Medical image retrieval and automatic annotation: OHSU at ImageCLEF 2007

Jayashree Kalpathy-Cramer, William (Bill) Hersh

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Oregon Health & Science University participated in the medical retrieval and medical annotation tasks of ImageCLEF 2007. In the medical retrieval task, we created a webbased retrieval system for the collection built on a full-text index of both image and case annotations. The text-based search engine was implemented in Ruby using Ferret, a port of Lucene, and a custom query parser. In addition to this textual index of annotations, supervised machine learning techniques using visual features were used to classify the images based on image acquisition modality. All images were annotated with the purported modality. Purely textual runs as well as mixed runs using the purported modality were submitted. Our runs performed moderately well using the MAP metric and better for the early precision (P10) metric. In the automatic annotation task, we used the 'gist' technique to create the feature vectors. Using statistics derived from a set of multi-scale oriented filters, we created a 512 dimensional vector. PCA was then used to create a 100-dimensional vector. This feature vector was fed into a two layer neural network. Our error rate on the 1000 test images was 67.8 using the hierarchical error calculations.

Original languageEnglish (US)
Title of host publicationCLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007
PublisherCEUR-WS
Volume1173
StatePublished - 2007
Event2007 Cross Language Evaluation Forum Workshop, CLEF 2007, co-located with the 11th European Conference on Digital Libraries, ECDL 2007 - Budapest, Hungary
Duration: Sep 19 2007Sep 21 2007

Other

Other2007 Cross Language Evaluation Forum Workshop, CLEF 2007, co-located with the 11th European Conference on Digital Libraries, ECDL 2007
CountryHungary
CityBudapest
Period9/19/079/21/07

Fingerprint

Image retrieval
Ruby
Image acquisition
Search engines
Learning systems
Health
Statistics
Neural networks

Keywords

  • Image modality classification
  • Neural networks
  • Query parsing
  • Text retrieval

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Kalpathy-Cramer, J., & Hersh, W. B. (2007). Medical image retrieval and automatic annotation: OHSU at ImageCLEF 2007. In CLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007 (Vol. 1173). CEUR-WS.

Medical image retrieval and automatic annotation : OHSU at ImageCLEF 2007. / Kalpathy-Cramer, Jayashree; Hersh, William (Bill).

CLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007. Vol. 1173 CEUR-WS, 2007.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kalpathy-Cramer, J & Hersh, WB 2007, Medical image retrieval and automatic annotation: OHSU at ImageCLEF 2007. in CLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007. vol. 1173, CEUR-WS, 2007 Cross Language Evaluation Forum Workshop, CLEF 2007, co-located with the 11th European Conference on Digital Libraries, ECDL 2007, Budapest, Hungary, 9/19/07.
Kalpathy-Cramer J, Hersh WB. Medical image retrieval and automatic annotation: OHSU at ImageCLEF 2007. In CLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007. Vol. 1173. CEUR-WS. 2007
Kalpathy-Cramer, Jayashree ; Hersh, William (Bill). / Medical image retrieval and automatic annotation : OHSU at ImageCLEF 2007. CLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007. Vol. 1173 CEUR-WS, 2007.
@inproceedings{b93a3591b27d4a0d9dff8a2b501777c8,
title = "Medical image retrieval and automatic annotation: OHSU at ImageCLEF 2007",
abstract = "Oregon Health & Science University participated in the medical retrieval and medical annotation tasks of ImageCLEF 2007. In the medical retrieval task, we created a webbased retrieval system for the collection built on a full-text index of both image and case annotations. The text-based search engine was implemented in Ruby using Ferret, a port of Lucene, and a custom query parser. In addition to this textual index of annotations, supervised machine learning techniques using visual features were used to classify the images based on image acquisition modality. All images were annotated with the purported modality. Purely textual runs as well as mixed runs using the purported modality were submitted. Our runs performed moderately well using the MAP metric and better for the early precision (P10) metric. In the automatic annotation task, we used the 'gist' technique to create the feature vectors. Using statistics derived from a set of multi-scale oriented filters, we created a 512 dimensional vector. PCA was then used to create a 100-dimensional vector. This feature vector was fed into a two layer neural network. Our error rate on the 1000 test images was 67.8 using the hierarchical error calculations.",
keywords = "Image modality classification, Neural networks, Query parsing, Text retrieval",
author = "Jayashree Kalpathy-Cramer and Hersh, {William (Bill)}",
year = "2007",
language = "English (US)",
volume = "1173",
booktitle = "CLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007",
publisher = "CEUR-WS",

}

TY - GEN

T1 - Medical image retrieval and automatic annotation

T2 - OHSU at ImageCLEF 2007

AU - Kalpathy-Cramer, Jayashree

AU - Hersh, William (Bill)

PY - 2007

Y1 - 2007

N2 - Oregon Health & Science University participated in the medical retrieval and medical annotation tasks of ImageCLEF 2007. In the medical retrieval task, we created a webbased retrieval system for the collection built on a full-text index of both image and case annotations. The text-based search engine was implemented in Ruby using Ferret, a port of Lucene, and a custom query parser. In addition to this textual index of annotations, supervised machine learning techniques using visual features were used to classify the images based on image acquisition modality. All images were annotated with the purported modality. Purely textual runs as well as mixed runs using the purported modality were submitted. Our runs performed moderately well using the MAP metric and better for the early precision (P10) metric. In the automatic annotation task, we used the 'gist' technique to create the feature vectors. Using statistics derived from a set of multi-scale oriented filters, we created a 512 dimensional vector. PCA was then used to create a 100-dimensional vector. This feature vector was fed into a two layer neural network. Our error rate on the 1000 test images was 67.8 using the hierarchical error calculations.

AB - Oregon Health & Science University participated in the medical retrieval and medical annotation tasks of ImageCLEF 2007. In the medical retrieval task, we created a webbased retrieval system for the collection built on a full-text index of both image and case annotations. The text-based search engine was implemented in Ruby using Ferret, a port of Lucene, and a custom query parser. In addition to this textual index of annotations, supervised machine learning techniques using visual features were used to classify the images based on image acquisition modality. All images were annotated with the purported modality. Purely textual runs as well as mixed runs using the purported modality were submitted. Our runs performed moderately well using the MAP metric and better for the early precision (P10) metric. In the automatic annotation task, we used the 'gist' technique to create the feature vectors. Using statistics derived from a set of multi-scale oriented filters, we created a 512 dimensional vector. PCA was then used to create a 100-dimensional vector. This feature vector was fed into a two layer neural network. Our error rate on the 1000 test images was 67.8 using the hierarchical error calculations.

KW - Image modality classification

KW - Neural networks

KW - Query parsing

KW - Text retrieval

UR - http://www.scopus.com/inward/record.url?scp=84921941105&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84921941105&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84921941105

VL - 1173

BT - CLEF 2007 - Working Notes for CLEF 2007 Workshop, co-located with the 11th European Conference on Digital Libraries, ECDL 2007

PB - CEUR-WS

ER -