Overview of the CLEF 2009 medical image retrieval track

  • Authors:
  • Henning Müller;Jayashree Kalpathy-Cramer;Ivan Eggel;Steven Bedrick;Saïd Radhouani;Brian Bakke;Charles E. Kahn, Jr.;William Hersh

  • Affiliations:
  • Geneva University Hospitals and University of Geneva, Switzerland and University of Applied Sciences Western Switzerland, Sierre, Switzerland;Oregon Health and Science University, Portland, OR;Geneva University Hospitals and University of Geneva, Switzerland;Oregon Health and Science University, Portland, OR;Oregon Health and Science University, Portland, OR;Oregon Health and Science University, Portland, OR;Department of Radiology, Medical College of Wisconsin, Milwaukee, WI;Oregon Health and Science University, Portland, OR

  • Venue:
  • CLEF'09 Proceedings of the 10th international conference on Cross-language evaluation forum: multimedia experiments
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

2009 was the sixth year for the ImageCLEF medical retrieval task. Participation was strong again with 38 registered research groups. 17 groups submitted runs and thus participated actively in the tasks. The database in 2009 was similar to the one used in 2008, containing scientific articles from two radiology journals, Radiology and Radiographics. The size of the database was increased to a total of 74,902 images. For the first time, 5 case-based topics were provided as an exploratory task. These topics' unit of retrieval was intended to be the source article and not the image itself. Case-based topics are designed to be closer to the clinical workflow, as clinicians often seek information about patient cases using incomplete information consisting of symptoms, findings, and a set of images. Supplying cases to a clinician from the scientific literature that are similar to the case (s)he is treating models what may become an important application of image retrieval in the future. We also introduced a lung nodule detection task in 2009. This task used the CT slices from the Lung Imaging Data Consortium (LIDC) includeding ground truth in the from of manual annotations. The goal of this task was to create algorithms to automatically detect lung nodules. Although there seemed to be significant interest in the task only two groups submitted results with a proprietary software from an industry participant achieving very good results.