Semi-supervised Multi-task Learning with Chest X-Ray Images View Full Text


Ontology type: schema:Chapter      Open Access: True


Chapter Info

DATE

2019-10-10

AUTHORS

Abdullah-Al-Zubaer Imran , Demetri Terzopoulos

ABSTRACT

Discriminative models that require full supervision are inefficacious in the medical imaging domain when large labeled datasets are unavailable. By contrast, generative modeling—i.e., learning data generation and classification—facilitates semi-supervised training with limited labeled data. Moreover, generative modeling can be advantageous in accomplishing multiple objectives for better generalization. We propose a novel multi-task learning model for jointly learning a classifier and a segmentor, from chest X-ray images, through semi-supervised learning. In addition, we propose a new loss function that combines absolute KL divergence with Tversky loss (KLTV) to yield faster convergence and better segmentation performance. Based on our experimental results using a novel segmentation model, an Adversarial Pyramid Progressive Attention U-Net (APPAU-Net), we hypothesize that KLTV can be more effective for generalizing multi-tasking models while being competitive in segmentation-only tasks. More... »

PAGES

151-159

Book

TITLE

Machine Learning in Medical Imaging

ISBN

978-3-030-32691-3
978-3-030-32692-0

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-030-32692-0_18

DOI

http://dx.doi.org/10.1007/978-3-030-32692-0_18

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1121614309


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": "Computer Science Department, University of California, Los Angeles, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.19006.3e", 
          "name": [
            "Computer Science Department, University of California, Los Angeles, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Imran", 
        "givenName": "Abdullah-Al-Zubaer", 
        "id": "sg:person.013214275461.49", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013214275461.49"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Computer Science Department, University of California, Los Angeles, CA, USA", 
          "id": "http://www.grid.ac/institutes/grid.19006.3e", 
          "name": [
            "Computer Science Department, University of California, Los Angeles, CA, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Terzopoulos", 
        "givenName": "Demetri", 
        "id": "sg:person.016347323445.35", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.016347323445.35"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2019-10-10", 
    "datePublishedReg": "2019-10-10", 
    "description": "Discriminative models that require full supervision are inefficacious in the medical imaging domain when large labeled datasets are unavailable. By contrast, generative modeling\u2014i.e., learning data generation and classification\u2014facilitates semi-supervised training with limited labeled data. Moreover, generative modeling can be advantageous in accomplishing multiple objectives for better generalization. We propose a novel multi-task learning model for jointly learning a classifier and a segmentor, from chest X-ray images, through semi-supervised learning. In addition, we propose a new loss function that combines absolute KL divergence with Tversky loss (KLTV) to yield faster convergence and better segmentation performance. Based on our experimental results using a novel segmentation model, an Adversarial Pyramid Progressive Attention U-Net (APPAU-Net), we hypothesize that KLTV can be more effective for generalizing multi-tasking models while being competitive in segmentation-only tasks.", 
    "editor": [
      {
        "familyName": "Suk", 
        "givenName": "Heung-Il", 
        "type": "Person"
      }, 
      {
        "familyName": "Liu", 
        "givenName": "Mingxia", 
        "type": "Person"
      }, 
      {
        "familyName": "Yan", 
        "givenName": "Pingkun", 
        "type": "Person"
      }, 
      {
        "familyName": "Lian", 
        "givenName": "Chunfeng", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-030-32692-0_18", 
    "inLanguage": "en", 
    "isAccessibleForFree": true, 
    "isPartOf": {
      "isbn": [
        "978-3-030-32691-3", 
        "978-3-030-32692-0"
      ], 
      "name": "Machine Learning in Medical Imaging", 
      "type": "Book"
    }, 
    "keywords": [
      "generative modeling", 
      "novel multi-task learning model", 
      "multi-task learning model", 
      "medical imaging domain", 
      "better segmentation performance", 
      "ray images", 
      "novel segmentation model", 
      "semi-supervised training", 
      "semi-supervised learning", 
      "new loss function", 
      "Multi-Task", 
      "imaging domain", 
      "segmentation performance", 
      "attention U", 
      "Tversky loss", 
      "segmentation model", 
      "better generalization", 
      "data generation", 
      "learning model", 
      "full supervision", 
      "loss function", 
      "KL divergence", 
      "multiple objectives", 
      "fast convergence", 
      "experimental results", 
      "images", 
      "classifier", 
      "segmentor", 
      "segmentation", 
      "dataset", 
      "task", 
      "learning", 
      "modeling", 
      "model", 
      "nets", 
      "performance", 
      "domain", 
      "supervision", 
      "convergence", 
      "chest X", 
      "training", 
      "generalization", 
      "data", 
      "generation", 
      "objective", 
      "results", 
      "function", 
      "addition", 
      "loss", 
      "divergence", 
      "contrast", 
      "classification\u2014facilitates semi-supervised training", 
      "absolute KL divergence", 
      "Adversarial Pyramid Progressive Attention U", 
      "Pyramid Progressive Attention U", 
      "Progressive Attention U", 
      "KLTV", 
      "multi-tasking models", 
      "Semi-supervised Multi-task"
    ], 
    "name": "Semi-supervised Multi-task Learning with Chest X-Ray Images", 
    "pagination": "151-159", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1121614309"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-030-32692-0_18"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-030-32692-0_18", 
      "https://app.dimensions.ai/details/publication/pub.1121614309"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-01-01T19:08", 
    "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/chapter/chapter_139.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-030-32692-0_18"
  }
]
 

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.10.1007/978-3-030-32692-0_18'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-32692-0_18'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-32692-0_18'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-32692-0_18'


 

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

141 TRIPLES      23 PREDICATES      84 URIs      77 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-030-32692-0_18 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N967d37bfcba84e2d839678b02aaa2a6a
4 schema:datePublished 2019-10-10
5 schema:datePublishedReg 2019-10-10
6 schema:description Discriminative models that require full supervision are inefficacious in the medical imaging domain when large labeled datasets are unavailable. By contrast, generative modeling—i.e., learning data generation and classification—facilitates semi-supervised training with limited labeled data. Moreover, generative modeling can be advantageous in accomplishing multiple objectives for better generalization. We propose a novel multi-task learning model for jointly learning a classifier and a segmentor, from chest X-ray images, through semi-supervised learning. In addition, we propose a new loss function that combines absolute KL divergence with Tversky loss (KLTV) to yield faster convergence and better segmentation performance. Based on our experimental results using a novel segmentation model, an Adversarial Pyramid Progressive Attention U-Net (APPAU-Net), we hypothesize that KLTV can be more effective for generalizing multi-tasking models while being competitive in segmentation-only tasks.
7 schema:editor Nbce3cd13584a4fc68d037d41b537454c
8 schema:genre chapter
9 schema:inLanguage en
10 schema:isAccessibleForFree true
11 schema:isPartOf N6a2e4b900c194872a2ac3b08da46b409
12 schema:keywords Adversarial Pyramid Progressive Attention U
13 KL divergence
14 KLTV
15 Multi-Task
16 Progressive Attention U
17 Pyramid Progressive Attention U
18 Semi-supervised Multi-task
19 Tversky loss
20 absolute KL divergence
21 addition
22 attention U
23 better generalization
24 better segmentation performance
25 chest X
26 classification—facilitates semi-supervised training
27 classifier
28 contrast
29 convergence
30 data
31 data generation
32 dataset
33 divergence
34 domain
35 experimental results
36 fast convergence
37 full supervision
38 function
39 generalization
40 generation
41 generative modeling
42 images
43 imaging domain
44 learning
45 learning model
46 loss
47 loss function
48 medical imaging domain
49 model
50 modeling
51 multi-task learning model
52 multi-tasking models
53 multiple objectives
54 nets
55 new loss function
56 novel multi-task learning model
57 novel segmentation model
58 objective
59 performance
60 ray images
61 results
62 segmentation
63 segmentation model
64 segmentation performance
65 segmentor
66 semi-supervised learning
67 semi-supervised training
68 supervision
69 task
70 training
71 schema:name Semi-supervised Multi-task Learning with Chest X-Ray Images
72 schema:pagination 151-159
73 schema:productId N5a08e4392d9a4335bf07853ee3014064
74 N75365ec8a5b64bcea25bf87ece1b5fa3
75 schema:publisher Nbf9b5bdc4d38439fa48038f02f730332
76 schema:sameAs https://app.dimensions.ai/details/publication/pub.1121614309
77 https://doi.org/10.1007/978-3-030-32692-0_18
78 schema:sdDatePublished 2022-01-01T19:08
79 schema:sdLicense https://scigraph.springernature.com/explorer/license/
80 schema:sdPublisher Nc0c04868752147f68a76ca3420812e6c
81 schema:url https://doi.org/10.1007/978-3-030-32692-0_18
82 sgo:license sg:explorer/license/
83 sgo:sdDataset chapters
84 rdf:type schema:Chapter
85 N3bfb35a3dc9243aa8f4cfbec0aad9f4b schema:familyName Lian
86 schema:givenName Chunfeng
87 rdf:type schema:Person
88 N426d474556ba4cdfbcf9b4ee58c09412 rdf:first Nba28df2d255248cc9656443b815ab868
89 rdf:rest N45df325a02414dd69115781bf9fd2e90
90 N45df325a02414dd69115781bf9fd2e90 rdf:first N3bfb35a3dc9243aa8f4cfbec0aad9f4b
91 rdf:rest rdf:nil
92 N5a08e4392d9a4335bf07853ee3014064 schema:name doi
93 schema:value 10.1007/978-3-030-32692-0_18
94 rdf:type schema:PropertyValue
95 N6a2e4b900c194872a2ac3b08da46b409 schema:isbn 978-3-030-32691-3
96 978-3-030-32692-0
97 schema:name Machine Learning in Medical Imaging
98 rdf:type schema:Book
99 N75365ec8a5b64bcea25bf87ece1b5fa3 schema:name dimensions_id
100 schema:value pub.1121614309
101 rdf:type schema:PropertyValue
102 N80d66bc4df1940e3b7bbc3475868ced4 schema:familyName Suk
103 schema:givenName Heung-Il
104 rdf:type schema:Person
105 N967d37bfcba84e2d839678b02aaa2a6a rdf:first sg:person.013214275461.49
106 rdf:rest Nb3185c51e5364b628553a4841dba1c35
107 Nb3185c51e5364b628553a4841dba1c35 rdf:first sg:person.016347323445.35
108 rdf:rest rdf:nil
109 Nb38b24e2532e48a2a1b29600fa06ff3e rdf:first Ndd0448a8736f4ce38f20b2a30d66fc2b
110 rdf:rest N426d474556ba4cdfbcf9b4ee58c09412
111 Nba28df2d255248cc9656443b815ab868 schema:familyName Yan
112 schema:givenName Pingkun
113 rdf:type schema:Person
114 Nbce3cd13584a4fc68d037d41b537454c rdf:first N80d66bc4df1940e3b7bbc3475868ced4
115 rdf:rest Nb38b24e2532e48a2a1b29600fa06ff3e
116 Nbf9b5bdc4d38439fa48038f02f730332 schema:name Springer Nature
117 rdf:type schema:Organisation
118 Nc0c04868752147f68a76ca3420812e6c schema:name Springer Nature - SN SciGraph project
119 rdf:type schema:Organization
120 Ndd0448a8736f4ce38f20b2a30d66fc2b schema:familyName Liu
121 schema:givenName Mingxia
122 rdf:type schema:Person
123 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
124 schema:name Information and Computing Sciences
125 rdf:type schema:DefinedTerm
126 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
127 schema:name Artificial Intelligence and Image Processing
128 rdf:type schema:DefinedTerm
129 sg:person.013214275461.49 schema:affiliation grid-institutes:grid.19006.3e
130 schema:familyName Imran
131 schema:givenName Abdullah-Al-Zubaer
132 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013214275461.49
133 rdf:type schema:Person
134 sg:person.016347323445.35 schema:affiliation grid-institutes:grid.19006.3e
135 schema:familyName Terzopoulos
136 schema:givenName Demetri
137 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.016347323445.35
138 rdf:type schema:Person
139 grid-institutes:grid.19006.3e schema:alternateName Computer Science Department, University of California, Los Angeles, CA, USA
140 schema:name Computer Science Department, University of California, Los Angeles, CA, USA
141 rdf:type schema:Organization
 




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


...