Training deep classifiers for B-line detection in lung ultrasound videos using crowdsourced labels

Draft Status

Ready - team will start page creating immediately

Presenter Location

In-person

Key Investigators

Mike Jin (Brigham and Women's Hospital, USA)
Tamas Ungi (Queen's University, Canada)
Ameneh Asgari-Targhi (Brigham and Women's Hospital, USA)
Tina Kapur (Brigham and Women's Hospital, USA)

Project Description

Crowdsourced labels on medical imaging data can bridge the gap in labeled data required for training AI models for clinical applications. Using crowdsourced labels on lung ultrasound videos that have been shown (Duggan 2023) to have comparable accuracy to medical experts, we will train a deep learning model using an approach (Lucassen 2023) shown to work with expert-created labels and use the previous model performance to compare the new model's performance.

Objective

Using crowdsourced labels, train a deep learning model to classify lung ultrasound videos as having B-lines or having no B-lines.

Approach and Plan

Run model training and evaluation code from (Lucassen 2023) on crowdsourced labels
Compare model performance to same model (Lucassen 2023) trained on expert-created labels.

Progress and Next Steps

Describe specific steps you have actually done.

Illustrations

No response

Background and References

No response

NA-MIC / ProjectWeek