Automated classification of cytogenetic abnormalities in hematolymphoid neoplasms.

Journal: Bioinformatics (Oxford, England)
Published Date:

Abstract

MOTIVATION: Algorithms for classifying chromosomes, like convolutional deep neural networks (CNNs), show promise to augment cytogeneticists' workflows; however, a critical limitation is their inability to accurately classify various structural chromosomal abnormalities. In hematopathology, recurrent structural cytogenetic abnormalities herald diagnostic, prognostic and therapeutic implications, but are laborious for expert cytogeneticists to identify. Non-recurrent cytogenetic abnormalities also occur frequently cancerous cells. Here, we demonstrate the feasibility of using CNNs to accurately classify many recurrent cytogenetic abnormalities while being able to reliably detect non-recurrent, spurious abnormal chromosomes, as well as provide insights into dataset assembly, model selection and training methodology that improve overall generalizability and performance for chromosome classification.

Authors

  • Andrew Cox
    Lyda Hill Department of Bioinformatics, UT Southwestern Medical Center, Dallas, TX 75235, USA.
  • Chanhee Park
    Sports(Movement Artificial-Intelligence Robotics Technology (SMART) Institute, Department of Physical Therapy, Yonsei University, Wonju, Republic of Korea.
  • Prasad Koduru
    Department of Pathology, UT Southwestern Medical Center, Dallas, TX 75235, USA.
  • Kathleen Wilson
    Department of Pathology, UT Southwestern Medical Center, Dallas, TX 75235, USA.
  • Olga Weinberg
    Department of Pathology, UT Southwestern Medical Center, Dallas, TX 75235, USA.
  • Weina Chen
    Department of Pathology, UT Southwestern Medical Center, Dallas, TX 75235, USA.
  • Rolando GarcĂ­a
    Department of Pathology, UT Southwestern Medical Center, Dallas, TX 75235, USA.
  • Daehwan Kim
    Lyda Hill Department of Bioinformatics, UT Southwestern Medical Center, Dallas, TX 75235, USA.