Interpretable segmentation of medical free-text records based on word embeddings

Journal of Intelligent Information Systems - Medical free-text records store a lot of useful information that can be exploited in developing computer-supported medicine. However, extracting the...     

Prognostic Role of Tumoral PD-L1 and IDO1 Expression,and Intratumoral CD8+ and FoxP3+ Lymphocyte Infiltrates in 132 Primary Cutaneous Merkel Cell Carcinomas

The association of immune markers and clinicopathologic features and patient outcome has not been extensively studied in Merkel cell carcinoma (MCC). We correlated tumoral PD-L1 and IDO1 expression, and intratumoral CD8+ and FoxP3+ lymphocytes count with clinicopathologic variables, Merkel cell polyomavirus (MCPyV) status, and patient outcomes in a series of 132 MCC. By univariate analyses, tumoral PD-L1 expression >1% and combined tumoral PD-L1 >1% and high intratumoral FoxP3+ lymphocyte count correlated with improved overall survival (OS) (p = 0.016, 0.0072), MCC-specific survival (MSS) (p = 0.019, 0.017), and progression-free survival (PFS) (p = 0.043, 0.004, respectively). High intratumoral CD8+ and FoxP3+ lymphocyte count correlated with longer MSS (p = 0.036) and improved PFS (p = 0.047), respectively. Ulceration correlated with worse OS and worse MSS. Age, male gender, and higher stage (3 and 4) significantly correlated with worse survival. MCPyV positivity correlated with immune response. By multivariate analyses, only ulceration and age remained as independent predictors of worse OS; gender and stage remained for shorter PFS. Tumoral PD-L1 expression and increased density of intratumoral CD8+ lymphocytes and FoxP+ lymphocytes may represent favorable prognosticators in a subset of MCCs. Tumoral PD-L1 expression correlated with intratumoral CD8+ and FoxP3+ lymphocytes, which is supportive of an adaptive immune response.     

Locating multiple interacting quantitative trait loci using robust model selection

One of the most popular criteria for model selection is the Bayesian Information Criterion (BIC). It is based on an asymptotic approximation using Bayes rule when the sample size tends to infinity and the dimension of the model is fixed. Although it works well in classical applications, it performs less satisfactorily for high dimensional problems, i.e. when the number of regressors is very large compared to the sample size. For this reason, an alternative version of the BIC has been proposed for the problem of mapping quantitative trait loci (QTLs) considered in genetics. One approach is to locate QTLs by using model selection in the context of a regression model with an extremely large number of potential regressors. Since the assumption of normally distributed errors is often unrealistic in such settings, we extend the idea underlying the modified BIC to the context of robust regression.