Camelyon17 - Breast Cancer Metastasis Detection in Pathology
Pathology slide annotation for breast cancer metastasis detection. Based on the Camelyon17 challenge (Bejnordi et al., JAMA 2017), annotators delineate tumor regions in whole-slide histopathology images and classify slides as positive or negative for metastasis.
ملف الإعدادconfig.yaml
# Camelyon17 - Breast Cancer Metastasis Detection in Pathology
# Based on Bejnordi et al., JAMA 2017
# Paper: https://jamanetwork.com/journals/jama/fullarticle/2665774
# Dataset: https://camelyon17.grand-challenge.org/
#
# Pathology whole-slide image annotation for detecting breast cancer metastases
# in sentinel lymph node tissue. Annotators outline tumor regions using polygon
# and bounding box tools, then classify the overall slide-level diagnosis.
#
# Region Labels:
# - Tumor Region: Areas containing metastatic cancer cells
# - Normal Tissue: Healthy lymph node tissue
# - Artifact: Tissue folding, bubbles, or staining artifacts
# - Background: Non-tissue areas (glass, mounting medium)
#
# Annotation Guidelines:
# 1. Examine the histopathology image at available magnification
# 2. Use polygon tool for precise tumor boundary delineation
# 3. Use bounding box for quick region-of-interest marking
# 4. Classify the slide as positive, negative, or uncertain for metastasis
# 5. Tumor cells appear as dense, darkly stained clusters with irregular nuclei
annotation_task_name: "Camelyon17 - Pathology Metastasis Detection"
task_dir: "."
data_files:
- sample-data.json
item_properties:
id_key: "id"
text_key: "text"
output_annotation_dir: "annotation_output/"
output_annotation_format: "json"
port: 8000
server_name: localhost
annotation_schemes:
# Step 1: Annotate regions in the pathology image
- annotation_type: image_annotation
name: tissue_regions
description: "Delineate tissue regions in the pathology slide. Use polygon for precise boundaries, bounding box for quick marking."
tools:
- polygon
- bbox
labels:
- "Tumor Region"
- "Normal Tissue"
- "Artifact"
- "Background"
# Step 2: Slide-level classification
- annotation_type: radio
name: slide_diagnosis
description: "What is the overall slide-level diagnosis for metastasis?"
labels:
- "Positive (Metastasis Present)"
- "Negative (No Metastasis)"
- "Uncertain"
keyboard_shortcuts:
"Positive (Metastasis Present)": "1"
"Negative (No Metastasis)": "2"
"Uncertain": "3"
tooltips:
"Positive (Metastasis Present)": "One or more tumor regions are clearly identifiable in the slide"
"Negative (No Metastasis)": "No tumor cells are visible; the tissue appears entirely normal"
"Uncertain": "Suspicious regions are present but a definitive diagnosis cannot be made"
annotation_instructions: |
You will annotate histopathology images of sentinel lymph node tissue sections
for the presence of breast cancer metastases.
For each slide:
1. Examine the tissue section carefully at available magnification.
2. Use the polygon tool to precisely outline any tumor regions you identify.
- Tumor cells typically appear as dense clusters with large, irregular,
darkly stained nuclei.
- Metastases may appear as isolated tumor cells (ITC), micrometastases,
or macrometastases.
3. Use the bounding box tool for quick marking of suspicious areas.
4. Label non-tumor regions as Normal Tissue, Artifact, or Background.
5. Provide an overall slide-level diagnosis.
Important:
- Artifacts (tissue folds, air bubbles, poor staining) should not be confused with tumor.
- When uncertain, mark as Uncertain and outline the suspicious region.
- This task requires familiarity with histopathology.
html_layout: |
<div style="padding: 15px; max-width: 900px; margin: auto;">
<div style="display: flex; gap: 12px; margin-bottom: 14px; flex-wrap: wrap;">
<div style="background: #fce4ec; padding: 7px 14px; border-radius: 8px;">
<strong>Tissue Type:</strong> {{tissue_type}}
</div>
</div>
<div style="text-align: center; margin-bottom: 16px; background: #212121; padding: 12px; border-radius: 8px;">
<img src="{{image_url}}" style="max-width: 100%; max-height: 600px; border-radius: 4px;" alt="Histopathology slide" />
</div>
<div style="background: #f0f9ff; border: 1px solid #bae6fd; border-radius: 8px; padding: 16px; margin-bottom: 16px;">
<strong style="color: #0369a1;">Slide Description:</strong>
<p style="font-size: 15px; line-height: 1.7; margin: 8px 0 0 0;">{{text}}</p>
</div>
</div>
allow_all_users: true
instances_per_annotator: 30
annotation_per_instance: 3
allow_skip: true
skip_reason_required: false
بيانات نموذجيةsample-data.json
[
{
"id": "camelyon_001",
"text": "Sentinel lymph node section from left axilla, H&E stained, 20x magnification. Dense cellular region visible in the subcapsular sinus area with irregular nuclear morphology.",
"image_url": "https://example.com/camelyon/slide_001.png",
"tissue_type": "Sentinel Lymph Node"
},
{
"id": "camelyon_002",
"text": "Lymph node tissue section showing predominantly normal germinal centers with reactive follicular hyperplasia. No obvious atypical cells at scanning magnification.",
"image_url": "https://example.com/camelyon/slide_002.png",
"tissue_type": "Sentinel Lymph Node"
}
]
// ... and 8 more itemsاحصل على هذا التصميم
Clone or download from the repository
بدء سريع:
git clone https://github.com/davidjurgens/potato-showcase.git cd potato-showcase/image/medical/camelyon-pathology potato start config.yaml
التفاصيل
أنواع التوسيم
المجال
حالات الاستخدام
الوسوم
وجدت مشكلة أو تريد تحسين هذا التصميم؟
افتح مشكلةتصاميم ذات صلة
CHART-Infographics: Chart and Infographic Analysis
Chart and infographic analysis with structured extraction. Annotators identify chart elements (axes, legends, data points, titles) with bounding boxes, classify chart types, and extract data values. Supports structured understanding of visual data representations.
CheXpert Chest X-Ray Classification
Multi-label classification of chest radiographs for 14 observations (Irvin et al., AAAI 2019). Annotate chest X-rays with pathology labels including uncertainty handling for clinical findings.
MIMIC-CXR Chest Radiograph Classification
Large-scale chest radiograph classification based on MIMIC-CXR (Johnson et al., Scientific Data 2019). Multi-label classification with 14 observations derived from radiology reports.