Before the use of a diagnostic test in a routine clinical setting, the rigorous evaluation of its diagnostic accuracy is an essential step. The receiver operating characteristic (ROC) curve is the measure of accuracy most widely used for continuous diagnostic tests. However, the possible impact of extra information about the patient (or even the environment) on diagnostic accuracy needs to be also assessed. In this paper, attention is focused on an estimator for the covariate-speciﬁc ROC curve based on direct regression modelling and nonparametric smoothing techniques. This approach deﬁnes the class of generalized additive models for the ROC curve (ROC-GAM). The main aim of the paper is to oﬀer new inferential procedures for testing the eﬀect of covariates over the conditional ROC curve within the ROC-GAM context. Speciﬁcally, two diﬀerent bootstrap-based tests are suggested to check (a) the possible eﬀect of continuous covariates on the ROC curve; and (b) the presence of factor-by-curve interaction terms. The validity of the proposed bootstrap-based procedures is supported by simulations. To facilitate the application of these new procedures in practice, an R-package, known as npROCRegression, is provided and brieﬂy described. Finally, data derived from a computed-aided diagnostic (CAD) system for the automatic detection of tumour masses in breast cancer is analysed.
Type of Publication:
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research was supported by the Spanish Ministry of Economy and Competitiveness MINECO grants MTM2014-55966-P, MTM2014-52975-C2-1-R and BCAM Severo Ochoa excellence accreditation SEV-2013-0323, and by the Basque Government through the BERC 360 2014–2017.