Toward Automated Detection of Silent Cerebral Infarcts in Children and Young Adults With Sickle Cell Anemia

Authors

Yasheng Chen, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Yan Wang, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Chia-Ling Phuah, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.Follow
Melanie E. Fields, Division of Pediatric Hematology/Oncology (M.E.F.), Washington University School of Medicine, St. Louis, MO.
Kristin P. Guilliams, Division of Pediatric Neurology (K.P.G.), Washington University School of Medicine, St. Louis, MO.
Slim Fellah, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Martin N. Reis, Mallinckrodt Institute of Radiology (M.N.R., H.A., J.-M.L., R.C.M., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Michael M. Binkley, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Hongyu An, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Jin-Moo Lee, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Robert C. McKinstry, Mallinckrodt Institute of Radiology (M.N.R., H.A., J.-M.L., R.C.M., A.L.F.), Washington University School of Medicine, St. Louis, MO.
Lori C. Jordan, Division of Pediatric Neurology, Department of Pediatrics, Vanderbilt University of Medicine, Nashville, TN (L.C.J.).
Michael R. DeBaun, Division of Hematology and Oncology, Department of Pediatrics, Vanderbilt-Meharry Center of Excellence in Sickle Cell Disease, Vanderbilt University Medical Center, Nashville, TN (M.R.D.).
Andria L. Ford, Department of Neurology (Y.C., Y.W., C.-L.P., S.F., M.M.B., H.A., J.-M.L., A.L.F.), Washington University School of Medicine, St. Louis, MO.

Document Type

Article

Abstract

BACKGROUND: Silent cerebral infarcts (SCI) in sickle cell anemia (SCA) are associated with future strokes and cognitive impairment, warranting early diagnosis and treatment. Detection of SCI, however, is limited by their small size, especially when neuroradiologists are unavailable. We hypothesized that deep learning may permit automated SCI detection in children and young adults with SCA as a tool to identify the presence and extent of SCI in clinical and research settings. METHODS: We utilized UNet-a deep learning model-for fully automated SCI segmentation. We trained and optimized UNet using brain magnetic resonance imaging from the SIT trial (Silent Infarct Transfusion). Neuroradiologists provided the ground truth for SCI diagnosis, while a vascular neurologist manually delineated SCI on fluid-attenuated inversion recovery and provided the ground truth for SCI segmentation. UNet was optimized for the highest spatial overlap between automatic and manual delineation (dice similarity coefficient). The optimized UNet was externally validated using an independent single-center prospective cohort of SCA participants. Model performance was evaluated through sensitivity and accuracy (%correct cases) for SCI diagnosis, dice similarity coefficient, intraclass correlation coefficient (metric of volumetric agreement), and Spearman correlation. RESULTS: The SIT trial (n=926; 31% with SCI; median age, 8.9 years) and external validation (n=80; 50% with SCI; age, 11.5 years) cohorts had small median lesion volumes of 0.40 and 0.25 mL, respectively. Compared with the neuroradiology diagnosis, UNet predicted SCI presence with 100% sensitivity and 74% accuracy. In magnetic resonance imaging with SCI, UNet reached a moderate spatial agreement (dice similarity coefficient, 0.48) and high volumetric agreement (intraclass correlation coefficient, 0.76; ρ=0.72; <0.001) between automatic and manual segmentations. CONCLUSIONS: UNet, trained using a large pediatric SCA magnetic resonance imaging data set, sensitively detected small SCI in children and young adults with SCA. While additional training is needed, UNet may be integrated into the clinical workflow as a screening tool, aiding in SCI diagnosis.

Medical Subject Headings

Child; Humans; Young Adult; Prospective Studies; Anemia, Sickle Cell (complications, diagnostic imaging, therapy); Cerebral Infarction (complications); Brain; Magnetic Resonance Imaging

Publication Date

8-1-2023

Publication Title

Stroke

E-ISSN

1524-4628

Volume

54

Issue

8

First Page

2096

Last Page

2104

PubMed ID

37387218

Digital Object Identifier (DOI)

10.1161/STROKEAHA.123.042683

Recommended Citation

Chen, Yasheng; Wang, Yan; Phuah, Chia-Ling; Fields, Melanie E.; Guilliams, Kristin P.; Fellah, Slim; Reis, Martin N.; Binkley, Michael M.; An, Hongyu; Lee, Jin-Moo; McKinstry, Robert C.; Jordan, Lori C.; DeBaun, Michael R.; and Ford, Andria L., "Toward Automated Detection of Silent Cerebral Infarcts in Children and Young Adults With Sickle Cell Anemia" (2023). Neurology. 1736.
https://scholar.barrowneuro.org/neurology/1736