overlay(img_path, organ, rle_str, img_shape, _alpha=0.999, _beta=0.35, _gamma=0):\n _img = tf_decode_tiff(img_path, to_numpy=True, to_rgb=True).astype(np.float32)\n _seg_rgb = (np.stack([rle_decode(rle_str, shape=img_shape, color=1),]*3, axis=-1)*O2C_MAP[organ]).astype(np.float32)\n seg_overlay = cv2.addWeighted(src1=_img, alpha=_alpha, \n src2=_seg_rgb, beta=_beta, gamma=_gamma)\n return seg_overlay/255.\n\ndef examine_id(df, ex_id=None, plot_overlay=True, print_meta=True, plot_original=False, plot_segmentation=False, _figsize=(20,20)):\n \"\"\" Wrapper function to allow for easy visual exploration of an example \"\"\"\n if ex_id is None:\n ex_id = df[\"id\"].sample(1).values[0]\n print(f\"\\n... NO ID GIVEN... RANDOM ID CHOSEN: ID={ex_id} ...\")\n \n print(f\"\\n... ID ({ex_id}) EXPLORATION STARTED ...\\n\\n\")\n demo_ex = df[df[\"id\"]==ex_id].squeeze()\n\n if print_meta:\n print(f\"\\n... WITH id=`{ex_id}` (organ=`{demo_ex['organ']}`) โ€“ WE HAVE THE FOLLOWING DEMO EXAMPLE TO WORK FROM ... \\n\\n\")\n display(demo_ex.to_frame())\n\n if plot_original:\n print(f\"\\n\\n... ORIGINAL IMAGE PLOT ...\\n\")\n plt.figure(figsize=_figsize)\n plt.imshow(tf_decode_tiff(demo_ex[\"img_path\"], to_numpy=True, to_rgb=True))\n plt.title(f\"Original RGB Image For ID: {ex_id}\", fontweight=\"bold\")\n plt.axis(False)\n plt.show()\n\n if plot_segmentation:\n print(f\"\\n\\n... SEGMENTATION MASK PLOT ({demo_ex['organ']})...\\n\")\n plt.figure(figsize=_figsize)\n plt.imshow((np.stack([rle_decode(demo_ex.rle, shape=(demo_ex.img_width, demo_ex.img_height), color=1),]*3, axis=-1)*O2C_MAP[demo_ex.organ]).astype(np.float32))\n plt.title(f\"Segmentation Mask ({demo_ex.organ})\", fontweight=\"bold\")\n plt.axis(False)\n plt.show()\n\n if plot_overlay:\n print(f\"\\n\\n... IMAGE WITH RGB SEGMENTATION MASK OVERLAY ({demo_ex['organ']}) ...\\n\")\n seg_overlay = get_overlay(demo_ex.img_path, demo_ex.organ, demo_ex.rle, img_shape=(demo_ex.img_width, demo_ex.img_height))\n\n plt.figure(figsize=_figsize)\n plt.imshow(seg_overlay)\n plt.title(f\"Segmentation Overlay id=`{ex_id}` (organ=`{demo_ex['organ']}`)\", fontweight=\"bold\")\n handles = [Rectangle((0,0),1,1, color=(*[__c/255. for __c in _c], 0.5)) for _c in _COLOURS]\n labels = ORGANS\n plt.legend(handles,labels)\n plt.axis(False)\n plt.show()\n\n print(\"\\n\\n... SINGLE ID EXPLORATION FINISHED ...\\n\\n\")","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:26.273395Z","iopub.execute_input":"2022-06-28T16:16:26.274007Z","iopub.status.idle":"2022-06-28T16:16:26.302276Z","shell.execute_reply.started":"2022-06-28T16:16:26.273953Z","shell.execute_reply":"2022-06-28T16:16:26.300598Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"for _o in ORGANS:\n print(f\"\\n\\n... RANDOM {_o.upper()} TRAIN EXAMPLE ...\\n\\n\")\n examine_id(train_df[train_df.organ==_o])\n\nprint(\"\\n\\n\\n\\n... TEST IMAGE ...\\n\\n\")\nexamine_id(test_df, plot_overlay=False, plot_original=True)","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:26.305457Z","iopub.execute_input":"2022-06-28T16:16:26.306112Z","iopub.status.idle":"2022-06-28T16:16:49.216199Z","shell.execute_reply.started":"2022-06-28T16:16:26.306071Z","shell.execute_reply":"2022-06-28T16:16:49.215053Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

4.1 ๋‹ค์–‘ํ•œ ๊ธฐ๊ด€ ๋ฐ์ดํ„ฐ๋“ค์˜ ๋นˆ๋„์ˆ˜(๋ฐ์ดํ„ฐ์ˆ˜) ํ™•์ธ

\n\n---\n\n
OBSERVATIONS\n\nWe can observe the following distribution:\n* **์‹ ์žฅ**\n * **99** images\n* **์ „๋ฆฝ์„ **\n * **93** images\n* **๋Œ€์žฅ**\n * **58** images\n* **๋น„์žฅ**\n * **53** images\n* **ํ**\n * **48** images\n \n๋‹ค์–‘ํ•œ ์ข…๋ฅ˜์˜ ๊ธฐ๊ด€ ๋ฐ์ดํ„ฐ๋“ค์ด ์ „์ฒด์ ์œผ๋กœ ์ž˜ ๋“ค์–ด์žˆ์ง€๋งŒ, **์‹ ์žฅ**๊ณผ **์ „๋ฆฝ์„ **์€ **๋Œ€์žฅ**, **๋น„์žฅ**,**ํ** ์— ๋น„ํ•ด 2๋ฐฐ ๋งŽ์€ ๋ฐ์ดํ„ฐ๋ฅผ ๋ณด์—ฌ์ฃผ๊ณ  ์žˆ๋‹ค.","metadata":{}},{"cell_type":"code","source":"fig = px.histogram(train_df, \"organ\", color_discrete_map=O2C_HEX_MAP, color=\"organ\", title=\"Number of Segmentation Masks Per Organ Type\",)\nfig.show()","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:49.217664Z","iopub.execute_input":"2022-06-28T16:16:49.21805Z","iopub.status.idle":"2022-06-28T16:16:50.161106Z","shell.execute_reply.started":"2022-06-28T16:16:49.218006Z","shell.execute_reply":"2022-06-28T16:16:50.160012Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

4.2 ์ด๋ฏธ์ง€ ์‚ฌ์ด์ฆˆ ๊ฒ€์‚ฌ

\n\n---\n\n๋ชจ๋“  ์ด๋ฏธ์ง€ ํฌ๊ธฐ๊ฐ€ ๊ฐ™์ง€๋Š” ์•Š์ง€๋งŒ, ๊ทธ๋ ‡๋‹ค๊ณ  ์—„์ฒญ๋‚˜๊ฒŒ ์ฐจ์ด๊ฐ€ ๋‚˜๋Š”ํŽธ์€ ์•„๋‹™๋‹ˆ๋‹ค.:\n\n
๊ด€์ฐฐ ๊ฒฐ\n* ์ „์ฒด์ ์œผ๋กœ ๋Œ€๋ถ€๋ถ„ ์ด๋ฏธ์ง€์˜ ์‚ฌ์ด์ฆˆ๋Š” $(3000 \\times 3000)$ ์ž…๋‹ˆ๋‹ค.\n* ๊ทธ๊ฒŒ ์•„๋‹Œ ์ด๋ฏธ์ง€ ์‚ฌ์ด์ฆˆ๋“ค์€ ๋Œ€๋ถ€๋ถ„ 1,2๋ฒˆ์˜ ๋นˆ๋„๋กœ ๋‚˜ํƒ€๋‚ฉ๋‹ˆ๋‹ค.\n* ๋ชจ๋“  ์ด๋ฏธ์ง€๋Š” ์ •์‚ฌ๊ฐํ˜•์ž…๋‹ˆ๋‹ค.\n\n
๊ธฐํƒ€ ์ •๋ณด\n\n* ๋Œ€๋žต 550๊ฐœ์˜ ์ด๋ฏธ์ง€๊ฐ€ hidden ํ…Œ์ŠคํŠธ ๋ฐ์ดํ„ฐ์…‹์— ์žˆ์Šต๋‹ˆ๋‹ค. \n* ๋ชจ๋“  HPA ์ด๋ฏธ์ง€ ๋ฐ์ดํ„ฐ๋Š” 3000__ ์ด๋ฉฐ ๊ฐ€์šด๋ฐ 2500*2500 ์˜์—ญ์— ์กฐ์ง์˜์—ญ์ด ์žˆ์Šต๋‹ˆ๋‹ค. \n* Hubmap ์ด๋ฏธ์ง€๋Š” 4500ํ”ฝ์…€๋ถ€ํ„ฐ 160ํ”ฝ์…€๊นŒ์ง€ ๋‹ค์–‘ํ•œ ํฌ๊ธฐ๋กœ ๊ตฌ์„ฑ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.\n\n
","metadata":{}},{"cell_type":"code","source":"fig = px.scatter(train_df.drop_duplicates(subset=[\"img_width\", \"img_height\"]), x=\"img_width\", y=\"img_height\", \n size=train_df.groupby([\"img_width\", \"img_height\"])[\"id\"].transform(\"count\").iloc[train_df.drop_duplicates(subset=[\"img_width\", \"img_height\"]).index], \n color=\"(\"+train_df.drop_duplicates(subset=[\"img_width\", \"img_height\"])[\"img_width\"].astype(str)+\",\"+train_df.drop_duplicates(subset=[\"img_width\", \"img_height\"])[\"img_height\"].astype(str)+\")\", \n title=\"Bubble Chart Showing The Various Image Sizes\",\n labels={\"color\":\"Size Legend\", \n \"size\":\"Number Of Observations\",\n \"img_height\":\"Image Height (pixels)\",\n \"img_width\":\"Image Width (pixels)\"},\n size_max=150)\nfig.show()","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:50.162477Z","iopub.execute_input":"2022-06-28T16:16:50.162799Z","iopub.status.idle":"2022-06-28T16:16:50.34535Z","shell.execute_reply.started":"2022-06-28T16:16:50.16277Z","shell.execute_reply":"2022-06-28T16:16:50.344209Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

4.3 ๋‚˜์ด ๊ฒ€์‚ฌ

\n\n---\n\n
๊ด€์ฐฐ ๊ฒฐ๊ณผ\n* ์—ฐ๋ น๋ถ„ํฌ๊ฐ€ 50+๋กœ ์น˜์šฐ์ณ์ ธ ์žˆ์Šต๋‹ˆ๋‹ค.\n* ๋Œ€์žฅ์—์„œ์˜ ์—ฐ๋ น๋ถ„ํฌ๋Š” ๋‹ค๋ฅธ ์žฅ๊ธฐ๋ณด๋‹ค ๋” ๋‚˜์ด๊ฐ€ ๋งŽ์€ ์ชฝ์œผ๋กœ ๊ธฐ์šธ์–ด์ ธ ์žˆ์Šต๋‹ˆ๋‹ค.\n* ์—ฐ๋ น ๊ทธ๋ฃน ์ „๋ฐ˜์— ๊ฑธ์นœ ์žฅ๊ธฐ ์„ธ๋ถ„ํ™” ์œ ํ˜•์˜ ์ผ๋ฐ˜์ ์ธ ๋ถ„ํฌ๋Š” ์ƒ๋Œ€์ ์œผ๋กœ ๊ณ„์ธตํ™”๋œ ๊ฒƒ์œผ๋กœ ๋‚˜ํƒ€๋‚ฉ๋‹ˆ๋‹ค(์ด์ „์— ์–ธ๊ธ‰ํ•œ ์˜ˆ์™ธ ์ œ์™ธ).\n\n
","metadata":{}},{"cell_type":"code","source":"fig = px.histogram(train_df, train_df[\"age\"]*age_divisor, color_discrete_map=O2C_HEX_MAP, color=\"organ\", \n title=\"Number of Segmentation Masks Per Age Bin by Organ Type \",\n labels={\"organ\":\"Organ Legend\", \n \"count\":\"Number Of Observations\",\n \"x\":\"Age\"}, text_auto=True,\n )\nfig.show()","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:50.346885Z","iopub.execute_input":"2022-06-28T16:16:50.347314Z","iopub.status.idle":"2022-06-28T16:16:50.434528Z","shell.execute_reply.started":"2022-06-28T16:16:50.347269Z","shell.execute_reply":"2022-06-28T16:16:50.433487Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

4.4 ์„ฑ๋ณ„ ๊ฒ€์‚ฌ

\n\n---\n\n
OBSERVATIONS\n* ์—ฌ์„ฑ์€ ์ „๋ฆฝ์„ ์ด ์—†์–ด์„œ ์ „๋ฆฝ์„ ์—๋Š” ๋‚จ์„ฑ๋งŒ ์žˆ์Šต๋‹ˆ๋‹ค.\n* ์‹ ์žฅ์€ ๋ถ„ํฌ๊ฐ€ ๋‚จ์„ฑ์ชฝ์œผ๋กœ ๋งŽ์ด ๊ธฐ์šธ์–ด ์žˆ์Šต๋‹ˆ๋‹ค.\n* ๋‹ค๋ฅธ ๊ธฐ๊ด€์€ ์ „์ฒด์ ์œผ๋กœ ๊ณ ๋ฅด๊ฒŒ ๋ถ„ํฌ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.\n\n
","metadata":{}},{"cell_type":"code","source":"fig = px.histogram(train_df, train_df[\"sex\"], color_discrete_map=O2C_HEX_MAP, color=\"organ\", \n title=\"Examples by Gender (0=Male, 1=Female) and Organ Type\",\n labels={\"organ\":\"Organ Legend\", \n \"count\":\"Number Of Observations\",\n \"sex\":\"Sex\"},text_auto=True, barmode=\"group\",\n )\nfig.show()","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:16:50.436159Z","iopub.execute_input":"2022-06-28T16:16:50.436471Z","iopub.status.idle":"2022-06-28T16:16:50.522584Z","shell.execute_reply.started":"2022-06-28T16:16:50.436441Z","shell.execute_reply":"2022-06-28T16:16:50.521499Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

4.5 ๋‹ค์–‘ํ•œ ์žฅ๊ธฐ์—์„œ์˜ FTU ํ™•์ธ

\n\n---\n\n์‹ ์žฅ์—์„œ ์žฅ๊ธฐ ์ˆ˜์ค€์—์„œ \n* **Notice that FTU is one level higher than cells.**\n\n
\n\n
\n\nFTUs in the organs presented in this competition are: \n* **glomeruli** in the **kidney** โ€“ (a)\n* **crypt** in the **large intestine** โ€“ (b)\n* **alveolus** in the **lung** โ€“ (c)\n* **glandular acinus** in the **prostate** โ€“ (d)\n* **white pulp** in the **spleen** โ€“ (e)\n\n
\n\n
\n\n
OBSERVATIONS\n* TBD","metadata":{}},{"cell_type":"code","source":"def crop_ftu_polygons(img, cnts, pad=5):\n def get_crop_from_cnt(_img, _cnt, _pad):\n x1,y1,w,h = cv2.boundingRect(np.array(_cnt))\n x2,y2 = x1+w+_pad, y1+h+_pad\n x1, y1 = max(0, x1-_pad), max(0, y1-_pad)\n return _img[y1:y2, x1:x2]\n ftu_crops = [get_crop_from_cnt(img, cnt, pad) for cnt in cnts]\n return ftu_crops \n \ndef plot_ftus(ftu_crops, n_cols=4, height_per_row=7, max_crops=50):\n ftu_crops = ftu_crops[:max_crops]\n n_rows = int(np.ceil(len(ftu_crops)/n_cols))\n \n plt.figure(figsize=(20, n_rows*height_per_row))\n for i, ftu_crop in enumerate(ftu_crops):\n plt.subplot(n_rows, n_cols, i+1)\n plt.imshow(ftu_crop)\n plt.title(f\"FTU #{i+1} โ€“โ€“ SHAPE={ftu_crop.shape[:-1]}\", fontweight=\"bold\")\n plt.axis(False)\n \n plt.tight_layout()\n plt.show()\n\nMAX_N_FTU = 50\nN_EX = 10\n\nftu_crop_map = {}\nfor _organ in ORGANS:\n sub_df = train_df[train_df.organ==_organ].sample(N_EX)\n imgs = [cv2.imread(img_path)[..., ::-1] for img_path in sub_df.img_path.values]\n cnts = [load_json_to_dict(json_path) for json_path in sub_df.lbl_path.values]\n ftu_crops = flatten_l_o_l([crop_ftu_polygons(img, _cnts, pad=0) for img, _cnts in zip(imgs, cnts)])[:MAX_N_FTU]\n ftu_crop_map[_organ] = ftu_crops.copy()\n print(f\"\\n\\n\\n... DISPLAYING {len(ftu_crops)} {_organ} FTU CROPS ...\\n\")\n plot_ftus(ftu_crops)\n\n# def get_ftus(df, organ, k=25, pad=25, one_per_img=False, data_source=\"HPA\", do_resize=False, resize_to=(128,128)):\n# pass","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:21:59.184235Z","iopub.execute_input":"2022-06-28T16:21:59.185503Z","iopub.status.idle":"2022-06-28T16:22:56.2919Z","shell.execute_reply.started":"2022-06-28T16:21:59.18543Z","shell.execute_reply":"2022-06-28T16:22:56.290581Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

4.6 AVERAGE FTU SIZE (BOUNDING-RECT) BY ORGAN

\n\n---\n\nSince it would take too much processing to comb through all the available images and respective FTUs we will look at the previously obtained sampling of crops (bounding-rects) and the respective dimensions therewithin.\n\nAs we know all the HPA images have a pixel resolution (size) of 0.4ยตm we can therefore determine the real dimensions of the bounding rectangles around each FTU.\n\nNote dimensions are represented as (HEIGHTxWIDTH) unless otherwise communicated\n\n
OBSERVATIONS\n* **Kidney**\n * **Actual** FTU size is on-average SQUARE (most images aren't square but it averages as such) at **~157x155ยตm**\n * **HPA** FTU size in pixels will be **~392x389px**\n * **HuBMAP** FTU size in pixels will be **~314x310px**\n* **Large Intestine**\n * **Actual** FTU size is on-average RECTANGULAR (1Hx2W) (images are mostly horizontal but occasionally vertical or square) at **~60x131ยตm**\n * **HPA** FTU size in pixels will be **~150x328px**\n * **HuBMAP** FTU size in pixels will be **~262x572px**\n* **Lung**\n * **Actual** FTU size is on-average SQUARE (images aren't square but average to mostly square) at **~90x80ยตm**\n * **HPA** FTU size in pixels will be **~225x200px**\n * **HuBMAP** FTU size in pixels will be **~119x106px**\n* **Prostate**\n * **Actual** FTU size is on-average SQUARE (images aren't square but average to mostly square) at **~68x77ยตm**\n * **HPA** FTU size in pixels will be **~172x193px**\n * **HuBMAP** FTU size in pixels will be **~11x13px**\n* **Spleen**\n * **Actual** FTU size is on-average RECTANGULAR (2Hx1W) at **~407x197ยตm**\n * **HPA** FTU size in pixels will be **~1018x493px**\n * **HuBMAP** FTU size in pixels will be **~823x399px**\n\n
","metadata":{}},{"cell_type":"code","source":"ftu_shape_map = {_organ:[c_map.shape[:-1] for c_map in crop_maps] for _organ, crop_maps in ftu_crop_map.items()}\nftu_area_map = {_organ:[c_map.shape[0]*c_map.shape[1] for c_map in crop_maps] for _organ, crop_maps in ftu_crop_map.items()}\n\norgan_ftu_df = pd.DataFrame({\"organ\":ORGANS})\norgan_ftu_df[\"average_area\"] = organ_ftu_df.organ.apply(lambda x: np.array(ftu_area_map[x]).mean())\norgan_ftu_df[\"average_area_ยตm\"] = organ_ftu_df[\"average_area\"]*0.4**2\norgan_ftu_df[\"average_height\"] = organ_ftu_df.organ.apply(lambda x: np.array(ftu_shape_map[x][0]).mean())\norgan_ftu_df[\"average_width\"] = organ_ftu_df.organ.apply(lambda x: np.array(ftu_shape_map[x][1]).mean())\norgan_ftu_df[\"average_shape_ยตm\"] = (\"(\"+(organ_ftu_df[\"average_height\"]*0.4).round(2).astype(str)+\",\"+(organ_ftu_df[\"average_width\"]*0.4).round(2).astype(str)+\")\").apply(ast.literal_eval)\norgan_ftu_df","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:45:38.359544Z","iopub.execute_input":"2022-06-28T16:45:38.360036Z","iopub.status.idle":"2022-06-28T16:45:38.390048Z","shell.execute_reply.started":"2022-06-28T16:45:38.359998Z","shell.execute_reply":"2022-06-28T16:45:38.389262Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n

4.7 JSON ํŒŒ์ผ๊ณผ RLE ์ฝ”๋“œ

\n\n---\n\n
OBSERVATIONS\n* The RLE is the entire binary mask as a simple compressed string\n* The JSON contains the description of the masks as a collection (list) of polygons representing the respective masks.\n * A polygon is represented as a collection (list) of vertices.\n * A vertex is a point represented by a pair of values โ€“โ€“ the x,y position\n * If you took these polygons and plotted them alongside the RLE decoded mask, they SHOULD be the same. I think it's simply another representation for us to use.\n\n
","metadata":{}},{"cell_type":"code","source":"# Use first row as example --> ex_id=10044\nex_row = train_df[train_df.id==10044]\n\nprint(\"\\n... First โ€“ Plot the image and the RLE overlapped ontop ...\\n\")\nexamine_id(ex_row)\n\nprint(\"\\n\\n\\n... Second โ€“ย Load the .json file and display mask polygons ...\\n\")\nex_poly = load_json_to_dict(ex_row.lbl_path.values[0])\nraw_img = np.array(cv2.imread(ex_row.img_path[0])[..., ::-1])\nimg = np.zeros_like(raw_img)\ncolor_space = np.linspace(235, 20, len(ex_poly), dtype=int)\nfor i, ex_p in enumerate(ex_poly):\n img = cv2.fillPoly(img, [np.expand_dims(np.array(ex_p, np.int32), axis=0)], color=(int(color_space[i]),245,245))\nplt.figure(figsize=(20,20))\nplt.title(\"Mask Polygons\", fontweight=\"bold\")\nplt.imshow(img)\nplt.show()\n\nprint(\"\\n\\n\\n Lastly โ€“ย Compare mask polygons to rle decoded mask\")\nrle_mask = rle_decode(ex_row.rle.values[0], (ex_row.img_width.values[0], ex_row.img_height.values[0]))\njson_mask = np.where(img>0, 1, 0)[..., 0].astype(np.float32)\ndiff_mask = np.abs(rle_mask-json_mask).astype(np.uint8)\n\nplt.figure(figsize=(20,20))\nplt.title(f\"Number of Non-Equal Pixels = {(rle_mask!=json_mask).sum()}\", fontweight=\"bold\")\nplt.imshow(diff_mask, cmap=\"gray\")\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:17:44.269274Z","iopub.execute_input":"2022-06-28T16:17:44.269659Z","iopub.status.idle":"2022-06-28T16:17:50.982719Z","shell.execute_reply.started":"2022-06-28T16:17:44.269628Z","shell.execute_reply":"2022-06-28T16:17:50.981803Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**VISUALIZE DAB STAIN AND HEMATOXYLIN COUNTERSTAIN**\n* DAB is brown and present at locations of enzymatic activity\n* HEMATOXYLIN is light blue when found within cells\n* HEMATOXYLIN is dark blue when found within nuclei","metadata":{"_kg_hide-output":true,"_kg_hide-input":true}},{"cell_type":"code","source":"# original_prostate_img = cv2.imread(train_df.img_path[0])[..., ::-1]\n# resized_prostate_img = cv2.resize(original_prostate_img, (int(3000*(0.4/6.263)),int(3000*(0.4/6.263))))\n\n# plt.figure(figsize=(20,40))\n\n# plt.subplot(2,1,1)\n# plt.imshow(original_prostate_img)\n# plt.title(\"Original Prostate Image (3000x3000) โ€“ 0.4\", fontweight=\"bold\")\n\n# plt.subplot(2,1,2)\n# plt.imshow(resized_prostate_img)\n# plt.title(\"Resized Prostate Image (191x191) โ€“ 6.263\", fontweight=\"bold\")\n\n# plt.savefig(\"comparison.png\", dpi=300)\n# plt.show()","metadata":{"_kg_hide-output":true,"_kg_hide-input":true,"execution":{"iopub.status.busy":"2022-06-28T16:17:50.984076Z","iopub.execute_input":"2022-06-28T16:17:50.984925Z","iopub.status.idle":"2022-06-28T16:17:50.989917Z","shell.execute_reply.started":"2022-06-28T16:17:50.984884Z","shell.execute_reply":"2022-06-28T16:17:50.988885Z"},"jupyter":{"source_hidden":true},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n\n\n\n\n

\n 5  MODELLING    ⤒\n

\n\n---","metadata":{}},{"cell_type":"markdown","source":"
\n\n

5.1 ALL ONES WITH SOME INTELLIGENCE

\n\n---\n\nWe simply submit a mask made up of ones in a circular shape in the centre of the respective image.\n\n
","metadata":{}},{"cell_type":"code","source":"rles = []\nSMART_CIRCLE_FRAC = 0.85\n\nfor i, (img_w, img_h) in enumerate(zip(test_df[\"img_width\"], test_df[\"img_height\"])):\n tmp_img = np.zeros((img_w, img_h))\n tmp_img = cv2.circle(tmp_img, (int(np.round(img_w/2)), int(np.round(img_h/2))), int(np.round((img_w/2)*SMART_CIRCLE_FRAC)), 1, -1)\n rles.append(rle_encode(tmp_img))\n if i==0:\n plt.figure(figsize=(10,10))\n plt.imshow(rle_decode(rles[0], (img_w, img_h)))\n plt.title(\"Example of Smart Circle Mask\", fontweight=\"bold\")\n plt.show()\n\nss_df[\"rle\"] = rles\nss_df = ss_df[[\"id\", \"rle\"]]\nss_df.to_csv(\"submission.csv\", index=False)\ndisplay(ss_df)","metadata":{"execution":{"iopub.status.busy":"2022-06-28T16:17:50.991598Z","iopub.execute_input":"2022-06-28T16:17:50.992566Z","iopub.status.idle":"2022-06-28T16:17:51.877938Z","shell.execute_reply.started":"2022-06-28T16:17:50.992518Z","shell.execute_reply":"2022-06-28T16:17:51.876691Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n\n\n\n\n\n

\n 9  APPENDIX    ⤒\n

\n\n---\n\nThis section is mostly if not entirely copied from various sources listed below:\n* Yashvardhan Jain's Background Info Notebook","metadata":{}},{"cell_type":"markdown","source":"

9.1 HUMAN BIOMOLECULAR ATLAS PROGRAM (HUBMAP)

\n\n---\n\nHuBMAP aims to create an open, global atlas of the healthy human body at the cellular level, see the [**HuBMAP Consortium**](https://hubmapconsortium.org/about/) and [**Visible Human Massive Open Online Course (VHMOOC)**](https://expand.iu.edu/browse/sice/cns/courses/hubmap-visible-human-mooc). \n\nOne component of this overarching goal is to identify medically relevant functional tissue units (FTUs) within whole slide microscopy images of human tissues. Once these FTUs are detected, information on size, shape, variability in number and location within the tissue samples can be used to help build a spatially accurate and semantically explicit model of the human body.","metadata":{}},{"cell_type":"markdown","source":"
\n\n

9.2 HUMAN PROTEIN ATLAS (HPA)

\n\n---\n\nThe [**Human Protein Atlas**](https://www.proteinatlas.org/) is a Swedish-based program initiated in 2003 with the aim to map all the human proteins in cells, tissues, and organs using an integration of various omics technologies, including antibody-based imaging, mass spectrometry-based proteomics, transcriptomics, and systems biology. All the data in the knowledge resource is open access to allow scientists both in academia and industry to freely access the data for exploration of the human proteome. The images for this competition are from the [**Tissue Atlas**](https://www.proteinatlas.org/humanproteome/tissue), which maps the protein expression profile across human tissues.","metadata":{}},{"cell_type":"markdown","source":"
\n\n

9.3 FUNCTIONAL TISSUE UNITS (FTUs)

\n\n---\n\nBernard de Bono and his team coined the term โ€œfunctional tissue unitโ€ to support tissue modelling (**[de Bono, 2013](https://www.ncbi.nlm.nih.gov/pubmed/24103658)**) as:\n> โ€œ... a three-dimensional block of cells centered around a capillary, such that each cell in this block is within diffusion distance from any other cell in the same blockโ€\n\nFor the purposes of Human Reference Atlas (**[Bรถrner, 2021](https://www.nature.com/articles/s41556-021-00788-6)**, **[Godwin, 2021](https://www.biorxiv.org/content/10.1101/2021.11.09.467810v1)**, **[Bรถrner, 2021](https://www.biorxiv.org/content/10.1101/2021.12.30.474265v1)**) construction and usage, **we define an FTU as the smallest level of tissue organization (i.e., a cell population neighborhood) that performs the organโ€™s major physiological function**. \n* One example of an FTU is the glomerulus found in the outer layer of kidney tissue known as the cortex, which in humans has an area of about 800 mm2 and average depth of about 9 mm (**[Mounier-Vehier, 2002](https://doi.org/10.1046/j.1523-1755.2002.00167.x)**). \n* Glomeruli consist of capillaries that facilitate filtration of waste products out of blood. \n* Normal glomeruli typically range from 100-350 ฮผm in diameter with a roughly spherical shape (**[Kannan, 2019](https://www.kireports.org/article/S2468-0249(19))**). \n\n
\n\nRefer to Fig. 1 for a zoom sequence from the human body to single-cell level for the kidney.\n\n
\n\n
\n\nSimilarly, FTUs in other organs presented in this competition are: crypt in the large intestine, alveolus in the lung, glandular acinus in the prostate, and white pulp in the spleen, see Fig. 2.\n\n
\n\n
","metadata":{}},{"cell_type":"markdown","source":"
\n\n

9.4 TERMINOLOGY

\n\n---\n\n* **Tissue Preservation**: \n * Different types of tissue preservation techniques are used to maintain the integrity of tissue for downstream biomolecular assays analysis.\n* **Fresh Frozen Tissue Preservation**: \n * Fresh frozen tissue is frozen in liquid nitrogen (-190ยฐC) within 30-60 minutes after surgical excision; this type of preservation has been the method of choice for transcriptomics and immunohistochemistry ([Robbe, 2018](https://doi.org/10.1038/gim.2017.241), [Stoeckli, 2007](https://doi.org/10.1016/j.ijms.2006.10.007)).\n* **Formalin Fixed, Paraffin Embedded (FFPE) Tissue Preservation**: \n * FFPE tissue is the preferred method for clinical pathology samples for histology assessment ([Bass, 2014](https://doi.org/10.5858/arpa.2013-0691-RA)).\n* **Periodic acid-Schiff (PAS) Stain Microscopy**: \n * [PAS](https://en.wikipedia.org/wiki/Periodic_acid%E2%80%93Schiff_stain) is a histology stain that detects complex sugars in tissue sections.\n* **Hematoxylin and Eosin (H&E) Stain Microscopy**: \n * [H&E](https://en.wikipedia.org/wiki/H%26E_stain) stain is one of the principal tissue stains used in histology.\n* **IHC staining**: \n * [Immunohistochemical](https://en.wikipedia.org/wiki/Immunohistochemistry) staining is accomplished with antibodies that recognize the target antigen.","metadata":{}},{"cell_type":"markdown","source":"
\n\n

9.5 DATASETS

\n\n---\n\nThe Ground Truth segmentations were manually created and further reviewed by Subject Matter Experts (SMEs) for quality assurance.\n\n* **HPA Data** comprises images of 1 mm diameter tissue microarray cores stained with antibodies visualized with 3,3'-diaminobenzidine (DAB) and counterstained with hematoxylin.\n* **HuBMAP Data** comprises Periodic acid-Schiff (PAS)/hematoxylin and eosin (H&E) stain images of human tissue.","metadata":{}},{"cell_type":"markdown","source":"
\n\n

9.6 RELATED WORK

\n\n---\n\n* Leah L. Godwin, Yingnan Ju, Naveksha Sood, Yashvardhan Jain, Ellen M. Quardokus, Andreas Bueckle, Teri Longacre, Aaron Horning, Yiing Lin, Edward D. Esplin, John W. Hickey, Michael P. Snyder, N. Heath Patterson, Jeffrey M. Spraggins, Katy Bรถrner. โ€œRobust and generalizable segmentation of human functional tissue unitsโ€; bioRxiv 2021.11.09.467810; doi: https://doi.org/10.1101/2021.11.09.467810\n* Sheehan, Susan M., and Ron Korstanje. 2018. โ€œAutomatic Glomerular Identification and Quantification of Histological Phenotypes Using Image Analysis and Machine Learning.โ€ American Journal of Physiology - Renal Physiology 315 (6): F1644โ€“51. https://doi.org/10.1152/ajprenal.00629.2017.\n* Gallego, Jaime, Anibal Pedraza, Samuel Lopez, Georg Steiner, Lucia Gonzalez, Arvydas Laurinavicius, and Gloria Bueno. 2018. โ€œGlomerulus Classification and Detection Based on Convolutional Neural Networks.โ€ Journal of Imaging 4 (1): 20. https://doi.org/10.3390/jimaging4010020.\n* Govind, Darshana, Brandon Ginley, Brendon Lutnick, John E. Tomaszewski, and Pinaki Sarder. 2018. โ€œGlomerular Detection and Segmentation from Multimodal Microscopy Images Using a Butterworth Band-Pass Filter.โ€ In Medical Imaging 2018: Digital Pathology, 10581:1058114. International Society for Optics and Photonics. https://doi.org/10.1117/12.2295446.\n* Bukowy, John D., Alex Dayton, Dustin Cloutier, Anna D. Manis, Alexander Staruschenko, Julian H. Lombard, Leah C. Solberg Woods, Daniel A. Beard, and Allen W. Cowley. 2018. โ€œRegion-Based Convolutional Neural Nets for Localization of Glomeruli in Trichrome-Stained Whole Kidney Sections.โ€ Journal of the American Society of Nephrology 29 (8): 2081โ€“88. https://doi.org/10.1681/ASN.2017111210.\n* Kannan, Shruti, Laura A. Morgan, Benjamin Liang, McKenzie G. Cheung, Christopher Q. Lin, Dan Mun, Ralph G. Nader, et al. 2019. โ€œSegmentation of Glomeruli Within Trichrome Images Using Deep Learning.โ€ Kidney International Reports 4 (7): 955โ€“62. https://doi.org/10.1016/j.ekir.2019.04.008.\n* Nassim Bouteldja, Barbara M. Klinkhammer, Roman D. Bรผlow, Patrick Droste, Simon W. Otten, Saskia Freifrau von Stillfried, Julia Moellmann, Susan M. Sheehan, Ron Korstanje, Sylvia Menzel, Peter Bankhead, Matthias Mietsch, Charis Drummer, Michael Lehrke, Rafael Kramann, Jรผrgen Floege, Peter Boor, Dorit Merhof. โ€œDeep Learningโ€“Based Segmentation and Quantification in Experimental Kidney Histopathologyโ€; JASN Jan 2021 32 (1) 52-68; https://doi.org/10.1681/ASN.2020050597\n* Brandon Ginley, Brendon Lutnick, Kuang-Yu Jen, Agnes B. Fogo, Sanjay Jain, Avi Rosenberg, Vighnesh Walavalkar, Gregory Wilding, John E. Tomaszewski, Rabi Yacoub, Giovanni Maria Rossi, Pinaki Sarder.โ€œComputational Segmentation and Classification of Diabetic Glomerulosclerosisโ€; JASN Oct 2019, 30 (10) 1953-1967; https://doi.org/10.1681/ASN.2018121259","metadata":{}},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}