Co-Inference of Data Mislabelings Reveals Improved Models in Genomics and Breast Cancer Diagnostics

Abstract

Mislabeling of cases as well as controls in case–control studies is a frequent source of strong bias in prognostic and diagnostic tests and algorithms. Common data processing methods available to the researchers in the biomedical community do not allow for consistent and robust treatment of labeled data in the situations where both, the case and the control groups, contain a non-negligible proportion of mislabeled data instances. This is an especially prominent issue in studies regarding late-onset conditions, where individuals who may convert to cases may populate the control group, and for screening studies that often have high false-positive/-negative rates. To address this problem, we propose a method for a simultaneous robust inference of Lasso reduced discriminative models and of latent group-specific mislabeling risks, not requiring any exactly labeled data. We apply it to a standard breast cancer imaging dataset and infer the mislabeling probabilities (being rates of false-negative and false-positive core-needle biopsies) together with a small set of simple diagnostic rules, outperforming the state-of-the-art BI-RADS diagnostics on these data. The inferred mislabeling rates for breast cancer biopsies agree with the published purely empirical studies. Applying the method to human genomic data from a healthy-ageing cohort reveals a previously unreported compact combination of single-nucleotide polymorphisms that are strongly associated with a healthy-ageing phenotype for Caucasians. It determines that 7.5% of Caucasians in the 1000 Genomes dataset (selected as a control group) carry a pattern characteristic of healthy ageing.

Cheap robust learning of data anomalies with analytically solvable entropic outlier sparsification

“…1 G and H , EOS with from SPA (EOS+SPA, red dashed lines), allows a statistically significant improvement of prediction performance (measured with the common performance measure area under curve [AUC]) for all of the tested mislabeling proportions p for all of the considered biomedical examples. As was shown recently, coinference of data mislabelings can significantly improve predictive performance of supervised classifiers ( 17 ). Application of the EOS algorithm with model loss function from SPA (EOS+SPA) allows achieving AUC of 0.96 and accuracy of ( SI Appendix , Fig.…”

Low-Cost Probabilistic 3D Denoising with Applications for Ultra-Low-Radiation Computed Tomography

“…The size of the training set, as well as the accuracy of prior data used in training, play a very central role also in the denoising and segmentation of CT images, where the number of instances in the training set T is significantly smaller than the feature space dimension D , corresponding to the number of voxels. A problem characterized by pertains to the so-called “small-data learning challenge” [ 34 , 35 , 36 , 37 , 38 ], and represents a scenario in which ML and DL approaches are prone to quickly overfit the small training set (which in addition often also contains missig data or incorrectly labeled data) and to achieve an unsatisfactory performance on the validation set [ 39 , 40 , 41 , 42 , 43 ]. To tackle this issue, several alternative approaches have been proposed [ 44 , 45 ], with transfer learning representing one of the most powerful alternatives [ 46 ].…”