Lesion-aware attention with neural support vector machine for retinopathy diagnosis View Full Text


Ontology type: schema:ScholarlyArticle     


Article Info

DATE

2021-10-18

AUTHORS

Nagur Shareef Shaik, Teja Krishna Cherukuri

ABSTRACT

Diabetic retinopathy (DR) is a severe eye disease which can lead to permanent blindness. Identifying DR in early stages by using computer-aided diagnosis (CAD) systems can help the ophthalmologists to give proper treatment rationally, there by preventing many people from going blind. Due to intra-class variations and imbalanced data distribution, it is highly difficult to design a CAD system for DR severity diagnosis with greater generalizability. In this article, we propose a multi-stage deep learning pipeline, lesion-aware attention with neural support vector machine, for diabetic retinopathy diagnosis. Proposed pipeline consists of a pre-trained convolution base for learning retinal image spatial representations, lesion-aware attention for weighting lesion specific features, convolution autoencoder for learning latent attention representations and a neural support vector machine for discrimination. Convolutional autoencoder and neural support vector machine are jointly trained in end-to-end fashion to obtain category based lesion specific latent attention features by complementing each other in re-constructor and discriminator paths. Proposed approach is validated using two benchmark retinal scan image datasets, Kaggle APTOS 2019 and ISBI 2018 IDRiD, for DR type and severity grade classification tasks. Our experimental studies expose that using lesion-aware attention along with the joint training of autoencoder and neural support vector machine boosted the performance of models used for DR diagnosis, thereby outperforming existing works presented in the literature for DR severity grading. Proposed model achieved the highest accuracy of 90.45%, 84.31% on APTOS dataset and an accuracy of 79.85%, 63.24% on IDRiD dataset for DR type and severity grade classification tasks, respectively. More... »

PAGES

126

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/s00138-021-01253-y

DOI

http://dx.doi.org/10.1007/s00138-021-01253-y

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1141966636


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": "TATA Consultancy Services Limited, Hyderabad, India", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "TATA Consultancy Services Limited, Hyderabad, India"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Shaik", 
        "givenName": "Nagur Shareef", 
        "id": "sg:person.012457303645.63", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012457303645.63"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "TATA Consultancy Services Limited, Hyderabad, India", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "TATA Consultancy Services Limited, Hyderabad, India"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Cherukuri", 
        "givenName": "Teja Krishna", 
        "id": "sg:person.015203525600.42", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015203525600.42"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "sg:pub.10.1007/s12652-020-02727-z", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1134282037", 
          "https://doi.org/10.1007/s12652-020-02727-z"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s00521-018-03974-0", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1111048678", 
          "https://doi.org/10.1007/s00521-018-03974-0"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s11760-020-01816-y", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1134316732", 
          "https://doi.org/10.1007/s11760-020-01816-y"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1186/s12938-019-0675-9", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1115977905", 
          "https://doi.org/10.1186/s12938-019-0675-9"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-3-030-50402-1_9", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1128714270", 
          "https://doi.org/10.1007/978-3-030-50402-1_9"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-3-319-65981-7_12", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1092665654", 
          "https://doi.org/10.1007/978-3-319-65981-7_12"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s11042-020-10238-4", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1134390122", 
          "https://doi.org/10.1007/s11042-020-10238-4"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s11760-020-01793-2", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1131538055", 
          "https://doi.org/10.1007/s11760-020-01793-2"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "2021-10-18", 
    "datePublishedReg": "2021-10-18", 
    "description": "Diabetic retinopathy (DR) is a severe eye disease which can lead to permanent blindness. Identifying DR in early stages by using computer-aided diagnosis (CAD) systems can help the ophthalmologists to give proper treatment rationally, there by preventing many people from going blind. Due to intra-class variations and imbalanced data distribution, it is highly difficult to design a CAD system for DR severity diagnosis with greater generalizability. In this article, we propose a multi-stage deep learning pipeline, lesion-aware attention with neural support vector machine, for diabetic retinopathy diagnosis. Proposed pipeline consists of a pre-trained convolution base for learning retinal image spatial representations, lesion-aware attention for weighting lesion specific features, convolution autoencoder for learning latent attention representations and a neural support vector machine for discrimination. Convolutional autoencoder and neural support vector machine are jointly trained in end-to-end fashion to obtain category based lesion specific latent attention features by complementing each other in re-constructor and discriminator paths. Proposed approach is validated using two benchmark retinal scan image datasets, Kaggle APTOS 2019 and ISBI 2018 IDRiD, for DR type and severity grade classification tasks. Our experimental studies expose that using lesion-aware attention along with the joint training of autoencoder and neural support vector machine boosted the performance of models used for DR diagnosis, thereby outperforming existing works presented in the literature for DR severity grading. Proposed model achieved the highest accuracy of 90.45%, 84.31% on APTOS dataset and an accuracy of 79.85%, 63.24% on IDRiD dataset for DR type and severity grade classification tasks, respectively.", 
    "genre": "article", 
    "id": "sg:pub.10.1007/s00138-021-01253-y", 
    "inLanguage": "en", 
    "isAccessibleForFree": false, 
    "isPartOf": [
      {
        "id": "sg:journal.1045266", 
        "issn": [
          "0932-8092", 
          "1432-1769"
        ], 
        "name": "Machine Vision and Applications", 
        "publisher": "Springer Nature", 
        "type": "Periodical"
      }, 
      {
        "issueNumber": "6", 
        "type": "PublicationIssue"
      }, 
      {
        "type": "PublicationVolume", 
        "volumeNumber": "32"
      }
    ], 
    "keywords": [
      "diabetic retinopathy", 
      "support vector machine", 
      "DR types", 
      "severe eye disease", 
      "vector machine", 
      "lesion-specific features", 
      "classification task", 
      "eye disease", 
      "retinopathy diagnosis", 
      "severity grading", 
      "permanent blindness", 
      "computer-aided diagnosis system", 
      "DR diagnosis", 
      "deep learning pipeline", 
      "diagnosis", 
      "intra-class variations", 
      "diabetic retinopathy diagnosis", 
      "proper treatment", 
      "learning pipeline", 
      "image datasets", 
      "convolutional autoencoder", 
      "attention representation", 
      "data distribution", 
      "diagnosis system", 
      "IDRiD dataset", 
      "CAD system", 
      "end fashion", 
      "latent attention", 
      "joint training", 
      "performance of models", 
      "severity diagnosis", 
      "machine", 
      "early stages", 
      "autoencoder", 
      "retinopathy", 
      "dataset", 
      "ophthalmologists", 
      "high accuracy", 
      "disease", 
      "blindness", 
      "greater generalizability", 
      "treatment", 
      "task", 
      "grading", 
      "pipeline", 
      "representation", 
      "accuracy", 
      "spatial representation", 
      "IDRiD", 
      "study", 
      "system", 
      "generalizability", 
      "convolution", 
      "types", 
      "training", 
      "people", 
      "specific features", 
      "fashion", 
      "stage", 
      "attention", 
      "categories", 
      "model", 
      "path", 
      "literature", 
      "performance", 
      "experimental study", 
      "discrimination", 
      "features", 
      "end", 
      "work", 
      "base", 
      "article", 
      "approach", 
      "variation", 
      "distribution", 
      "DR severity diagnosis", 
      "multi-stage deep learning pipeline", 
      "lesion-aware attention", 
      "neural support vector machine", 
      "pre-trained convolution base", 
      "convolution base", 
      "retinal image spatial representations", 
      "image spatial representations", 
      "latent attention representations", 
      "lesion specific latent attention", 
      "specific latent attention", 
      "discriminator paths", 
      "benchmark retinal scan image datasets", 
      "retinal scan image datasets", 
      "scan image datasets", 
      "Kaggle APTOS 2019", 
      "APTOS 2019", 
      "ISBI 2018 IDRiD", 
      "severity grade classification tasks", 
      "grade classification tasks", 
      "DR severity grading", 
      "APTOS dataset"
    ], 
    "name": "Lesion-aware attention with neural support vector machine for retinopathy diagnosis", 
    "pagination": "126", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1141966636"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/s00138-021-01253-y"
        ]
      }
    ], 
    "sameAs": [
      "https://doi.org/10.1007/s00138-021-01253-y", 
      "https://app.dimensions.ai/details/publication/pub.1141966636"
    ], 
    "sdDataset": "articles", 
    "sdDatePublished": "2022-01-01T19:04", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/article/article_913.jsonl", 
    "type": "ScholarlyArticle", 
    "url": "https://doi.org/10.1007/s00138-021-01253-y"
  }
]
 

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/s00138-021-01253-y'

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/s00138-021-01253-y'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s00138-021-01253-y'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s00138-021-01253-y'


 

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

194 TRIPLES      22 PREDICATES      129 URIs      113 LITERALS      6 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/s00138-021-01253-y schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N87acf1c984654202aaf4738eba4279c2
4 schema:citation sg:pub.10.1007/978-3-030-50402-1_9
5 sg:pub.10.1007/978-3-319-65981-7_12
6 sg:pub.10.1007/s00521-018-03974-0
7 sg:pub.10.1007/s11042-020-10238-4
8 sg:pub.10.1007/s11760-020-01793-2
9 sg:pub.10.1007/s11760-020-01816-y
10 sg:pub.10.1007/s12652-020-02727-z
11 sg:pub.10.1186/s12938-019-0675-9
12 schema:datePublished 2021-10-18
13 schema:datePublishedReg 2021-10-18
14 schema:description Diabetic retinopathy (DR) is a severe eye disease which can lead to permanent blindness. Identifying DR in early stages by using computer-aided diagnosis (CAD) systems can help the ophthalmologists to give proper treatment rationally, there by preventing many people from going blind. Due to intra-class variations and imbalanced data distribution, it is highly difficult to design a CAD system for DR severity diagnosis with greater generalizability. In this article, we propose a multi-stage deep learning pipeline, lesion-aware attention with neural support vector machine, for diabetic retinopathy diagnosis. Proposed pipeline consists of a pre-trained convolution base for learning retinal image spatial representations, lesion-aware attention for weighting lesion specific features, convolution autoencoder for learning latent attention representations and a neural support vector machine for discrimination. Convolutional autoencoder and neural support vector machine are jointly trained in end-to-end fashion to obtain category based lesion specific latent attention features by complementing each other in re-constructor and discriminator paths. Proposed approach is validated using two benchmark retinal scan image datasets, Kaggle APTOS 2019 and ISBI 2018 IDRiD, for DR type and severity grade classification tasks. Our experimental studies expose that using lesion-aware attention along with the joint training of autoencoder and neural support vector machine boosted the performance of models used for DR diagnosis, thereby outperforming existing works presented in the literature for DR severity grading. Proposed model achieved the highest accuracy of 90.45%, 84.31% on APTOS dataset and an accuracy of 79.85%, 63.24% on IDRiD dataset for DR type and severity grade classification tasks, respectively.
15 schema:genre article
16 schema:inLanguage en
17 schema:isAccessibleForFree false
18 schema:isPartOf Na34336a5691c48a08513f96345cb2883
19 Nfea852f1ac324b59a5a8a2d7126418fb
20 sg:journal.1045266
21 schema:keywords APTOS 2019
22 APTOS dataset
23 CAD system
24 DR diagnosis
25 DR severity diagnosis
26 DR severity grading
27 DR types
28 IDRiD
29 IDRiD dataset
30 ISBI 2018 IDRiD
31 Kaggle APTOS 2019
32 accuracy
33 approach
34 article
35 attention
36 attention representation
37 autoencoder
38 base
39 benchmark retinal scan image datasets
40 blindness
41 categories
42 classification task
43 computer-aided diagnosis system
44 convolution
45 convolution base
46 convolutional autoencoder
47 data distribution
48 dataset
49 deep learning pipeline
50 diabetic retinopathy
51 diabetic retinopathy diagnosis
52 diagnosis
53 diagnosis system
54 discrimination
55 discriminator paths
56 disease
57 distribution
58 early stages
59 end
60 end fashion
61 experimental study
62 eye disease
63 fashion
64 features
65 generalizability
66 grade classification tasks
67 grading
68 greater generalizability
69 high accuracy
70 image datasets
71 image spatial representations
72 intra-class variations
73 joint training
74 latent attention
75 latent attention representations
76 learning pipeline
77 lesion specific latent attention
78 lesion-aware attention
79 lesion-specific features
80 literature
81 machine
82 model
83 multi-stage deep learning pipeline
84 neural support vector machine
85 ophthalmologists
86 path
87 people
88 performance
89 performance of models
90 permanent blindness
91 pipeline
92 pre-trained convolution base
93 proper treatment
94 representation
95 retinal image spatial representations
96 retinal scan image datasets
97 retinopathy
98 retinopathy diagnosis
99 scan image datasets
100 severe eye disease
101 severity diagnosis
102 severity grade classification tasks
103 severity grading
104 spatial representation
105 specific features
106 specific latent attention
107 stage
108 study
109 support vector machine
110 system
111 task
112 training
113 treatment
114 types
115 variation
116 vector machine
117 work
118 schema:name Lesion-aware attention with neural support vector machine for retinopathy diagnosis
119 schema:pagination 126
120 schema:productId N259a6c08a2ee471bafde3e9c61493f5f
121 Na426143180b947d3beef2c350342e29d
122 schema:sameAs https://app.dimensions.ai/details/publication/pub.1141966636
123 https://doi.org/10.1007/s00138-021-01253-y
124 schema:sdDatePublished 2022-01-01T19:04
125 schema:sdLicense https://scigraph.springernature.com/explorer/license/
126 schema:sdPublisher N87fb4b8e1fa144958c7bd177633c03a9
127 schema:url https://doi.org/10.1007/s00138-021-01253-y
128 sgo:license sg:explorer/license/
129 sgo:sdDataset articles
130 rdf:type schema:ScholarlyArticle
131 N259a6c08a2ee471bafde3e9c61493f5f schema:name doi
132 schema:value 10.1007/s00138-021-01253-y
133 rdf:type schema:PropertyValue
134 N87acf1c984654202aaf4738eba4279c2 rdf:first sg:person.012457303645.63
135 rdf:rest Needf474f6d8444b8bb579fb1fb51fb32
136 N87fb4b8e1fa144958c7bd177633c03a9 schema:name Springer Nature - SN SciGraph project
137 rdf:type schema:Organization
138 Na34336a5691c48a08513f96345cb2883 schema:volumeNumber 32
139 rdf:type schema:PublicationVolume
140 Na426143180b947d3beef2c350342e29d schema:name dimensions_id
141 schema:value pub.1141966636
142 rdf:type schema:PropertyValue
143 Needf474f6d8444b8bb579fb1fb51fb32 rdf:first sg:person.015203525600.42
144 rdf:rest rdf:nil
145 Nfea852f1ac324b59a5a8a2d7126418fb schema:issueNumber 6
146 rdf:type schema:PublicationIssue
147 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
148 schema:name Information and Computing Sciences
149 rdf:type schema:DefinedTerm
150 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
151 schema:name Artificial Intelligence and Image Processing
152 rdf:type schema:DefinedTerm
153 sg:journal.1045266 schema:issn 0932-8092
154 1432-1769
155 schema:name Machine Vision and Applications
156 schema:publisher Springer Nature
157 rdf:type schema:Periodical
158 sg:person.012457303645.63 schema:affiliation grid-institutes:None
159 schema:familyName Shaik
160 schema:givenName Nagur Shareef
161 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012457303645.63
162 rdf:type schema:Person
163 sg:person.015203525600.42 schema:affiliation grid-institutes:None
164 schema:familyName Cherukuri
165 schema:givenName Teja Krishna
166 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015203525600.42
167 rdf:type schema:Person
168 sg:pub.10.1007/978-3-030-50402-1_9 schema:sameAs https://app.dimensions.ai/details/publication/pub.1128714270
169 https://doi.org/10.1007/978-3-030-50402-1_9
170 rdf:type schema:CreativeWork
171 sg:pub.10.1007/978-3-319-65981-7_12 schema:sameAs https://app.dimensions.ai/details/publication/pub.1092665654
172 https://doi.org/10.1007/978-3-319-65981-7_12
173 rdf:type schema:CreativeWork
174 sg:pub.10.1007/s00521-018-03974-0 schema:sameAs https://app.dimensions.ai/details/publication/pub.1111048678
175 https://doi.org/10.1007/s00521-018-03974-0
176 rdf:type schema:CreativeWork
177 sg:pub.10.1007/s11042-020-10238-4 schema:sameAs https://app.dimensions.ai/details/publication/pub.1134390122
178 https://doi.org/10.1007/s11042-020-10238-4
179 rdf:type schema:CreativeWork
180 sg:pub.10.1007/s11760-020-01793-2 schema:sameAs https://app.dimensions.ai/details/publication/pub.1131538055
181 https://doi.org/10.1007/s11760-020-01793-2
182 rdf:type schema:CreativeWork
183 sg:pub.10.1007/s11760-020-01816-y schema:sameAs https://app.dimensions.ai/details/publication/pub.1134316732
184 https://doi.org/10.1007/s11760-020-01816-y
185 rdf:type schema:CreativeWork
186 sg:pub.10.1007/s12652-020-02727-z schema:sameAs https://app.dimensions.ai/details/publication/pub.1134282037
187 https://doi.org/10.1007/s12652-020-02727-z
188 rdf:type schema:CreativeWork
189 sg:pub.10.1186/s12938-019-0675-9 schema:sameAs https://app.dimensions.ai/details/publication/pub.1115977905
190 https://doi.org/10.1186/s12938-019-0675-9
191 rdf:type schema:CreativeWork
192 grid-institutes:None schema:alternateName TATA Consultancy Services Limited, Hyderabad, India
193 schema:name TATA Consultancy Services Limited, Hyderabad, India
194 rdf:type schema:Organization
 




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


...