Validation of the deep learning models on NLST data (Chest radiographs)
Principal Investigator
Name
yoona hwang
Degrees
Ph.D.
Institution
VUNO Inc.
Position Title
team lead (researcher)
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1274
Initial CDAS Request Approval
Jul 8, 2024
Title
Validation of the deep learning models on NLST data (Chest radiographs)
Summary
We have developed deep-learning algorithms that can detect lung nodules and predict the risk of malignancy on chest radiographs.
Most deep learning algorithms perform well when tested on datasets that are similar to the training set. However, the NLST data is sourced from the US population and has not been seen by our algorithms before. It is a valuable resource for testing our nodule detection and malignancy risk prediction algorithms against radiology opinions and confirmed diagnoses. Our project aims to evaluate the performance of pre-trained deep learning algorithms on the NLST data.
Most deep learning algorithms perform well when tested on datasets that are similar to the training set. However, the NLST data is sourced from the US population and has not been seen by our algorithms before. It is a valuable resource for testing our nodule detection and malignancy risk prediction algorithms against radiology opinions and confirmed diagnoses. Our project aims to evaluate the performance of pre-trained deep learning algorithms on the NLST data.
Aims
- Evaluation is needed to generalize the detection of nodules and other findings on the chest X-ray.
- The performance of our models in detecting features suggestive of lung cancer on chest X-rays.
- The performance of our models in detecting other pathologies on chest X-rays.
- The ability of our models to localize nodules/masses on chest X-rays.
By using the final indications and diagnoses provided by biopsy or CT scans, we aim to assess whether our algorithm could have aided in the early detection of actionable nodules and other abnormal findings on chest radiographs, potentially reducing the number of cases missed by radiologists.
Collaborators
All members of the R&D team of VUNO Inc.,