View in EDS

Interpretability-Driven Sample Selection Using Self Supervised Learning for Disease Classification and Segmentation.

Saved in:
Bibliographic Details
Title: Interpretability-Driven Sample Selection Using Self Supervised Learning for Disease Classification and Segmentation.
Authors: Mahapatra, Dwarikanath1 (AUTHOR) dwarikanath.mahapatra@inceptioniai.org, Poellinger, Alexander2 (AUTHOR) alexander.poellinger@insel.ch, Shao, Ling1 (AUTHOR) ling.shao@inceptioniai.org, Reyes, Mauricio3 (AUTHOR) mauricio.reyes@med.unibe.ch
Source: IEEE Transactions on Medical Imaging. Oct2021, Vol. 40 Issue 10, p2548-2562. 15p.
Subjects: National Institutes of Health (U.S.), Nosology, Active learning, Supervised learning, Image segmentation, Image analysis, Self, Lung diseases
Abstract: In supervised learning for medical image analysis, sample selection methodologies are fundamental to attain optimum system performance promptly and with minimal expert interactions (e.g. label querying in an active learning setup). In this article we propose a novel sample selection methodology based on deep features leveraging information contained in interpretability saliency maps. In the absence of ground truth labels for informative samples, we use a novel self supervised learning based approach for training a classifier that learns to identify the most informative sample in a given batch of images. We demonstrate the benefits of the proposed approach, termed Interpretability-Driven Sample Selection (IDEAL), in an active learning setup aimed at lung disease classification and histopathology image segmentation. We analyze three different approaches to determine sample informativeness from interpretability saliency maps: (i) an observational model stemming from findings on previous uncertainty-based sample selection approaches, (ii) a radiomics-based model, and (iii) a novel data-driven self-supervised approach. We compare IDEAL to other baselines using the publicly available NIH chest X-ray dataset for lung disease classification, and a public histopathology segmentation dataset (GLaS), demonstrating the potential of using interpretability information for sample selection in active learning systems. Results show our proposed self supervised approach outperforms other approaches in selecting informative samples leading to state of the art performance with fewer samples. [ABSTRACT FROM AUTHOR]
Copyright of IEEE Transactions on Medical Imaging is the property of IEEE and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
FullText Text:
  Availability: 0
Header DbId: egs
DbLabel: Engineering Source
An: 153710572
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Interpretability-Driven Sample Selection Using Self Supervised Learning for Disease Classification and Segmentation.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Mahapatra%2C+Dwarikanath%22">Mahapatra, Dwarikanath</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> dwarikanath.mahapatra@inceptioniai.org</i><br /><searchLink fieldCode="AR" term="%22Poellinger%2C+Alexander%22">Poellinger, Alexander</searchLink><relatesTo>2</relatesTo> (AUTHOR)<i> alexander.poellinger@insel.ch</i><br /><searchLink fieldCode="AR" term="%22Shao%2C+Ling%22">Shao, Ling</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> ling.shao@inceptioniai.org</i><br /><searchLink fieldCode="AR" term="%22Reyes%2C+Mauricio%22">Reyes, Mauricio</searchLink><relatesTo>3</relatesTo> (AUTHOR)<i> mauricio.reyes@med.unibe.ch</i>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22IEEE+Transactions+on+Medical+Imaging%22">IEEE Transactions on Medical Imaging</searchLink>. Oct2021, Vol. 40 Issue 10, p2548-2562. 15p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22National+Institutes+of+Health+%28U%2ES%2E%29%22">National Institutes of Health (U.S.)</searchLink><br /><searchLink fieldCode="DE" term="%22Nosology%22">Nosology</searchLink><br /><searchLink fieldCode="DE" term="%22Active+learning%22">Active learning</searchLink><br /><searchLink fieldCode="DE" term="%22Supervised+learning%22">Supervised learning</searchLink><br /><searchLink fieldCode="DE" term="%22Image+segmentation%22">Image segmentation</searchLink><br /><searchLink fieldCode="DE" term="%22Image+analysis%22">Image analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Self%22">Self</searchLink><br /><searchLink fieldCode="DE" term="%22Lung+diseases%22">Lung diseases</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: In supervised learning for medical image analysis, sample selection methodologies are fundamental to attain optimum system performance promptly and with minimal expert interactions (e.g. label querying in an active learning setup). In this article we propose a novel sample selection methodology based on deep features leveraging information contained in interpretability saliency maps. In the absence of ground truth labels for informative samples, we use a novel self supervised learning based approach for training a classifier that learns to identify the most informative sample in a given batch of images. We demonstrate the benefits of the proposed approach, termed Interpretability-Driven Sample Selection (IDEAL), in an active learning setup aimed at lung disease classification and histopathology image segmentation. We analyze three different approaches to determine sample informativeness from interpretability saliency maps: (i) an observational model stemming from findings on previous uncertainty-based sample selection approaches, (ii) a radiomics-based model, and (iii) a novel data-driven self-supervised approach. We compare IDEAL to other baselines using the publicly available NIH chest X-ray dataset for lung disease classification, and a public histopathology segmentation dataset (GLaS), demonstrating the potential of using interpretability information for sample selection in active learning systems. Results show our proposed self supervised approach outperforms other approaches in selecting informative samples leading to state of the art performance with fewer samples. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of IEEE Transactions on Medical Imaging is the property of IEEE and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=153710572
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1109/TMI.2021.3061724
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 15
        StartPage: 2548
    Subjects:
      – SubjectFull: National Institutes of Health (U.S.)
        Type: general
      – SubjectFull: Nosology
        Type: general
      – SubjectFull: Active learning
        Type: general
      – SubjectFull: Supervised learning
        Type: general
      – SubjectFull: Image segmentation
        Type: general
      – SubjectFull: Image analysis
        Type: general
      – SubjectFull: Self
        Type: general
      – SubjectFull: Lung diseases
        Type: general
    Titles:
      – TitleFull: Interpretability-Driven Sample Selection Using Self Supervised Learning for Disease Classification and Segmentation.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Mahapatra, Dwarikanath
      – PersonEntity:
          Name:
            NameFull: Poellinger, Alexander
      – PersonEntity:
          Name:
            NameFull: Shao, Ling
      – PersonEntity:
          Name:
            NameFull: Reyes, Mauricio
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 10
              Text: Oct2021
              Type: published
              Y: 2021
          Identifiers:
            – Type: issn-print
              Value: 02780062
          Numbering:
            – Type: volume
              Value: 40
            – Type: issue
              Value: 10
          Titles:
            – TitleFull: IEEE Transactions on Medical Imaging
              Type: main
ResultId 1