Learning a Frequency–Based Weighting for Medical Image Classification View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

2008-01-01

AUTHORS

Tobias Gass , Adrien Depeursinge , Antoine Geissbuhler , Henning Müller

ABSTRACT

This article describes the use of a frequency–based weighting developed for image retrieval to perform automatic annotation of images (medical and non–medical). The techniques applied are based on a simple tf/idf (term frequency, inverse document frequency) weighting scheme of GIFT (GNU Image Finding Tool), which is augmented by feature weights extracted from training data. The additional weights represent a measure of discrimination by taking into account the number of occurrences of the features in pairs of images of the same class or in pairs of images from different classes. The approach is fit to the image classification task by pruning parts of the training data. Further investigations were performed showing that weightings lead to significantly worse classification quality in certain feature domains. A classifier using a mixture of tf/idf weighted scoring, learned feature weights, and regular Euclidean distance gave best results using only the simple features. Using the aspect–ratio of images as feature improved results significantly. More... »

PAGES

99-108

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-540-79490-5_14

DOI

http://dx.doi.org/10.1007/978-3-540-79490-5_14

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1001067383


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Information and Computing Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0801", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Artificial Intelligence and Image Processing", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "Lehrstuhl f\u00fcr Informatik 6, RWTH Aachen, Germany", 
          "id": "http://www.grid.ac/institutes/grid.1957.a", 
          "name": [
            "Lehrstuhl f\u00fcr Informatik 6, RWTH Aachen, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Gass", 
        "givenName": "Tobias", 
        "id": "sg:person.01230434667.05", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01230434667.05"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Medical Informatics, University and Hospitals of Geneva, Switzerland", 
          "id": "http://www.grid.ac/institutes/grid.150338.c", 
          "name": [
            "Medical Informatics, University and Hospitals of Geneva, Switzerland"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Depeursinge", 
        "givenName": "Adrien", 
        "id": "sg:person.01132575260.26", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01132575260.26"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Medical Informatics, University and Hospitals of Geneva, Switzerland", 
          "id": "http://www.grid.ac/institutes/grid.150338.c", 
          "name": [
            "Medical Informatics, University and Hospitals of Geneva, Switzerland"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Geissbuhler", 
        "givenName": "Antoine", 
        "id": "sg:person.0600360343.20", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0600360343.20"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Medical Informatics, University and Hospitals of Geneva, Switzerland", 
          "id": "http://www.grid.ac/institutes/grid.150338.c", 
          "name": [
            "Medical Informatics, University and Hospitals of Geneva, Switzerland"
          ], 
          "type": "Organization"
        }, 
        "familyName": "M\u00fcller", 
        "givenName": "Henning", 
        "id": "sg:person.07552063233.67", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07552063233.67"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2008-01-01", 
    "datePublishedReg": "2008-01-01", 
    "description": "This article describes the use of a frequency\u2013based weighting developed for image retrieval to perform automatic annotation of images (medical and non\u2013medical). The techniques applied are based on a simple tf/idf (term frequency, inverse document frequency) weighting scheme of GIFT (GNU Image Finding Tool), which is augmented by feature weights extracted from training data. The additional weights represent a measure of discrimination by taking into account the number of occurrences of the features in pairs of images of the same class or in pairs of images from different classes. The approach is fit to the image classification task by pruning parts of the training data. Further investigations were performed showing that weightings lead to significantly worse classification quality in certain feature domains. A classifier using a mixture of tf/idf weighted scoring, learned feature weights, and regular Euclidean distance gave best results using only the simple features. Using the aspect\u2013ratio of images as feature improved results significantly.", 
    "editor": [
      {
        "familyName": "Gao", 
        "givenName": "Xiaohong", 
        "type": "Person"
      }, 
      {
        "familyName": "M\u00fcller", 
        "givenName": "Henning", 
        "type": "Person"
      }, 
      {
        "familyName": "Loomes", 
        "givenName": "Martin J.", 
        "type": "Person"
      }, 
      {
        "familyName": "Comley", 
        "givenName": "Richard", 
        "type": "Person"
      }, 
      {
        "familyName": "Luo", 
        "givenName": "Shuqian", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-540-79490-5_14", 
    "inLanguage": "en", 
    "isAccessibleForFree": false, 
    "isPartOf": {
      "isbn": [
        "978-3-540-79489-9", 
        "978-3-540-79490-5"
      ], 
      "name": "Medical Imaging and Informatics", 
      "type": "Book"
    }, 
    "keywords": [
      "pair of images", 
      "training data", 
      "feature weights", 
      "medical image classification", 
      "image classification tasks", 
      "TF/IDF", 
      "image retrieval", 
      "automatic annotation", 
      "image classification", 
      "feature domain", 
      "classification task", 
      "classification quality", 
      "simple features", 
      "Euclidean distance", 
      "weighting scheme", 
      "images", 
      "number of occurrences", 
      "same class", 
      "good results", 
      "classifier", 
      "different classes", 
      "retrieval", 
      "annotation", 
      "weighting", 
      "features", 
      "task", 
      "scheme", 
      "classification", 
      "class", 
      "data", 
      "domain", 
      "technique", 
      "quality", 
      "additional weight", 
      "IDF", 
      "results", 
      "pairs", 
      "number", 
      "distance", 
      "use", 
      "part", 
      "account", 
      "measures", 
      "article", 
      "weight", 
      "discrimination", 
      "frequency", 
      "occurrence", 
      "investigation", 
      "gift", 
      "further investigation", 
      "measures of discrimination", 
      "mixture", 
      "approach", 
      "frequency\u2013based weighting", 
      "simple tf/idf (term frequency, inverse document frequency) weighting scheme", 
      "tf/idf (term frequency, inverse document frequency) weighting scheme", 
      "idf (term frequency, inverse document frequency) weighting scheme", 
      "worse classification quality", 
      "certain feature domains", 
      "regular Euclidean distance"
    ], 
    "name": "Learning a Frequency\u2013Based Weighting for Medical Image Classification", 
    "pagination": "99-108", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1001067383"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-540-79490-5_14"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-540-79490-5_14", 
      "https://app.dimensions.ai/details/publication/pub.1001067383"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2021-12-01T20:02", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20211201/entities/gbq_results/chapter/chapter_258.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-540-79490-5_14"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/978-3-540-79490-5_14'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-540-79490-5_14'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-540-79490-5_14'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-540-79490-5_14'


 

This table displays all metadata directly associated to this object as RDF triples.

165 TRIPLES      23 PREDICATES      86 URIs      79 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-540-79490-5_14 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N3843581cd2c744a88f6773064470c2cc
4 schema:datePublished 2008-01-01
5 schema:datePublishedReg 2008-01-01
6 schema:description This article describes the use of a frequency–based weighting developed for image retrieval to perform automatic annotation of images (medical and non–medical). The techniques applied are based on a simple tf/idf (term frequency, inverse document frequency) weighting scheme of GIFT (GNU Image Finding Tool), which is augmented by feature weights extracted from training data. The additional weights represent a measure of discrimination by taking into account the number of occurrences of the features in pairs of images of the same class or in pairs of images from different classes. The approach is fit to the image classification task by pruning parts of the training data. Further investigations were performed showing that weightings lead to significantly worse classification quality in certain feature domains. A classifier using a mixture of tf/idf weighted scoring, learned feature weights, and regular Euclidean distance gave best results using only the simple features. Using the aspect–ratio of images as feature improved results significantly.
7 schema:editor Nafa9cc2ecab44bd79b0c97ef0bd1f53f
8 schema:genre chapter
9 schema:inLanguage en
10 schema:isAccessibleForFree false
11 schema:isPartOf N955c9d09d7ca4262b5033489fe907179
12 schema:keywords Euclidean distance
13 IDF
14 TF/IDF
15 account
16 additional weight
17 annotation
18 approach
19 article
20 automatic annotation
21 certain feature domains
22 class
23 classification
24 classification quality
25 classification task
26 classifier
27 data
28 different classes
29 discrimination
30 distance
31 domain
32 feature domain
33 feature weights
34 features
35 frequency
36 frequency–based weighting
37 further investigation
38 gift
39 good results
40 idf (term frequency, inverse document frequency) weighting scheme
41 image classification
42 image classification tasks
43 image retrieval
44 images
45 investigation
46 measures
47 measures of discrimination
48 medical image classification
49 mixture
50 number
51 number of occurrences
52 occurrence
53 pair of images
54 pairs
55 part
56 quality
57 regular Euclidean distance
58 results
59 retrieval
60 same class
61 scheme
62 simple features
63 simple tf/idf (term frequency, inverse document frequency) weighting scheme
64 task
65 technique
66 tf/idf (term frequency, inverse document frequency) weighting scheme
67 training data
68 use
69 weight
70 weighting
71 weighting scheme
72 worse classification quality
73 schema:name Learning a Frequency–Based Weighting for Medical Image Classification
74 schema:pagination 99-108
75 schema:productId N6a828594bc1240f490018effde22f576
76 N8fc7f5cec20e44a5b69247a3d7011690
77 schema:publisher Ncd8ccab5f0194a9e8a09859a2315cfc7
78 schema:sameAs https://app.dimensions.ai/details/publication/pub.1001067383
79 https://doi.org/10.1007/978-3-540-79490-5_14
80 schema:sdDatePublished 2021-12-01T20:02
81 schema:sdLicense https://scigraph.springernature.com/explorer/license/
82 schema:sdPublisher N3d48032b4a3c4b14a1ae260b82d9eb1c
83 schema:url https://doi.org/10.1007/978-3-540-79490-5_14
84 sgo:license sg:explorer/license/
85 sgo:sdDataset chapters
86 rdf:type schema:Chapter
87 N2ca35c01a6574732ad76253100067d5f schema:familyName Luo
88 schema:givenName Shuqian
89 rdf:type schema:Person
90 N3843581cd2c744a88f6773064470c2cc rdf:first sg:person.01230434667.05
91 rdf:rest Nba7e4079fc7744ae9c2401e734273cca
92 N3d48032b4a3c4b14a1ae260b82d9eb1c schema:name Springer Nature - SN SciGraph project
93 rdf:type schema:Organization
94 N500ac25bc1ba4e6e83b6078d3ba63c06 rdf:first N2ca35c01a6574732ad76253100067d5f
95 rdf:rest rdf:nil
96 N55735b69c2264cf6977122b5261008ef schema:familyName Loomes
97 schema:givenName Martin J.
98 rdf:type schema:Person
99 N6a828594bc1240f490018effde22f576 schema:name dimensions_id
100 schema:value pub.1001067383
101 rdf:type schema:PropertyValue
102 N7b0f6d4096cb46949ad73be77ec8611b schema:familyName Müller
103 schema:givenName Henning
104 rdf:type schema:Person
105 N88bb5e12a26244b4a7df0a999415c5b0 rdf:first sg:person.0600360343.20
106 rdf:rest Nf797067912e44bd483088dced39f87e4
107 N8fc7f5cec20e44a5b69247a3d7011690 schema:name doi
108 schema:value 10.1007/978-3-540-79490-5_14
109 rdf:type schema:PropertyValue
110 N955c9d09d7ca4262b5033489fe907179 schema:isbn 978-3-540-79489-9
111 978-3-540-79490-5
112 schema:name Medical Imaging and Informatics
113 rdf:type schema:Book
114 Nafa9cc2ecab44bd79b0c97ef0bd1f53f rdf:first Ne2c992228dc3434ab850ae77303feec3
115 rdf:rest Ne37ff65528d84d5586170004aaf40b0e
116 Nba7e4079fc7744ae9c2401e734273cca rdf:first sg:person.01132575260.26
117 rdf:rest N88bb5e12a26244b4a7df0a999415c5b0
118 Ncd2d570801144fec824b71058c73f83b rdf:first Nd1a420bd974a424da45d79410e06585d
119 rdf:rest N500ac25bc1ba4e6e83b6078d3ba63c06
120 Ncd8ccab5f0194a9e8a09859a2315cfc7 schema:name Springer Nature
121 rdf:type schema:Organisation
122 Nd1a420bd974a424da45d79410e06585d schema:familyName Comley
123 schema:givenName Richard
124 rdf:type schema:Person
125 Ne2c992228dc3434ab850ae77303feec3 schema:familyName Gao
126 schema:givenName Xiaohong
127 rdf:type schema:Person
128 Ne37ff65528d84d5586170004aaf40b0e rdf:first N7b0f6d4096cb46949ad73be77ec8611b
129 rdf:rest Nf6598c3054c341aeac2c0c3deb83bbb7
130 Nf6598c3054c341aeac2c0c3deb83bbb7 rdf:first N55735b69c2264cf6977122b5261008ef
131 rdf:rest Ncd2d570801144fec824b71058c73f83b
132 Nf797067912e44bd483088dced39f87e4 rdf:first sg:person.07552063233.67
133 rdf:rest rdf:nil
134 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
135 schema:name Information and Computing Sciences
136 rdf:type schema:DefinedTerm
137 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
138 schema:name Artificial Intelligence and Image Processing
139 rdf:type schema:DefinedTerm
140 sg:person.01132575260.26 schema:affiliation grid-institutes:grid.150338.c
141 schema:familyName Depeursinge
142 schema:givenName Adrien
143 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01132575260.26
144 rdf:type schema:Person
145 sg:person.01230434667.05 schema:affiliation grid-institutes:grid.1957.a
146 schema:familyName Gass
147 schema:givenName Tobias
148 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01230434667.05
149 rdf:type schema:Person
150 sg:person.0600360343.20 schema:affiliation grid-institutes:grid.150338.c
151 schema:familyName Geissbuhler
152 schema:givenName Antoine
153 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0600360343.20
154 rdf:type schema:Person
155 sg:person.07552063233.67 schema:affiliation grid-institutes:grid.150338.c
156 schema:familyName Müller
157 schema:givenName Henning
158 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07552063233.67
159 rdf:type schema:Person
160 grid-institutes:grid.150338.c schema:alternateName Medical Informatics, University and Hospitals of Geneva, Switzerland
161 schema:name Medical Informatics, University and Hospitals of Geneva, Switzerland
162 rdf:type schema:Organization
163 grid-institutes:grid.1957.a schema:alternateName Lehrstuhl für Informatik 6, RWTH Aachen, Germany
164 schema:name Lehrstuhl für Informatik 6, RWTH Aachen, Germany
165 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...