{
"cells": [
{
"cell_type": "markdown",
"id": "casual-lunch",
"metadata": {},
"source": [
"# Trailer: Biobildanalyse mit Python\n",
"In den folgenden Kapiteln werden wir uns in die Bildanalyse, maschinelles Lernen und Biostatistik mit Python vertiefen. Dieses erste Notebook dient als Vorschau auf das, was wir machen werden.\n",
"\n",
"Python-Notebooks beginnen typischerweise mit den Importen der Python-Bibliotheken, die das Notebook verwenden wird. Der Leser kann zun\u00e4chst \u00fcberpr\u00fcfen, ob all diese Bibliotheken installiert sind, bevor er das gesamte Notebook durchgeht."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "4b5018db-c80f-49f9-bece-0409972e5a68",
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"from skimage.io import imread, imshow\n",
"import pyclesperanto_prototype as cle\n",
"from skimage import measure\n",
"import pandas as pd\n",
"import seaborn\n",
"import apoc\n",
"import stackview"
]
},
{
"cell_type": "markdown",
"id": "14b582e5-5e0e-4bf0-9e09-9495e98578c7",
"metadata": {},
"source": [
"## Arbeiten mit Bilddaten\n",
"Wir beginnen mit dem Laden der interessierenden Bilddaten. In diesem Beispiel laden wir ein Bild, das ein Zebrafischauge zeigt, mit freundlicher Genehmigung von Mauricio Rocha Martins, Norden-Labor, MPI CBG Dresden."
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "helpful-purpose",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"\n",
" \n",
" \n",
"\n",
"\n",
"\n",
"shape (400, 400) \n",
"dtype uint16 \n",
"size 312.5 kB \n",
"min 9187 max 57619 \n",
"
\n",
" \n",
" \n",
" \n",
"
"
],
"text/plain": [
"StackViewNDArray([[14709, 17195, 15101, ..., 10977, 10615, 10864],\n",
" [15234, 16786, 16048, ..., 11350, 11110, 11061],\n",
" [15709, 15896, 16645, ..., 11334, 10853, 10584],\n",
" ...,\n",
" [22707, 22025, 23783, ..., 16773, 17757, 16991],\n",
" [22484, 24699, 21860, ..., 16546, 16348, 17791],\n",
" [23357, 21872, 23148, ..., 18280, 18107, 17240]],\n",
" dtype=uint16)"
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# open an image file\n",
"multichannel_image = imread(\"../../data/zfish_eye.tif\")\n",
"\n",
"# extract a channel\n",
"single_channel_image = multichannel_image[:,:,0]\n",
"\n",
"cropped_image = single_channel_image[200:600, 500:900]\n",
"\n",
"stackview.insight(cropped_image)"
]
},
{
"cell_type": "markdown",
"id": "6a0c73b5-ec4f-4d7b-bdcc-cbbef9e7b0b2",
"metadata": {},
"source": [
"## Bildfilterung\n",
"\n",
"Ein \u00fcblicher Schritt bei der Arbeit mit Fluoreszenz-Mikroskopiebildern ist das Subtrahieren der Hintergrundintensit\u00e4t, z.B. resultierend aus unfokussiertem Licht. Dies kann die Bildsegmentierungsergebnisse weiter unten im Workflow verbessern."
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "ed5f952a-8215-4cce-a908-a69c727e1fad",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"\n",
" \n",
" \n",
"\n",
"\n",
"\n",
"shape (400, 400) \n",
"dtype float32 \n",
"size 625.0 kB \n",
"min 0.0 max 40758.0 \n",
"
\n",
" \n",
" \n",
" \n",
"
"
],
"text/plain": [
"StackViewNDArray([[1279., 3765., 1671., ..., 1790., 1428., 1677.],\n",
" [1804., 3356., 2618., ..., 2163., 1923., 1874.],\n",
" [2279., 2466., 3215., ..., 2147., 1666., 1397.],\n",
" ...,\n",
" [1608., 926., 2684., ..., 1342., 2326., 1560.],\n",
" [1385., 3600., 761., ..., 1115., 917., 2360.],\n",
" [2280., 795., 2071., ..., 2849., 2676., 1809.]],\n",
" dtype=float32)"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# subtract background using a top-hat filter\n",
"background_subtracted_image = cle.top_hat_box(cropped_image, radius_x=20, radius_y=20)\n",
"\n",
"stackview.insight(background_subtracted_image)"
]
},
{
"cell_type": "markdown",
"id": "meaning-campus",
"metadata": {},
"source": [
"## Bildsegmentierung\n",
"F\u00fcr die Segmentierung der Zellkerne im gegebenen Bild existiert eine enorme Anzahl von Algorithmen. Hier verwenden wir einen klassischen Ansatz namens Voronoi-Otsu-Labeling, der sicherlich nicht perfekt ist."
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "linear-estate",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"\n",
" \n",
" \n",
"\n",
"\n",
"\n",
"shape (400, 400) \n",
"dtype uint32 \n",
"size 625.0 kB \n",
"min 0 max 113 \n",
"
\n",
"\n",
" \n",
" \n",
"
"
],
"text/plain": [
"StackViewNDArray([[0, 0, 0, ..., 0, 0, 0],\n",
" [0, 0, 0, ..., 0, 0, 0],\n",
" [0, 0, 0, ..., 0, 0, 0],\n",
" ...,\n",
" [0, 0, 0, ..., 0, 0, 0],\n",
" [0, 0, 0, ..., 0, 0, 0],\n",
" [0, 0, 0, ..., 0, 0, 0]], dtype=uint32)"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"label_image = np.asarray(cle.voronoi_otsu_labeling(background_subtracted_image, spot_sigma=4))\n",
"\n",
"# show result\n",
"stackview.insight(label_image)"
]
},
{
"cell_type": "markdown",
"id": "incredible-explorer",
"metadata": {},
"source": [
"## Messungen und Merkmalsextraktion\n",
"Nachdem das Bild segmentiert ist, k\u00f6nnen wir Eigenschaften der einzelnen Objekte messen. Diese Eigenschaften sind typischerweise beschreibende statistische Parameter, die als Merkmale bezeichnet werden. Wenn wir Messungen wie die Fl\u00e4che oder die mittlere Intensit\u00e4t ableiten, extrahieren wir diese beiden Merkmale."
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "interstate-emperor",
"metadata": {},
"outputs": [],
"source": [
"statistics = measure.regionprops_table(label_image, \n",
" intensity_image=cropped_image,\n",
" properties=('area', 'mean_intensity', 'major_axis_length', 'minor_axis_length'))"
]
},
{
"cell_type": "markdown",
"id": "executive-centre",
"metadata": {},
"source": [
"## Arbeiten mit Tabellen\n",
"Das oben erstellte `statistics`-Objekt enth\u00e4lt eine Python-Datenstruktur, ein W\u00f6rterbuch von Messvektoren, was nicht am intuitivsten zu betrachten ist. Daher konvertieren wir es in eine Tabelle. Datenwissenschaftler nennen diese Tabellen oft DataFrames, die in der [pandas](https://pandas.pydata.org/)-Bibliothek verf\u00fcgbar sind."
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "indian-girlfriend",
"metadata": {},
"outputs": [],
"source": [
"dataframe = pd.DataFrame(statistics)"
]
},
{
"cell_type": "markdown",
"id": "37cde8e6-8368-4b08-ab92-d61a27469e3a",
"metadata": {},
"source": [
"Wir k\u00f6nnen bestehende Tabellenspalten verwenden, um andere Messungen zu berechnen, wie zum Beispiel das `aspect_ratio`."
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "df467d3a-49a0-4c7e-a329-11a08608bfc9",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" area \n",
" mean_intensity \n",
" major_axis_length \n",
" minor_axis_length \n",
" aspect_ratio \n",
" \n",
" \n",
" \n",
" \n",
" 0 \n",
" 294.0 \n",
" 36604.625850 \n",
" 25.656180 \n",
" 18.800641 \n",
" 1.364644 \n",
" \n",
" \n",
" 1 \n",
" 91.0 \n",
" 37379.769231 \n",
" 20.821990 \n",
" 6.053507 \n",
" 3.439658 \n",
" \n",
" \n",
" 2 \n",
" 246.0 \n",
" 44895.308943 \n",
" 21.830827 \n",
" 14.916032 \n",
" 1.463581 \n",
" \n",
" \n",
" 3 \n",
" 574.0 \n",
" 44394.637631 \n",
" 37.788705 \n",
" 19.624761 \n",
" 1.925563 \n",
" \n",
" \n",
" 4 \n",
" 518.0 \n",
" 45408.903475 \n",
" 26.917447 \n",
" 24.872908 \n",
" 1.082199 \n",
" \n",
" \n",
" ... \n",
" ... \n",
" ... \n",
" ... \n",
" ... \n",
" ... \n",
" \n",
" \n",
" 108 \n",
" 568.0 \n",
" 48606.121479 \n",
" 37.357606 \n",
" 19.808121 \n",
" 1.885974 \n",
" \n",
" \n",
" 109 \n",
" 175.0 \n",
" 25552.074286 \n",
" 17.419031 \n",
" 13.675910 \n",
" 1.273702 \n",
" \n",
" \n",
" 110 \n",
" 460.0 \n",
" 39031.419565 \n",
" 26.138592 \n",
" 23.522578 \n",
" 1.111213 \n",
" \n",
" \n",
" 111 \n",
" 407.0 \n",
" 39343.292383 \n",
" 28.544027 \n",
" 19.563792 \n",
" 1.459023 \n",
" \n",
" \n",
" 112 \n",
" 31.0 \n",
" 29131.322581 \n",
" 6.892028 \n",
" 5.711085 \n",
" 1.206781 \n",
" \n",
" \n",
"
\n",
"
113 rows \u00d7 5 columns
\n",
"
"
],
"text/plain": [
" area mean_intensity major_axis_length minor_axis_length aspect_ratio\n",
"0 294.0 36604.625850 25.656180 18.800641 1.364644\n",
"1 91.0 37379.769231 20.821990 6.053507 3.439658\n",
"2 246.0 44895.308943 21.830827 14.916032 1.463581\n",
"3 574.0 44394.637631 37.788705 19.624761 1.925563\n",
"4 518.0 45408.903475 26.917447 24.872908 1.082199\n",
".. ... ... ... ... ...\n",
"108 568.0 48606.121479 37.357606 19.808121 1.885974\n",
"109 175.0 25552.074286 17.419031 13.675910 1.273702\n",
"110 460.0 39031.419565 26.138592 23.522578 1.111213\n",
"111 407.0 39343.292383 28.544027 19.563792 1.459023\n",
"112 31.0 29131.322581 6.892028 5.711085 1.206781\n",
"\n",
"[113 rows x 5 columns]"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"dataframe['aspect_ratio'] = dataframe['major_axis_length'] / dataframe['minor_axis_length']\n",
"dataframe"
]
},
{
"cell_type": "markdown",
"id": "c9bafb0e-ee84-4c2d-9a6e-472c4c0f08ba",
"metadata": {},
"source": [
"## Plotten\n",
"Messungen k\u00f6nnen mithilfe von Diagrammen visualisiert werden."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "aba68386-4220-47ea-bf4a-8caf599b4920",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"seaborn.scatterplot(dataframe, x='area', y='aspect_ratio', hue='mean_intensity')"
]
},
{
"cell_type": "markdown",
"id": "aerial-release",
"metadata": {},
"source": [
"## Deskriptive Statistik\n",
"Ausgehend von dieser Tabelle k\u00f6nnen wir Statistiken verwenden, um einen \u00dcberblick \u00fcber die gemessenen Daten zu erhalten."
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "scheduled-motel",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Mean nucleus area is 524.4247787610619 +- 231.74703195433014 pixels\n"
]
}
],
"source": [
"mean_area = np.mean(dataframe['area'])\n",
"stddev_area = np.std(dataframe['area'])\n",
"\n",
"print(\"Die durchschnittliche Zellkernfl\u00e4che betr\u00e4gt\", mean_area, \"+-\", stddev_area, \"Pixel\")"
]
},
{
"cell_type": "markdown",
"id": "46804cc8-4307-4620-ba3f-6993c62182d7",
"metadata": {},
"source": [
"## Klassifizierung\n",
"F\u00fcr ein besseres Verst\u00e4ndnis der internen Struktur von Geweben, aber auch zur Korrektur von Artefakten in Bildverarbeitungs-Workflows, k\u00f6nnen wir Zellen klassifizieren, zum Beispiel nach ihrer Gr\u00f6\u00dfe und Form."
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "fb6d7f66-b209-4c32-8963-dfe61615b8b2",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"object_classifier = apoc.ObjectClassifier('../../data/blobs_classifier.cl')\n",
"classification_image = object_classifier.predict(label_image, cropped_image)\n",
"\n",
"stackview.imshow(classification_image)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.19"
}
},
"nbformat": 4,
"nbformat_minor": 5
}