TY - JOUR
T1 - Automated classification of cytogenetic abnormalities in hematolymphoid neoplasms
AU - Cox, Andrew
AU - Park, Chanhee
AU - Koduru, Prasad
AU - Wilson, Kathleen
AU - Weinberg, Olga
AU - Chen, Weina
AU - Garciá, Rolando
AU - Kim, Daehwan
N1 - Publisher Copyright:
© 2021 The Author(s) 2021. Published by Oxford University Press. All rights reserved.
PY - 2022/3/1
Y1 - 2022/3/1
N2 - 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. Results: Our top-performing model achieved a mean weighted F1 score of 96.86% on the validation set and 94.03% on the test set. Gradient class activation maps indicated that our model learned biologically meaningful feature maps, reinforcing the clinical utility of our proposed approach. Altogether, this work: proposes a new dataset framework for training chromosome classifiers for use in a clinical environment, reveals that residual CNNs and cyclical learning rates confer superior performance, and demonstrates the feasibility of using this approach to automatically screen for many recurrent cytogenetic abnormalities while adeptly classifying non-recurrent abnormal chromosomes.
AB - 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. Results: Our top-performing model achieved a mean weighted F1 score of 96.86% on the validation set and 94.03% on the test set. Gradient class activation maps indicated that our model learned biologically meaningful feature maps, reinforcing the clinical utility of our proposed approach. Altogether, this work: proposes a new dataset framework for training chromosome classifiers for use in a clinical environment, reveals that residual CNNs and cyclical learning rates confer superior performance, and demonstrates the feasibility of using this approach to automatically screen for many recurrent cytogenetic abnormalities while adeptly classifying non-recurrent abnormal chromosomes.
UR - http://www.scopus.com/inward/record.url?scp=85125505114&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85125505114&partnerID=8YFLogxK
U2 - 10.1093/bioinformatics/btab822
DO - 10.1093/bioinformatics/btab822
M3 - Article
C2 - 34874998
AN - SCOPUS:85125505114
SN - 1367-4803
VL - 38
SP - 1420
EP - 1426
JO - Bioinformatics
JF - Bioinformatics
IS - 5
ER -