Applying machine learning classifiers to automate quality assessment of paediatric dynamic susceptibility contrast (DSC-) MRI data.

Autor: Powell SJ; Physical Sciences for Health CDT, University of Birmingham, Birmingham, United Kingdom.; Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, United Kingdom., Withey SB; Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, United Kingdom.; Department of Oncology, Birmingham Children's Hospital, Birmingham, United Kingdom.; RRPPS, University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom., Sun Y; Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, United Kingdom.; School of Biological Sciences and Medical Engineering, Southeast University, Nanjing, China., Grist JT; Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, United Kingdom., Novak J; Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, United Kingdom.; Department of Oncology, Birmingham Children's Hospital, Birmingham, United Kingdom.; Department of Psychology, Aston Brain Centre, School of Life and Health Sciences, Aston University, Birmingham, United Kingdom., MacPherson L; Radiology, Birmingham Children's Hospital, Birmingham, United Kingdom., Abernethy L; Radiology, Alder Hey Children's NHS Foundation Trust, Liverpool, United Kingdom., Pizer B; Oncology, Alder Hey Children's NHS Foundation Trust, Liverpool, United Kingdom., Grundy R; The Children's Brain Tumour Research Centre, University of Nottingham, Nottingham, United Kingdom., Morgan PS; The Children's Brain Tumour Research Centre, University of Nottingham, Nottingham, United Kingdom.; Medical Physics, Nottingham University Hospitals, Nottingham, United Kingdom.; NIHR Nottingham Biomedical Research Centre, Nottingham, United Kingdom., Jaspan T; The Children's Brain Tumour Research Centre, University of Nottingham, Nottingham, United Kingdom.; Radiology, Nottingham University Hospitals, Nottingham, United Kingdom., Bailey S; Sir James Spence Institute of Child Health, Royal Victoria Infirmary, Newcastle upon Tyne, United Kingdom., Mitra D; Neuroradiology, The Newcastle upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, United Kingdom., Auer DP; Sir Peter Mansfield Imaging Centre, University of Nottingham, Nottingham, United Kingdom., Avula S; Radiology, Alder Hey Children's NHS Foundation Trust, Liverpool, United Kingdom., Arvanitis TN; Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, United Kingdom.; Department of Oncology, Birmingham Children's Hospital, Birmingham, United Kingdom.; Institute of Digital Healthcare, WMG, University of Warwick, Coventry, United Kingdom., Peet A; Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, United Kingdom.; Department of Oncology, Birmingham Children's Hospital, Birmingham, United Kingdom.
Jazyk: angličtina
Zdroj: The British journal of radiology [Br J Radiol] 2023 Apr 01; Vol. 96 (1145), pp. 20201465. Date of Electronic Publication: 2023 Feb 20.
DOI: 10.1259/bjr.20201465
Abstrakt: Objective: Investigate the performance of qualitative review (QR) for assessing dynamic susceptibility contrast (DSC-) MRI data quality in paediatric normal brain and develop an automated alternative to QR.
Methods: 1027 signal-time courses were assessed by Reviewer 1 using QR. 243 were additionally assessed by Reviewer 2 and % disagreements and Cohen's κ (κ) were calculated. The signal drop-to-noise ratio (SDNR), root mean square error (RMSE), full width half maximum (FWHM) and percentage signal recovery (PSR) were calculated for the 1027 signal-time courses. Data quality thresholds for each measure were determined using QR results. The measures and QR results trained machine learning classifiers. Sensitivity, specificity, precision, classification error and area under the curve from a receiver operating characteristic curve were calculated for each threshold and classifier.
Results: Comparing reviewers gave 7% disagreements and κ = 0.83. Data quality thresholds of: 7.6 for SDNR; 0.019 for RMSE; 3 s and 19 s for FWHM; and 42.9 and 130.4% for PSR were produced. SDNR gave the best sensitivity, specificity, precision, classification error and area under the curve values of 0.86, 0.86, 0.93, 14.2% and 0.83. Random forest was the best machine learning classifier, giving sensitivity, specificity, precision, classification error and area under the curve of 0.94, 0.83, 0.93, 9.3% and 0.89.
Conclusion: The reviewers showed good agreement. Machine learning classifiers trained on signal-time course measures and QR can assess quality. Combining multiple measures reduces misclassification.
Advances in Knowledge: A new automated quality control method was developed, which trained machine learning classifiers using QR results.
Databáze: MEDLINE
Externí odkaz: