Optimal threshold selection methods under tree or umbrella ordering

Cutoff estimation and construction of their confidence intervals for continuous biomarkers under ternary umbrella and tree stochastic ordering settings

Tuberculosis (TB) studies often involve four different states under consideration, namely, "healthy," "latent infection," "pulmonary active disease," and "extra-pulmonary active disease." While highly accurate clinical diagnosis tests do exist, they are expensive and generally not accessible in regions where they are most needed; thus, there is an interest in assessing the accuracy of new and easily obtainable biomarkers. For some such biomarkers, the typical stochastic ordering assumption might not be justified for all disease classes under study, and usual ROC methodologies that involve ROC surfaces and hypersurfaces are inadequate. Different types of orderings may be appropriate depending on the setting, and these may involve a number of ambiguously ordered groups that stochastically exhibit larger (or lower) marker scores than the remaining groups. Recently, there has been scientific interest on ROC methods that can accommodate these so-called "tree" or "umbrella" orderings. However, there is limited work discussing the estimation of cutoffs in such settings. In this article, we discuss the estimation and inference around optimized cutoffs when accounting for such configurations. We explore different cutoff alternatives and provide parametric, flexible parametric, and non-parametric kernel-based approaches for estimation and inference. We evaluate our approaches using simulations and illustrate them through a real data set that involves TB patients.

Post-test diagnostic accuracy measures under tree ordering of disease classes

The medical field commonly employs post-test measures such as predictive values and likelihood ratios to assess diagnostic accuracy. Predictive values, including positive and negative values (PPV and NPV), indicate the probability that individuals have a target health condition based on test results. On the other hand, likelihood ratios, including positive and negative ratios (LR+ and LR- respectively), compare the probability of a particular test result between the diseased and non-diseased groups. While predictive values are useful in evaluating diagnostic test accuracy in populations with varying disease prevalence, likelihood ratios provide a direct link between pre-test and post-test probabilities in specific patients. In this study, we introduce and analyze a new approach called generalized predictive values and likelihood ratios, using a tree ordering of disease classes. We evaluate the effectiveness of these methods through simulation studies and illustrate their use with real data on lung cancer.

Deep learning model calibration for improving performance in class-imbalanced medical image classification tasks

In medical image classification tasks, it is common to find that the number of normal samples far exceeds the number of abnormal samples. In such class-imbalanced situations, reliable training of deep neural networks continues to be a major challenge, therefore biasing the predicted class probabilities toward the majority class. Calibration has been proposed to alleviate some of these effects. However, there is insufficient analysis explaining whether and when calibrating a model would be beneficial. In this study, we perform a systematic analysis of the effect of model calibration on its performance on two medical image modalities, namely, chest X-rays and fundus images, using various deep learning classifier backbones. For this, we study the following variations: (i) the degree of imbalances in the dataset used for training; (ii) calibration methods; and (iii) two classification thresholds, namely, default threshold of 0.5, and optimal threshold from precision-recall (PR) curves. Our results indicate that at the default classification threshold of 0.5, the performance achieved through calibration is significantly superior (p < 0.05) to using uncalibrated probabilities. However, at the PR-guided threshold, these gains are not significantly different (p > 0.05). This observation holds for both image modalities and at varying degrees of imbalance. The code is available at https://github.com/sivaramakrishnan-rajaraman/Model_calibration.

Towards Amazon Forest Restoration: Automatic Detection of Species from UAV Imagery

Precise assessments of forest species’ composition help analyze biodiversity patterns, estimate wood stocks, and improve carbon stock estimates. Therefore, the objective of this work was to evaluate the use of high-resolution images obtained from Unmanned Aerial Vehicle (UAV) for the identification of forest species in areas of forest regeneration in the Amazon. For this purpose, convolutional neural networks (CNN) were trained using the Keras–Tensorflow package with the faster_rcnn_inception_v2_pets model. Samples of six forest species were used to train CNN. From these, attempts were made with the number of thresholds, which is the cutoff value of the function; any value below this output is considered 0, and values above are treated as an output 1; that is, values above the value stipulated in the Threshold are considered as identified species. The results showed that the reduction in the threshold decreases the accuracy of identification, as well as the overlap of the polygons of species identification. However, in comparison with the data collected in the field, it was observed that there exists a high correlation between the trees identified by the CNN and those observed in the plots. The statistical metrics used to validate the classification results showed that CNN are able to identify species with accuracy above 90%. Based on our results, which demonstrate good accuracy and precision in the identification of species, we conclude that convolutional neural networks are an effective tool in classifying objects from UAV images.

A comprehensive and comparative review of optimal cut-points selection methods for diseases with multiple ordinal stages

Cut-points selection is a key topic in the field of diagnostic studies. For binary classification, there exist several well-developed methods, some of which have been extended to three-class settings and beyond. This paper focuses on optimal cut-points selection methods for diseases with multiple ordinal stages. The purpose of this paper is two-fold: 1) to propose three new cut-points selection methods; and 2) to present a comprehensive simulation study to assess and compare the performance of all the available methods. Two real data sets, one from ovarian cancer and the other from pancreatic cancer, are analyzed.