Overview of the VISCERAL Retrieval Benchmark 2015

• Published in: Multimodal Retrieval in the Medical Domain pp. 115 - 123
• Main Authors: Jiménez–del–Toro, Oscar Alfonso, Hanbury, Allan, Langs, Georg, Foncubierta–Rodríguez, Antonio, Müller, Henning
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing 2015
• Series: Lecture Notes in Computer Science
• Subjects: Information retrieval infrastructures; Medical content–based retrieval; Multimodal retrieval
• ISBN: 3319244701; 9783319244709
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-24471-6_10

The results of the VISCERAL 3D case retrieval benchmark were presented during the Multimodal Retrieval in the Medical Domain (MRMD) 2015 workshop in Vienna, Austria on March 29, 2015. The main task for the participanta was to find and rank similar medical cases from a large multimodal (semantic RadLex terms extracted from text and visual 3D data) data set using a query case as input. The approaches that integrated information from both the RadLex terms and the 3D volumes provided in the benchmark obtained the best results based on 5 standard evaluation metrics. The benchmark set up, data set description and result analysis from the benchmark are presented for all the submitted methods.