Voxelwise statistical methods to localize practice variation in brain tumor surgery

English version

Autoři: Roelant Eijgelaar ^aff001; Philip C. De Witt Hamer ^aff002; Carel F. W. Peeters ^aff003; Frederik Barkhof ^aff004; Marcel van Herk ^aff006; Marnix G. Witte ^aff001
Působiště autorů: Cluster Radiation Oncology, The Netherlands Cancer Institute, Amsterdam, The Netherlands ^aff001; Neurosurgical Center Amsterdam, Brain Tumor Center Amsterdam, Amsterdam University Medical Center, location VUmc, Amsterdam, The Netherlands ^aff002; Department of Epidemiology & Biostatistics, Amsterdam Public Health Research Institute, Amsterdam University Medical Centers, location VUmc, Amsterdam, The Netherlands ^aff003; Department of Radiology & Nuclear Medicine, Amsterdam University Medical Center, Location VUmc, Amsterdam, The Netherlands ^aff004; Institutes of Neurology & Healthcare Engineering, University College London, London, United Kingdom ^aff005; Division of Cancer Sciences, Manchester Cancer Research Centre, School of Medical Sciences, Faculty of Biology, Medicine & Health, University of Manchester, Manchester Academic Health Sciences Centre, Manchester, United Kingdom ^aff006
Vyšlo v časopise: PLoS ONE 14(9)
Kategorie: Research Article
prolekare.web.journal.doi_sk: https://doi.org/10.1371/journal.pone.0222939

Souhrn

Purpose

During resections of brain tumors, neurosurgeons have to weigh the risk between residual tumor and damage to brain functions. Different perspectives on these risks result in practice variation. We present statistical methods to localize differences in extent of resection between institutions which should enable to reveal brain regions affected by such practice variation.

Methods

Synthetic data were generated by simulating spheres for brain, tumors, resection cavities, and an effect region in which a likelihood of surgical avoidance could be varied between institutions. Three statistical methods were investigated: a non-parametric permutation based approach, Fisher’s exact test, and a full Bayesian Markov chain Monte Carlo (MCMC) model. For all three methods the false discovery rate (FDR) was determined as a function of the cut-off value for the q-value or the highest density interval, and receiver operating characteristic and precision recall curves were created. Sensitivity to variations in the parameters of the synthetic model were investigated. Finally, all these methods were applied to retrospectively collected data of 77 brain tumor resections in two academic hospitals.

Results

Fisher’s method provided an accurate estimation of observed FDR in the synthetic data, whereas the permutation approach was too liberal and underestimated FDR. AUC values were similar for Fisher and Bayes methods, and superior to the permutation approach. Fisher’s method deteriorated and became too liberal for reduced tumor size, a smaller size of the effect region, a lower overall extent of resection, fewer patients per cohort, and a smaller discrepancy in surgical avoidance probabilities between the different surgical practices. In the retrospective patient data, all three methods identified a similar effect region, with lower estimated FDR in Fisher’s method than using the permutation method.

Conclusions

Differences in surgical practice may be detected using voxel statistics. Fisher’s test provides a fast method to localize differences but could underestimate true FDR. Bayesian MCMC is more flexible and easily extendable, and leads to similar results, but at increased computational cost.

Klíčová slova:

Surgical and invasive medical procedures – Permutation – Surgical resection – Surgical oncology – Bayesian method – Tumor resection

Zdroje

1. De Witt Hamer PC, Hendriks EJ, Mandonnet E, Barkhof F, Zwinderman AH, Duffau H. Resection probability maps for quality assessment of glioma surgery without brain location bias. PloS one. 2013;8(9):e73353. doi: 10.1371/journal.pone.0073353 24039922

2. Müller DMJ, Robe PAJT, Eijgelaar RS, Witte MG, Visser M, de Munck JC, et al. Comparing Glioblastoma Surgery Decisions Between Teams Using Brain Maps of Tumor Locations, Biopsies, and Resections. JCO Clinical Cancer Informatics. 2019;3:1–12. doi: 10.1200/CCI.18.00089 30673344

3. Storey JD. The positive false discovery rate: a Bayesian intepretation and the q-value. Annals of Statistics. 2003;3(6):2013–2035. doi: 10.1214/aos/1074290335

4. Nichols T, Holmes A. Nonparametric Permutation Tests for Functional Neuroimaging. Human Brain Function. 2004; p. 887–910.

5. Bates E, Wilson SM, Saygin AP, Dick F, Sereno MI, Knight RT, et al. Voxel-based lesion–symptom mapping. Nature neuroscience. 2003;6:448–50. doi: 10.1038/nn1050 12704393

6. Chen R, Herskovits EH. Voxel-based Bayesian lesion-symptom mapping. NeuroImage. 2010;49(1):597–602. doi: 10.1016/j.neuroimage.2009.07.061 19647797

7. Gilbert PB. A modified false discovery rate multiple-comparisons procedure for discrete data, applied to human immunodeficiency virus genetics. Journal of the Royal Statistical Society Series C: Applied Statistics. 2005;54(1):143–158. doi: 10.1111/j.1467-9876.2005.00475.x

8. Nettleton D, Hwang J, Caldo R, Wise R. Estimating the number of true null hypotheses from a histogram of p values. Journal of Agricultural, Biological, and Environmental Statistics. 2006;11(3):337–356. doi: 10.1198/108571106X129135

9. Dialsingh I, Austin SR, Altman NS. Estimating the proportion of true null hypotheses when the statistics are discrete. Bioinformatics. 2015;31(14):2303–2309. doi: 10.1093/bioinformatics/btv104 25735771

10. Carpenter B, Gelman A, Hoffman MD, Lee D, Goodrich B, Betancourt M, et al. Stan: A Probabilistic Programming Language. Journal of Statistical Software. 2017;76(1). doi: 10.18637/jss.v076.i01

11. Betancourt M, Girolami M. Hamiltonian Monte Carlo for Hierarchical Models. In: Upadhyay SK, Singh U, Dey D, Loganathan A, editors. Current Trends in Bayesian Methodology with Applications. Chapman and Hall; 2015. p. 79–102.

12. Kruschke JK, Vanpaemel W. Bayesian Estimation in Hierarchical Models. In: Busemeyer JR, Wang Z, Townsend JT, Eidels A, editors. The Oxford Handbook of Computational and Mathematical Psychology. Oxford, UK: Oxford University Press; 2015. p. 279–299.

13. Fonov V, Evans A, McKinstry R, Almli C, Collins D. Unbiased nonlinear average age-appropriate brain templates from birth to adulthood. NeuroImage. 2009;47:S102. https://doi.org/10.1016/S1053-8119(09)70884-5.

14. Carlson J, Heckerman D, Shani G. Estimating false discovery rates for contingency tables. Microsoft, Redmond, WA. 2009;MSR-TR-2009-53:1–23.

15. Rorden C, Karnath Ho, Bonilha L. Improving Lesion—Symptom Mapping. Journal of cognitive neuroscience. 2007; p. 1081–1088. doi: 10.1162/jocn.2007.19.7.1081 17583985

16. Visser M, Müller DMJ, van Duijn RJM, Smits M, Verburg N, Hendriks EJ, et al. Inter-rater agreement in glioma segmentations on longitudinal MRI. NeuroImage: Clinical. 2019;22(February):101727. doi: 10.1016/j.nicl.2019.101727

17. Benjamini Y, Heller R. False Discovery Rates for Spatial Signals. Journal of the American Statistical Association. 2007;102(480):1272–1281. doi: 10.1198/016214507000000941