RESUMEN
Random Forests are a powerful and frequently applied Machine Learning tool. The permutation variable importance (VIMP) has been proposed to improve the explainability of such a pure prediction model. It describes the expected increase in prediction error after randomly permuting a variable and disturbing its association with the outcome. However, VIMPs measure a variable's marginal influence only, that can make its interpretation difficult or even misleading. In the present work we address the general need for improving the explainability of prediction models by exploring VIMPs in the presence of correlated variables. In particular, we propose to use a variable's residual information for investigating if its permutation importance partially or totally originates from correlated predictors. Hypotheses tests are derived by a resampling algorithm that can further support results by providing test decisions and p-values. In simulation studies we show that the proposed test controls type I error rates. When applying the methods to a Random Forest analysis of post-transplant survival after kidney transplantation, the importance of kidney donor quality for predicting post-transplant survival is shown to be high. However, the transplant allocation policy introduces correlations with other well-known predictors, which raises the concern that the importance of kidney donor quality may simply originate from these predictors. By using the proposed method, this concern is addressed and it is demonstrated that kidney donor quality plays an important role in post-transplant survival, regardless of correlations with other predictors.
Asunto(s)
Trasplante de Riñón , Bosques Aleatorios , Humanos , Algoritmos , Simulación por Computador , Aprendizaje AutomáticoRESUMEN
Progression-free survival (PFS) is a frequently used endpoint in oncological clinical studies. In case of PFS, potential events are progression and death. Progressions are usually observed delayed as they can be diagnosed not before the next study visit. For this reason potential bias of treatment effect estimates for progression-free survival is a concern. In randomized trials and for relative treatment effects measures like hazard ratios, bias-correcting methods are not necessarily required or have been proposed before. However, less is known on cross-trial comparisons of absolute outcome measures like median survival times. This paper proposes a new method for correcting the assessment time bias of progression-free survival estimates to allow a fair cross-trial comparison of median PFS. Using median PFS for example, the presented method approximates the unknown posterior distribution by a Bayesian approach based on simulations. It is shown that the proposed method leads to a substantial reduction of bias as compared to estimates derived from maximum likelihood or Kaplan-Meier estimates. Bias could be reduced by more than 90% over a broad range of considered situations differing in assessment times and underlying distributions. By coverage probabilities of at least 94% based on the credibility interval of the posterior distribution the resulting parameters hold common confidence levels. In summary, the proposed approach is shown to be useful for a cross-trial comparison of median PFS.
Asunto(s)
Supervivencia sin Progresión , Teorema de Bayes , Sesgo , Supervivencia sin Enfermedad , Humanos , Estimación de Kaplan-MeierRESUMEN
A 5-inch-diameter Frisch Grid gas-proportional ionization chamber was utilized at Brookhaven National Laboratory (BNL) to rapidly characterize and quantify alpha-emitting actinides in unprocessed water, soil, air-filter, urine, and solid matrices. Instrument calibrations for the various matrices were performed by spiking representative samples with National Institute of Standards and Technology traceable isotopes of 230Th, 232U, 236Pu, and 243Am. Detection efficiencies were typically 15-20% for solid matrices (soil, concrete, filters, dry urine) and 45% for mass-less water samples. Instrument background over a 512-channel alpha-energy range of 3-8 MeV is very low at 0.01 cps. At optimum efficiency, minimum detectable levels of 0.56 mBq Kg(-1), 74 mBq L(-1) and 14.8 mBq filter(-1) were achievable for 40 x 10(-6) Kg soil, 1 x 10(-3) L tap water (or urine), and 4.5 cm diameter air-filter samples, respectively, each counted for 60 min. Data and spectra are presented showing the quality of results obtained using untreated samples obtained from the BNL Graphite Research Reactor Decommissioning Project. These samples contained Bq to MBq per gram amounts of (239,240)Pu, 241Am, and/or (234,235/238)U (as well as other beta/gamma emitters). Data and spectra are also presented for a very finely pulverized and homogeneous U.S. DOE/RESL soil reference standard (spiked with 239Pu, 241Am, and 233U) that was used to assess precision, accuracy, and reproducibility. Although this technique has its limitations, the advantages are (1) minimal sample preparation, (2) no separation chemistry required, (3) no chemical or hazardous waste generated, and (4) ability to immediately characterize and quantify alpha-emitting nuclides in most matrices. The benefits of this technique to the BNL/DOE Project Managers were rapid (1-2 d) turn-around times coupled with significant cost savings, as compared to commercial off-site analyses.