Childhood Cancer Data Initiative Annual Symposium (Abstract Registration): Submission #60

serial: '60'
sid: '150493'
uuid: 9b5a2d3c-8cd5-4e2f-a64d-0a3db5ad1be8
uri: /nci/ccdisymposium/abstract
created: '1756775621'
completed: '1756777562'
changed: '1756777562'
in_draft: '0'
current_page: ''
remote_addr: 10.208.28.30
uid: '0'
langcode: en
webform_id: ccdi_symposium_abstract
entity_type: node
entity_id: '2139'
locked: '0'
sticky: '0'
notes: ''
metatag: meta
data:
  authors_:
    - add_author_degree: ''
      add_author_first_name: Adam
      add_author_last_name: Thiesen
      add_author_middle: H.
      add_author_organization: 'The Jackson Laboratory, UConn School of Medicine'
    - add_author_degree: Ph.D.
      add_author_first_name: Sergii
      add_author_last_name: Domanskyi
      add_author_middle: ''
      add_author_organization: 'The Jackson Laboratory'
    - add_author_degree: Ph.D.
      add_author_first_name: 'Ali '
      add_author_last_name: 'Foroughi pour'
      add_author_middle: ''
      add_author_organization: "St. Jude Children's Research Hospital"
    - add_author_degree: ''
      add_author_first_name: Jingyan
      add_author_last_name: Zhang
      add_author_middle: ''
      add_author_organization: 'Johns Hopkins University'
    - add_author_degree: M.D.
      add_author_first_name: Todd
      add_author_last_name: Sheridan
      add_author_middle: B.
      add_author_organization: 'Hartford Healthcare'
    - add_author_degree: ''
      add_author_first_name: Steven
      add_author_last_name: Neuhauser
      add_author_middle: B.
      add_author_organization: 'The Jackson Laboratory'
    - add_author_degree: 'M.D. MPH'
      add_author_first_name: Alyssa
      add_author_last_name: Stetson
      add_author_middle: E.
      add_author_organization: 'Massachusetts General Hospital Department of Surgery'
    - add_author_degree: M.D.
      add_author_first_name: Katelyn
      add_author_last_name: Dannheim
      add_author_middle: ''
      add_author_organization: 'Massachusetts General Hospital Department of Pathology'
    - add_author_degree: 'M.D. MPH'
      add_author_first_name: Danielle
      add_author_last_name: Cameron
      add_author_middle: B.
      add_author_organization: 'Massachusetts General Hospital '
    - add_author_degree: 'M.D. Ph.D.'
      add_author_first_name: 'Shawn '
      add_author_last_name: Anh
      add_author_middle: ''
      add_author_organization: 'University of Pennsylvania Department of Surgery'
    - add_author_degree: 'M.D. Ph.D.'
      add_author_first_name: Hao
      add_author_last_name: Wu
      add_author_middle: ''
      add_author_organization: 'Yale School of Medicine Department of Pathology'
    - add_author_degree: M.D.
      add_author_first_name: 'Emily '
      add_author_last_name: Christison-Lagay
      add_author_middle: R.
      add_author_organization: 'Yale School of Medicine Department of Surgery'
    - add_author_degree: Ph.D.
      add_author_first_name: Carol
      add_author_last_name: Bult
      add_author_middle: J.
      add_author_organization: 'The Jackson Laboratory'
    - add_author_degree: Ph.D.
      add_author_first_name: Jeffrey
      add_author_last_name: Chuang
      add_author_middle: H.
      add_author_organization: 'The Jackson Laboratory, UConn School of Medicine'
    - add_author_degree: 'M.D. Ph.D.'
      add_author_first_name: 'Jill '
      add_author_last_name: Rubinstein
      add_author_middle: C.
      add_author_organization: 'The Jackson Laboratory, Hartford Healthcare, UConn School of Medicine'
  abstract: |
    Intro: Digitization of histopathology slides has allowed for the use of computational machine learning and artificial intelligence (AI)–based approaches to aid in diagnostics. These tools could be especially helpful for classifying pediatric sarcoma subtypes, which are rare and heterogeneous, and whose diagnoses often require costly genetic and molecular testing that may not be available to every patient. These machine learning models offer great promise but come with the caveat of being prone to overfitting to an individual institution’s microscope, scanner, and staining protocol, which can affect performance in real-world clinical settings where instruments and protocols differ across hospitals. Therefore, it can be difficult to make these models generalizable for use globally if they are not trained on a large and diverse dataset.

    Methods: We have curated over 700 H&E images from four institutions spanning over 10 different sarcoma subtypes. We utilize an in-house, open-source pipeline for stain normalization, focus checking, and cropping to harmonize the images. AI models are then used to extract features from these images, which can be used for downstream SAMPLER-based machine learning.

    Results: We achieve state-of-the-art results in classifying images as rhabdomyosarcoma vs non-rhabdomyosarcoma soft tissue sarcomas (AUC 0.969 ± 0.026), alveolar vs embryonal rhabdomyosarcoma (AUC 0.961 ± 0.021), and Ewing sarcoma (AUC 0.929). Importantly, our models generalize well when tested on data from previously unseen institutions, outperforming similar methods.

    Conclusion: Our pipeline is well suited for additional collaboration and could be a tool to help bridge access to clinical resources globally.
  abstract_title_: 'Generalizable Pediatric Sarcoma Histopathology Classification with Multi-Institutional Machine Learning'
  email_address_: adam.thiesen@jax.org
  institution_: 'The Jackson Laboratory'
  presenting_author_: 'Adam Thiesen'