Academic Journal
View in EDS
PDF Full Text

Identification of veterinary and medically important blood parasites using contrastive loss-based self-supervised learning

Bibliographic Details
Title: Identification of veterinary and medically important blood parasites using contrastive loss-based self-supervised learning
Authors: Supasuta Busayakanon, Morakot Kaewthamasorn, Natchapon Pinetsuksai, Teerawat Tongloy, Santhad Chuwongin, Siridech Boonsang, Veerayuth Kittichai
Source: Veterinary World, Vol 17, Iss 11, Pp 2619-2634 (2024)
Publisher Information: Veterinary World, 2024.
Publication Year: 2024
Collection: LCC:Animal culture
LCC:Veterinary medicine
Subject Terms: bootstrap your own latent, fractioned data, microscopic image, pre-trained, self-supervised learning, zoonotic disease, Animal culture, SF1-1100, Veterinary medicine, SF600-1100
More Details: Background and Aim: Zoonotic diseases caused by various blood parasites are important public health concerns that impact animals and humans worldwide. The traditional method of microscopic examination for parasite diagnosis is labor-intensive, time-consuming, and prone to variability among observers, necessitating highly skilled and experienced personnel. Therefore, an innovative approach is required to enhance the conventional method. This study aimed to develop a self-supervised learning (SSL) approach to identify zoonotic blood parasites from microscopic images, with an initial focus on parasite species classification. Materials and Methods: We acquired a public dataset featuring microscopic images of Giemsa-stained thin blood films of trypanosomes and other blood parasites, including Babesia, Leishmania, Plasmodium, Toxoplasma, and Trichomonad, as well as images of both white and red blood cells. The input data were subjected to SSL model training using the Bootstrap Your Own Latent (BYOL) algorithm with Residual Network 50 (ResNet50), ResNet101, and ResNet152 as the backbones. The performance of the proposed SSL model was then compared to that of baseline models. Results: The proposed BYOL SSL model outperformed supervised learning models across all classes. Among the SSL models, ResNet50 consistently achieved high accuracy, reaching 0.992 in most classes, which aligns well with the patterns observed in the pre-trained uniform manifold approximation and projection representations. Fine-tuned SSL models exhibit high performance, achieving 95% accuracy and a 0.960 area under the curve of the receiver operating characteristics (ROC) curve even when fine-tuned with 1% of the data in the downstream process. Furthermore, 20% of the data for training with SSL models yielded ≥95% in all other statistical metrics, including accuracy, recall, precision, specification, F1 score, and ROC curve. As a result, multi-class classification prediction demonstrated that model performance exceeded 91% for the F1 score, except for the early stage of Trypanosoma evansi, which showed an F1 score of 87%. This may be due to the model being exposed to high levels of variation during the developmental stage. Conclusion: This approach can significantly enhance active surveillance efforts to improve disease control and prevent outbreaks, particularly in resource-limited settings. In addition, SSL addresses significant challenges, such as data variability and the requirement for extensive class labeling, which are common in biology and medical fields.
Document Type: article
File Description: electronic resource
Language: English
ISSN: 0972-8988
Relation: https://www.veterinaryworld.org/Vol.17/November-2024/22.pdf; https://doaj.org/toc/0972-8988
DOI: 10.14202/vetworld.2024.2619-2634
Access URL: https://doaj.org/article/3b04860f8eeb494c8d8ec6296e8050c3
Accession Number: edsdoj.3b04860f8eeb494c8d8ec6296e8050c3
Database: Directory of Open Access Journals
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHjPtM4BHU3ZchRwgzYmadcigk49r9CVlbU7V5F6lgH7WwH3DcMOk2bOje2pTs4xQJJoAAAA4TCB3gYJKoZIhvcNAQcGoIHQMIHNAgEAMIHHBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDI4bTmsRY7dYg3FqFQIBEICBmTMI-i_vkD6IWQPNtQTdKPH8CuQ_Di6V0FtpzB3MlZHNoFEmP1CBHRUpaZyPrmC3cawGmwnnmEAWq41_mH7H0OA-gI_yu3Petp0v-OvVJqQbNDypDn1shmA_dJWHYRsOMWAKojNp21oF-0hKU18WiM1av89q9EK1LdPBA1Lps4aglUt0gRZr4UsK-SWIRFYccQ7fr6r9GmiebQ==
Text:
  Availability: 0
CustomLinks:
  – Url: https://resolver.ebsco.com/c/xy5jbn/result?sid=EBSCO:edsdoj&genre=article&issn=09728988&ISBN=&volume=17&issue=11&date=20241101&spage=2619&pages=2619-2634&title=Veterinary World&atitle=Identification%20of%20veterinary%20and%20medically%20important%20blood%20parasites%20using%20contrastive%20loss-based%20self-supervised%20learning&aulast=Supasuta%20Busayakanon&id=DOI:10.14202/vetworld.2024.2619-2634
    Name: Full Text Finder (for New FTF UI) (s8985755)
    Category: fullText
    Text: Find It @ SCU Libraries
    MouseOverText: Find It @ SCU Libraries
  – Url: https://doaj.org/article/3b04860f8eeb494c8d8ec6296e8050c3
    Name: EDS - DOAJ (s8985755)
    Category: fullText
    Text: View record from DOAJ
    MouseOverText: View record from DOAJ
Header DbId: edsdoj
DbLabel: Directory of Open Access Journals
An: edsdoj.3b04860f8eeb494c8d8ec6296e8050c3
RelevancyScore: 1047
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 1047.28198242188
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Identification of veterinary and medically important blood parasites using contrastive loss-based self-supervised learning
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Supasuta+Busayakanon%22">Supasuta Busayakanon</searchLink><br /><searchLink fieldCode="AR" term="%22Morakot+Kaewthamasorn%22">Morakot Kaewthamasorn</searchLink><br /><searchLink fieldCode="AR" term="%22Natchapon+Pinetsuksai%22">Natchapon Pinetsuksai</searchLink><br /><searchLink fieldCode="AR" term="%22Teerawat+Tongloy%22">Teerawat Tongloy</searchLink><br /><searchLink fieldCode="AR" term="%22Santhad+Chuwongin%22">Santhad Chuwongin</searchLink><br /><searchLink fieldCode="AR" term="%22Siridech+Boonsang%22">Siridech Boonsang</searchLink><br /><searchLink fieldCode="AR" term="%22Veerayuth+Kittichai%22">Veerayuth Kittichai</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: Veterinary World, Vol 17, Iss 11, Pp 2619-2634 (2024)
– Name: Publisher
  Label: Publisher Information
  Group: PubInfo
  Data: Veterinary World, 2024.
– Name: DatePubCY
  Label: Publication Year
  Group: Date
  Data: 2024
– Name: Subset
  Label: Collection
  Group: HoldingsInfo
  Data: LCC:Animal culture<br />LCC:Veterinary medicine
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22bootstrap+your+own+latent%22">bootstrap your own latent</searchLink><br /><searchLink fieldCode="DE" term="%22fractioned+data%22">fractioned data</searchLink><br /><searchLink fieldCode="DE" term="%22microscopic+image%22">microscopic image</searchLink><br /><searchLink fieldCode="DE" term="%22pre-trained%22">pre-trained</searchLink><br /><searchLink fieldCode="DE" term="%22self-supervised+learning%22">self-supervised learning</searchLink><br /><searchLink fieldCode="DE" term="%22zoonotic+disease%22">zoonotic disease</searchLink><br /><searchLink fieldCode="DE" term="%22Animal+culture%22">Animal culture</searchLink><br /><searchLink fieldCode="DE" term="%22SF1-1100%22">SF1-1100</searchLink><br /><searchLink fieldCode="DE" term="%22Veterinary+medicine%22">Veterinary medicine</searchLink><br /><searchLink fieldCode="DE" term="%22SF600-1100%22">SF600-1100</searchLink>
– Name: Abstract
  Label: Description
  Group: Ab
  Data: Background and Aim: Zoonotic diseases caused by various blood parasites are important public health concerns that impact animals and humans worldwide. The traditional method of microscopic examination for parasite diagnosis is labor-intensive, time-consuming, and prone to variability among observers, necessitating highly skilled and experienced personnel. Therefore, an innovative approach is required to enhance the conventional method. This study aimed to develop a self-supervised learning (SSL) approach to identify zoonotic blood parasites from microscopic images, with an initial focus on parasite species classification. Materials and Methods: We acquired a public dataset featuring microscopic images of Giemsa-stained thin blood films of trypanosomes and other blood parasites, including Babesia, Leishmania, Plasmodium, Toxoplasma, and Trichomonad, as well as images of both white and red blood cells. The input data were subjected to SSL model training using the Bootstrap Your Own Latent (BYOL) algorithm with Residual Network 50 (ResNet50), ResNet101, and ResNet152 as the backbones. The performance of the proposed SSL model was then compared to that of baseline models. Results: The proposed BYOL SSL model outperformed supervised learning models across all classes. Among the SSL models, ResNet50 consistently achieved high accuracy, reaching 0.992 in most classes, which aligns well with the patterns observed in the pre-trained uniform manifold approximation and projection representations. Fine-tuned SSL models exhibit high performance, achieving 95% accuracy and a 0.960 area under the curve of the receiver operating characteristics (ROC) curve even when fine-tuned with 1% of the data in the downstream process. Furthermore, 20% of the data for training with SSL models yielded ≥95% in all other statistical metrics, including accuracy, recall, precision, specification, F1 score, and ROC curve. As a result, multi-class classification prediction demonstrated that model performance exceeded 91% for the F1 score, except for the early stage of Trypanosoma evansi, which showed an F1 score of 87%. This may be due to the model being exposed to high levels of variation during the developmental stage. Conclusion: This approach can significantly enhance active surveillance efforts to improve disease control and prevent outbreaks, particularly in resource-limited settings. In addition, SSL addresses significant challenges, such as data variability and the requirement for extensive class labeling, which are common in biology and medical fields.
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: article
– Name: Format
  Label: File Description
  Group: SrcInfo
  Data: electronic resource
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0972-8988
– Name: NoteTitleSource
  Label: Relation
  Group: SrcInfo
  Data: https://www.veterinaryworld.org/Vol.17/November-2024/22.pdf; https://doaj.org/toc/0972-8988
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.14202/vetworld.2024.2619-2634
– Name: URL
  Label: Access URL
  Group: URL
  Data: <link linkTarget="URL" linkTerm="https://doaj.org/article/3b04860f8eeb494c8d8ec6296e8050c3" linkWindow="_blank">https://doaj.org/article/3b04860f8eeb494c8d8ec6296e8050c3</link>
– Name: AN
  Label: Accession Number
  Group: ID
  Data: edsdoj.3b04860f8eeb494c8d8ec6296e8050c3
PLink https://login.libproxy.scu.edu/login?url=https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&scope=site&db=edsdoj&AN=edsdoj.3b04860f8eeb494c8d8ec6296e8050c3
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.14202/vetworld.2024.2619-2634
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 16
        StartPage: 2619
    Subjects:
      – SubjectFull: bootstrap your own latent
        Type: general
      – SubjectFull: fractioned data
        Type: general
      – SubjectFull: microscopic image
        Type: general
      – SubjectFull: pre-trained
        Type: general
      – SubjectFull: self-supervised learning
        Type: general
      – SubjectFull: zoonotic disease
        Type: general
      – SubjectFull: Animal culture
        Type: general
      – SubjectFull: SF1-1100
        Type: general
      – SubjectFull: Veterinary medicine
        Type: general
      – SubjectFull: SF600-1100
        Type: general
    Titles:
      – TitleFull: Identification of veterinary and medically important blood parasites using contrastive loss-based self-supervised learning
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Supasuta Busayakanon
      – PersonEntity:
          Name:
            NameFull: Morakot Kaewthamasorn
      – PersonEntity:
          Name:
            NameFull: Natchapon Pinetsuksai
      – PersonEntity:
          Name:
            NameFull: Teerawat Tongloy
      – PersonEntity:
          Name:
            NameFull: Santhad Chuwongin
      – PersonEntity:
          Name:
            NameFull: Siridech Boonsang
      – PersonEntity:
          Name:
            NameFull: Veerayuth Kittichai
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 11
              Type: published
              Y: 2024
          Identifiers:
            – Type: issn-print
              Value: 09728988
          Numbering:
            – Type: volume
              Value: 17
            – Type: issue
              Value: 11
          Titles:
            – TitleFull: Veterinary World
              Type: main
ResultId 1