View in EDS HTML Full Text PDF Full Text

Domain adaptation to enhance (2 + 1)D CNN dynamic analysis of cell collective behavior in time-lapse microscopy videos.

Saved in:
Bibliographic Details
Title: Domain adaptation to enhance (2 + 1)D CNN dynamic analysis of cell collective behavior in time-lapse microscopy videos.
Authors: D'Orazio, Michele1,2 (AUTHOR), Pastore, Donatella2 (AUTHOR), Mencattini, Arianna1,2 (AUTHOR) mencattini@ing.uniroma2.it, Filippi, Joanna1,2 (AUTHOR), Antonelli, Gianni1,2 (AUTHOR), Corsi, Francesca3,4 (AUTHOR), Casti, Paola1,2 (AUTHOR), Curci, Giorgia1,2 (AUTHOR), Salmeri, Marcello1 (AUTHOR), Pacifici, Francesca2 (AUTHOR), Ghibelli, Lina3 (AUTHOR), Canosci, David Della-Morte2 (AUTHOR), Martinelli, Eugenio1,2 (AUTHOR)
Source: Neural Computing & Applications. Feb2025, Vol. 37 Issue 6, p4133-4153. 21p.
Subjects: Video microscopy, Feature selection, Convolutional neural networks, Biological systems, Artificial intelligence
Abstract: In recent years, 2D CNNs have excelled in analyzing single-frame video sequences, prompting the evolution of standard architectures toward full 3D CNNs. This transition, while enhancing the modeling of spatial and temporal information in video activities, demanded substantial data for effective training. To alleviate this challenge, we introduced a switched multitask training strategy. This approach involves factorizing 3D layers into (2 + 1)D convolutions, training only 2D (1D) layers for spatial (temporal) tasks and switching off the training of the 1D (2D) layers. Additionally, we addressed data scarcity by generating synthetic stylized video sequences. These were crafted using stochastic particle models of collective cell motions, further modified through neural style transfer to mimic real video data. Such a domain adaptation strategy facilitated the creation of training data impractical to obtain in the real world. Transferring knowledge from the switched (2 + 1) CNN to real video data, we encoded wound healing experiments of three distinct cell lines—human melanocytes cells M14, mouse neuroblastoma cells N1, and human prostate cells PC3—into deep features. Employing a novel feature selection strategy based on robustness to disturbances, we discriminated the three wound healing processes. Average classification accuracy of 92.91% (0.11%), 91.50% (0.37%), and 88.81% (0.29%) was obtained for the original real videos, the real videos with progressive altered levels of focus, and levels of brightness, respectively. The proposed approach proved to be a powerful tool for analyzing the spatiotemporal dynamics of biological systems, even in the presence of fluctuations. [ABSTRACT FROM AUTHOR]
Copyright of Neural Computing & Applications is the property of Springer Nature 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
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: egs
DbLabel: Engineering Source
An: 182882378
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Domain adaptation to enhance (2 + 1)D CNN dynamic analysis of cell collective behavior in time-lapse microscopy videos.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22D'Orazio%2C+Michele%22">D'Orazio, Michele</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Pastore%2C+Donatella%22">Pastore, Donatella</searchLink><relatesTo>2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Mencattini%2C+Arianna%22">Mencattini, Arianna</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<i> mencattini@ing.uniroma2.it</i><br /><searchLink fieldCode="AR" term="%22Filippi%2C+Joanna%22">Filippi, Joanna</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Antonelli%2C+Gianni%22">Antonelli, Gianni</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Corsi%2C+Francesca%22">Corsi, Francesca</searchLink><relatesTo>3,4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Casti%2C+Paola%22">Casti, Paola</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Curci%2C+Giorgia%22">Curci, Giorgia</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Salmeri%2C+Marcello%22">Salmeri, Marcello</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Pacifici%2C+Francesca%22">Pacifici, Francesca</searchLink><relatesTo>2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Ghibelli%2C+Lina%22">Ghibelli, Lina</searchLink><relatesTo>3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Canosci%2C+David+Della-Morte%22">Canosci, David Della-Morte</searchLink><relatesTo>2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Martinelli%2C+Eugenio%22">Martinelli, Eugenio</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Neural+Computing+%26+Applications%22">Neural Computing & Applications</searchLink>. Feb2025, Vol. 37 Issue 6, p4133-4153. 21p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Video+microscopy%22">Video microscopy</searchLink><br /><searchLink fieldCode="DE" term="%22Feature+selection%22">Feature selection</searchLink><br /><searchLink fieldCode="DE" term="%22Convolutional+neural+networks%22">Convolutional neural networks</searchLink><br /><searchLink fieldCode="DE" term="%22Biological+systems%22">Biological systems</searchLink><br /><searchLink fieldCode="DE" term="%22Artificial+intelligence%22">Artificial intelligence</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: In recent years, 2D CNNs have excelled in analyzing single-frame video sequences, prompting the evolution of standard architectures toward full 3D CNNs. This transition, while enhancing the modeling of spatial and temporal information in video activities, demanded substantial data for effective training. To alleviate this challenge, we introduced a switched multitask training strategy. This approach involves factorizing 3D layers into (2 + 1)D convolutions, training only 2D (1D) layers for spatial (temporal) tasks and switching off the training of the 1D (2D) layers. Additionally, we addressed data scarcity by generating synthetic stylized video sequences. These were crafted using stochastic particle models of collective cell motions, further modified through neural style transfer to mimic real video data. Such a domain adaptation strategy facilitated the creation of training data impractical to obtain in the real world. Transferring knowledge from the switched (2 + 1) CNN to real video data, we encoded wound healing experiments of three distinct cell lines—human melanocytes cells M14, mouse neuroblastoma cells N1, and human prostate cells PC3—into deep features. Employing a novel feature selection strategy based on robustness to disturbances, we discriminated the three wound healing processes. Average classification accuracy of 92.91% (0.11%), 91.50% (0.37%), and 88.81% (0.29%) was obtained for the original real videos, the real videos with progressive altered levels of focus, and levels of brightness, respectively. The proposed approach proved to be a powerful tool for analyzing the spatiotemporal dynamics of biological systems, even in the presence of fluctuations. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Neural Computing & Applications is the property of Springer Nature 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=182882378
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1007/s00521-024-10767-1
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 21
        StartPage: 4133
    Subjects:
      – SubjectFull: Video microscopy
        Type: general
      – SubjectFull: Feature selection
        Type: general
      – SubjectFull: Convolutional neural networks
        Type: general
      – SubjectFull: Biological systems
        Type: general
      – SubjectFull: Artificial intelligence
        Type: general
    Titles:
      – TitleFull: Domain adaptation to enhance (2 + 1)D CNN dynamic analysis of cell collective behavior in time-lapse microscopy videos.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: D'Orazio, Michele
      – PersonEntity:
          Name:
            NameFull: Pastore, Donatella
      – PersonEntity:
          Name:
            NameFull: Mencattini, Arianna
      – PersonEntity:
          Name:
            NameFull: Filippi, Joanna
      – PersonEntity:
          Name:
            NameFull: Antonelli, Gianni
      – PersonEntity:
          Name:
            NameFull: Corsi, Francesca
      – PersonEntity:
          Name:
            NameFull: Casti, Paola
      – PersonEntity:
          Name:
            NameFull: Curci, Giorgia
      – PersonEntity:
          Name:
            NameFull: Salmeri, Marcello
      – PersonEntity:
          Name:
            NameFull: Pacifici, Francesca
      – PersonEntity:
          Name:
            NameFull: Ghibelli, Lina
      – PersonEntity:
          Name:
            NameFull: Canosci, David Della-Morte
      – PersonEntity:
          Name:
            NameFull: Martinelli, Eugenio
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 21
              M: 02
              Text: Feb2025
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 09410643
          Numbering:
            – Type: volume
              Value: 37
            – Type: issue
              Value: 6
          Titles:
            – TitleFull: Neural Computing & Applications
              Type: main
ResultId 1