Pulmonary ultrasound may be an alternative tool to screen patients with or suspected of having COVID-19. A prospective study carried out at the AP-HP evaluated the correlation between a pulmonary ultrasound severity score and the unfavorable clinical evolution of the patient under 28 days.

In this data challenge, you will need to use a machine Learning approach on data from this study to construct a global predictor of the patient's clinical outcome.

More details can be found HERE

At the end of the hackathon, you will be required to share your methods and source code with the open community under CC by NC 2.0 license

We suggest the following submission template:

Name of your project

A short description of what makes your project unique !

• Ada Lovelace
• Claude Shannon
• Stephen Hawking

Explain in details the preprocessing and training methods(dataset split and stratification, hyperparameters selection) you are using to build your global predictor of the patient's clinical outcome. You will also provide a link to the source code implementing these different steps.

The model performances are evaluated on a hidden subset of the dataset following the F1-score applied to patient's clinical outcome variable.

In this section, you will provide meaningfull visualizations and interpretations that will help to understand the underlying dynamic of the model decision. It will nurture a medical discussion with the radiologists involved in the challenge.

• Epita
• echOpen foundation
• AP-HP
• EIT Health