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.
Configuration Fileconfig.yaml
This Potato config reproduces the annotation task. Save it as config.yaml and run potato start config.yaml to try it.
# CheXpert Chest X-Ray Classification Configuration
# Based on Irvin et al., AAAI 2019
annotation_task_name: "CheXpert Chest X-Ray Classification"
task_dir: "."
data_files:
- "sample-data.json"
item_properties:
id_key: "id"
text_key: "image_url"
context_key: "context"
user_config:
allow_all_users: true
annotation_schemes:
- annotation_type: "radio"
name: "cardiomegaly"
description: "Cardiomegaly (enlarged heart)"
labels:
- name: "positive"
tooltip: "Finding is present"
- name: "negative"
tooltip: "Finding is absent"
- name: "uncertain"
tooltip: "Cannot determine with certainty"
- name: "not_mentioned"
tooltip: "Not applicable or not assessed"
- annotation_type: "radio"
name: "edema"
description: "Pulmonary edema"
labels:
- name: "positive"
- name: "negative"
- name: "uncertain"
- name: "not_mentioned"
- annotation_type: "radio"
name: "consolidation"
description: "Lung consolidation"
labels:
- name: "positive"
- name: "negative"
- name: "uncertain"
- name: "not_mentioned"
- annotation_type: "radio"
name: "atelectasis"
description: "Atelectasis (lung collapse)"
labels:
- name: "positive"
- name: "negative"
- name: "uncertain"
- name: "not_mentioned"
- annotation_type: "radio"
name: "pleural_effusion"
description: "Pleural effusion (fluid around lungs)"
labels:
- name: "positive"
- name: "negative"
- name: "uncertain"
- name: "not_mentioned"
- annotation_type: "radio"
name: "pneumonia"
description: "Pneumonia"
labels:
- name: "positive"
- name: "negative"
- name: "uncertain"
- name: "not_mentioned"
- annotation_type: "radio"
name: "pneumothorax"
description: "Pneumothorax (collapsed lung)"
labels:
- name: "positive"
- name: "negative"
- name: "uncertain"
- name: "not_mentioned"
- annotation_type: "radio"
name: "no_finding"
description: "No significant findings"
labels:
- name: "positive"
tooltip: "Normal study, no findings"
- name: "negative"
tooltip: "Abnormalities present"
- annotation_type: "radio"
name: "image_quality"
description: "Rate the image quality"
labels:
- name: "diagnostic"
tooltip: "Adequate for diagnosis"
- name: "suboptimal"
tooltip: "Limited but usable"
- name: "non_diagnostic"
tooltip: "Inadequate for diagnosis"
interface_config:
item_display_format: "<img src='{{text}}' style='max-width:100%; max-height:500px; background:#000;'/><br/><small>{{context}}</small>"
output_annotation_format: "json"
output_annotation_dir: "annotations"
Sample Datasample-data.json
[
{
"id": "cxr_001",
"image_url": "https://upload.wikimedia.org/wikipedia/commons/c/c5/Normal_posteroanterior_%28PA%29_chest_radiograph_%28X-ray%29.jpg",
"context": "Chest X-ray PA view. Assess for cardiomegaly, edema, consolidation, atelectasis, pleural effusion, pneumonia, and pneumothorax."
},
{
"id": "cxr_002",
"image_url": "https://upload.wikimedia.org/wikipedia/commons/2/29/Chest_X-ray_of_RTI.jpg",
"context": "Chest radiograph. Label all visible pathological findings. Use 'uncertain' when findings are equivocal."
}
]
// ... and 1 more itemsTry it live — no install
Boot the real Potato server in your browser (WebAssembly) and annotate with this exact config. Nothing leaves your machine.
▶ Run live in your browserGet This Design
Clone or download from the repository
Quick start:
git clone https://github.com/davidjurgens/potato-showcase.git cd potato-showcase/image/medical/chexpert potato start config.yaml
Dataset & paper
Irvin et al., AAAI 2019
Citation (BibTeX)
@inproceedings{irvin2019chexpert,
title = "{CheXpert}: A Large Chest Radiograph Dataset with Uncertainty Labels and Expert Comparison",
author = "Irvin, Jeremy and Rajpurkar, Pranav and Ko, Michael and Yu, Yifan and Ciurea-Ilcus, Silviana and Chute, Chris and Marklund, Henrik and Haghgoo, Behzad and Ball, Robyn and Shpanskaya, Katie and others",
booktitle = "Proceedings of the AAAI Conference on Artificial Intelligence",
volume = "33",
pages = "590--597",
year = "2019",
url = "https://ojs.aaai.org/index.php/AAAI/article/view/3834"
}Details
Annotation Types
Domain
Use Cases
Tags
Found an issue or want to improve this design?
Open an IssueRelated Designs
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.
Camelyon17 - Breast Cancer Metastasis Detection in Pathology
Pathology slide annotation for breast cancer metastasis detection. Based on the CAMELYON grand challenge (Ehteshami Bejnordi et al., JAMA 2017), annotators delineate tumor regions in whole-slide histopathology images and classify slides as positive or negative for metastasis.
RadQA: Radiology Report Question Answering Dataset
RadQA is an extractive QA dataset of physician questions answered by spans in MIMIC-III radiology reports, including unanswerable questions. This Potato config reproduces the span-marking and answerability task.