The use of decision trees for obtaining and representing clustering solutions is advantageous, due to their interpretability property. We propose a method called Decision Trees for Axis Unimodal Clustering (DTAUC), which constructs unsupervised binary decision trees for clustering by exploiting the concept of unimodality. Unimodality is a key property indicating the grouping behavior of data around a single density mode. Our approach is based on the notion of an axis unimodal cluster: a cluster where all features are unimodal, i.e., the set of values of each feature is unimodal as decided by a unimodality test. The proposed method follows the typical top-down splitting paradigm for building axis-aligned decision trees and aims to partition the initial dataset into axis unimodal clusters by applying thresholding on multimodal features. To determine the decision rule at each node, we propose a criterion that combines unimodality and separation. The method automatically terminates when all clusters are axis unimodal. Unlike typical decision tree methods, DTAUC does not require user-defined hyperparameters, such as maximum tree depth or the minimum number of points per leaf, except for the significance level of the unimodality test. Comparative experimental results on various synthetic and real datasets indicate the effectiveness of our method.

Introduction: Despite proven efficacy of anti-TNF agents in inflammatory bowel disease, primary non-response affects up to one-third of patients, while secondary loss of response occurs at 13–21% per patient-year, often requiring dose optimization or switching to alternative advanced therapies. Methods: The present single-center cohort study at the University Hospital of Ioannina included biologic-naïve patients receiving anti-TNF therapy as their first biologic treatment. First anti-TNF treatment failure was defined as discontinuation due to persistent IBD activity despite maximal dose optimization (infliximab 10 mg/kg every 4 weeks, adalimumab 40 mg weekly). Patients with measurable anti-drug antibodies prior to anti-TNF dose intensification or discontinuation were excluded. Of 528 anti-TNF-treated patients, 286 (173 with CD, 113 with UC) met the inclusion criteria and were included in the final statistical analysis. Results: Anti-TNF failure occurred in 32.7% of Crohn’s (CD) and 32.9% of ulcerative colitis (UC) patients. Multivariable Cox regression identified complicated phenotype (stricturing or/and penetrating CD; HR = 1.9, p = 0.032) and concomitant corticosteroid use at anti-TNF initiation (HR = 2.03, p = 0.012) as independent predictors of anti-TNF failure in CD. Age at CD diagnosis showed a trend for statistical significance (HR = 1.02, p = 0.061), and after stratification, age at diagnosis ≥ 40 years conferred higher risk (HR = 1.93, p = 0.016), alongside persistent effects of complicated phenotype (HR = 1.83, p = 0.027) and corticosteroid use (HR = 2.01, p = 0.013). In UC patients, female sex predicted anti-TNF failure (HR = 2.13, p = 0.025). IBD-related bowel resection occurred in 26.6% of patients with CD and in 5.3% of patients with UC. Conclusions: Anti-TNF failure remains common despite optimization. Identifying immunogenicity-independent predictors may enable personalized treatment strategies and improve outcomes.

Objective: Mild malformation of cortical development with oligodendroglial hyperplasia in epilepsy (MOGHE) is an underrecognized pediatric cortical lesion associated with somatic X‐linked SLC35A2 variants in approximately 50% of individuals. The genetic etiology in individuals without detectable SLC35A2 mutations remains undefined, which limits our understanding of its pathomechanism and clinical characterization. The aim is to identify genetic correlates of MOGHE in individuals lacking pathogenic SLC35A2 variants and to investigate the clinicopathological and genomic correlations. Methods: We investigated archival brain tissue of 29 individuals (19 males, 10 females) with histopathologically confirmed MOGHE. Twenty individuals carried SLC35A2 variants, and nine individuals lacked detectable mutations in SLC35A2 or other epilepsy‐related genes. Brain tissue samples were analyzed using DNA methylation‐derived copy number variation (CNV) analysis, polymerase chain reaction (PCR), and chromogenic/fluorescent in situ hybridization (ISH). Results: CNV analysis identified Y‐chromosomal material in five of 10 females (50%) and mosaic Y‐chromosome gains in 16 of 19 males (84%), confirmed by PCR and ISH, and associated with lesional tissue. Exploratory stratification of the cohort by SLC35A2 status and Y‐chromosome gain revealed a tendency toward earlier seizure onset, multilobar involvement, and cognitive impairment in patients carrying both SLC35A2 mutations and Y‐chromosome gain; however, these differences did not reach statistical significance in our small cohort. Significance: The combined presence of Y‐chromosomal material in females, mosaic Y‐chromosome gains in males, and SLC35A2 variants represents a recurrent genomic feature in patients with MOGHE. These findings suggest that sex‐chromosome anomalies may contribute to the pathobiology of this epilepsy‐associated cortical malformation, highlighting the need to clarify the underlying molecular mechanisms of this clinicohistopathologic entity. Although the root cause, biological consequence, and clinical significance of brain somatic Y‐chromosomal gain remain to be further clarified, our findings identify a previously unrecognized genomic feature of MOGHE and provide a rationale for future research in MOGHE cohorts.

Large language models (LLMs) have rapidly advanced natural language processing and are increasingly used to analyze clinical narratives. Their ability to extract information, summarize records, and support clinical workflows makes them potential tools for enhancing documentation efficiency and the secondary application in the analysis of electronic health record (EHR) data. The aim of this work is to synthesize recent evidence on methodological approaches and applications of LLMs for clinical narrative processing, and to assess their performance, benefits, limitations, and implications for clinical practice. Across 2022–2026 studies, LLMs demonstrated strong performance in information extraction, summarization, triage prediction, section classification, and synthetic text generation, often surpassing traditional machine-learning models. Overall, LLMs improved the conversion of unstructured notes into actionable clinical insights, reduced documentation burden, and supported decision-making tasks. Key challenges included hallucinations, variable reproducibility, sensitivity to prompting, domain adaptation gaps, and limited transparency. Our findings indicate that LLMs show substantial promise for transforming clinical narrative processing, but safe adoption requires rigorous evaluation and continuous model auditing. This work provides a structured, non-systematic synthesis of representative studies and is intended as a high-level overview of emerging applications rather than a comprehensive systematic review.

Deciding on the unimodality of a dataset is an important problem in data analysis and statistical modeling. It allows to obtain knowledge about the structure of the dataset, ie. whether data points have been generated by a probability distribution with a single or more than one peaks. Such knowledge is very useful for several data analysis problems, such as for deciding on the number of clusters and determining unimodal projections. We propose a technique called UU-test (Unimodal Uniform test) to decide on the unimodality of a one-dimensional dataset. The method operates on the empirical cumulative density function (ecdf) of the dataset. It attempts to build a piecewise linear approximation of the ecdf that is unimodal and models the data sufficiently in the sense that the data corresponding to each linear segment follows the uniform distribution. A unique feature of this approach is that in the case of unimodality, it also provides a statistical model of the data in the form of a Uniform Mixture Model. We present experimental results in order to assess the ability of the method to decide on unimodality and perform comparisons with the well-known dip-test approach. In addition, in the case of unimodal datasets we evaluate the Uniform Mixture Models provided by the proposed method using the test set log-likelihood and the two-sample Kolmogorov-Smirnov (KS) test.