| Applicability Criteria |
Diagnostic CT and PET/CT reports |
| Applicability Criteria |
Reports saved as a CSV file |
| Applicability Criteria |
Reports are written in English |
| Applicability Criteria |
CSV file containing 2 columns: report date and report |
| Applicability Criteria |
Thoracic DMG reports |
| Applicability Criteria |
Exclude MRI&USG reports |
| Applicability Criteria |
Exclude CT guided biopsy reports |
| Applicability Criteria |
Exclude notes without radiology report |
| Applicability Criteria |
Exclude reports with blanc rows |
| Applicability Criteria |
Exclude radiology reports without unique ID and not ICD-9 codes |
| Name |
Model Card metadata |
| Based on template |
https://repo.metadatacenter.org/templates/c1252f26-9075-4934-a8ee-51695b0b892d |
| Creation date |
2023-06-12 |
| Email |
snehacnair@gmail.com |
| Out-of-scope use cases |
None mentioned |
| Description |
Model Card implementation in CEDAR forms |
| Model Input |
Radiology report |
| Created on |
2023-12-08T03:56:24-08:00 |
| License |
Creative Commons |
| Primary intended use |
Automating cohort building from radiology reports |
| Created by |
S. Mithun |
| Outcome |
Lung cancer |
| Reference |
https://doi.org/10.1016/j.imu.2023.101294 |
| Primary intended users |
Medical professionals and researchers |
| Title |
Development and validation of deep learning and BERT models for classification of lung cancer radiology reports |
| Created by |
https://metadatacenter.org/users/0428302b-d7f9-45ff-a4ff-a27f62df6fb6 |
| Modified by |
https://metadatacenter.org/users/0428302b-d7f9-45ff-a4ff-a27f62df6fb6 |
| Foundational model or algorithm used |
Bi-LSTM (Bidirectional Long Short-Term Memory) neural networks, with and without dropout regularization, and a pre-trained BERT (Bidirectional Encoder Representations from Transformers) model |
| Last updated on |
2024-01-18T01:52:34-08:00 |
| Ethical considerations - Data |
The study was approved by the institutional ethics committee as a retrospective study with a waiver of informed consent |
| https://schema.metadatacenter.org/properties/5eefd8ad-66b4-4fff-b5a0-b5cb7b832d2d |
The study can be used to automate cohort building from radiology reports for lung cancer research |