Skip to content Python client library
Branch: master
Clone or download
Latest commit 0e5890e Aug 9, 2019 Python Client Library

Currently pre-alpha -- API may change significantly in future releases.

The python client library is designed to work with the datasets and annotations generated by the Medical AI platform.

You can download datasets consisting of images and annotations (i.e., JSON file), create train/validation/test datasets, integrate with various machine learing libraries (e.g., TensorFlow/Keras, for developing machine learning algorithms.

To get started, check out the examples in the notebooks section, or our intro to deep learning for medical imaging lessons.


Install and update using pip:

pip install --upgrade mdai


Overview Documentation: API Documentation coming soon.

The Annotator

The annotator is a powerful web based application, to store and view anonymized medical images (e.g, DICOM) on the cloud, create annotations collaboratively, in real-time, and export annotations, images and labels for training. The python client library can be used to download images and annotations, prepare the datasets, and then used to train and evaluate deep learning models. Annotator Annotation JSON Format

More detailed information regarding the annotation JSON export format, see:

Example Notebooks

Introductory lessons to Deep Learning for medical imaging by

The following are several Jupyter notebooks covering the basics of downloading and parsing annotation data, and training and evaluating different deep learning models for classification, semantic and instance segmentation and object detection problems in the medical imaging domain. The notebooks can be run on Google's colab with GPU (see instruction below).

  • Lesson 1. Classification of chest vs. adominal X-rays using TensorFlow/Keras Github | Annotator
  • Lesson 2. Lung X-Rays Semantic Segmentation using UNets. Github | Annotator
  • Lesson 3. RSNA Pneumonia detection using Kaggle data format Github | Annotator
  • Lesson 3. RSNA Pneumonia detection using python client library Github | Annotator


See contributing guidelines to set up a development environemnt and how to make contributions to mdai.

Running Jupyter notebooks Colab

It’s easy to run a Jupyter notebook on Google's Colab with free GPU use (time limited). For example, you can add the Github jupyter notebook path to Select the "GITHUB" tab, and add the Lesson 1 URL:

To use the GPU, in the notebook menu, go to Runtime -> Change runtime type -> switch to Python 3, and turn on GPU. See more colab tips.

Advanced: How to run on Google Cloud Platform with Deep Learning Images

You can also run the notebook with powerful GPUs on the Google Cloud Platform. In this case, you need to authenticate to the Google Cloug Platform, create a private virtual machine instance running a Google's Deep Learning image, and import the lessons. See instructions below.

GCP Deep Learnings Images How To

© 2019, Inc.

You can’t perform that action at this time.