# Summary #
1. Purpose: To develop computer-aided diagnostic (CAD) methods for lung nodule detection, classification, and quantitative assessment, they established the Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI) to complete the database.
2. Methods: The LIDC/IDRI Database contains 1018 cases, each consisting of images from a clinical thoracic CT scan and an associated XML file that records the results of a two-phase image annotation process performed by four experienced thoracic radiologists.
3. Results: The database contains 7371 lesions marked as “nodules” by at least one radiologist, among which 2669 were marked as “nodule ≥ 3 mm” by at least one radiologist, with 928 (34.7%) receiving such marks from all four radiologists.
# Review #
1. Introduction
- The collections of images acquired during comprehensive lung cancer screening trials have the potential to become valuable database resources.
- Publicly available image databases designed to facilitate computerized image analysis research were first introduced in mammography (Digital Database for Screening Mammography; DDSM).
- The National Lung Screening Trial (NLST) randomized 26,724 subjects to the CT screening.
- The NELSON (Nederlands Leuvens Longkanker Screeningsonderzoek) trial has accrued 15,523 participants across four institutions since 2003.
- The Lung Image Database Consortium (LIDC), which has been working since 2001 to develop a web-accessible research resource for the development, training, and evaluation of CAD methods for lung nodules, collaborated with the Foundation for the National Institutes of Health (FNIH) to create the Image Database Resource Initiative (IDRI) in 2004, further advancing the efforts of the LIDC.
- The IDRI joined the five LIDC institutions with two additional academic centers, and eight medical imaging companies.
- LIDC-IDRI include, (1) an image repository of screening and diagnostic thoracic CT scans, (2) associated metadata such as technical scan parameters and patient information, (3) nodule truth information based on the subjective assessments of multiple experienced radiologists.
- This paper aims to describe the now-completed, publicly available LIDC/IDRI Database of 1018 thoracic CT scans and associated radiologist annotations.
- A total of 12 radiologists participated in the image annotation process across all five sites over the course of the project.
- In the initial blinded-read phase, each radiologist independently reviewed each CT scan and marked lesions belonging to one of three categories (“nodule >= 3 mm,” “nodule < 3 mm,” and “non-nodule >= 3 mm”).
- In the subsequent unblinded-read phase, each radiologist independently reviewed their own marks along with the anonymized marks of the three other radiologists to render a final opinion.
2. Methods and Results
* Reference: Armato III, Samuel G., et al. "The lung image database consortium (LIDC) and image database resource initiative (IDRI): a completed reference database of lung nodules on CT scans." Medical physics 38.2 (2011): 915-931
댓글