No abstract

Gliomas are primary brain tumors, originating from the glial cells in the brain. In contrast to the more traditional view of glioma as a localized disease, it is becoming clear that global brain functioning is impacted, even with respect to functional communication between brain regions remote from the tumor itself. However, a thorough investigation of glioma-related functional connectomic profiles is lacking. Therefore, we constructed functional brain networks using functional MR scans of 71 glioma patients and 19 matched healthy controls using the automated anatomical labelling (AAL) atlas and interregional Pearson correlation coefficients. The frequency distributions across connectivity values were calculated to depict overall connectomic profiles and quantitative features of these distributions (full-width half maximum (FWHM), peak position, peak height) were calculated. Next, we investigated the spatial distribution of the connectomic profile. We defined hub locations based on the literature and determined connectivity (1) between hubs, (2) between hubs and non-hubs, and (3) between non-hubs. Results show that patients had broader and flatter connectivity distributions compared to controls. Spatially, glioma patients particularly showed increased connectivity between non-hubs and hubs. Furthermore, connectivity distributions and hub-non-hub connectivity differed within the patient group according to tumor grade, while relating to Karnofsky performance status and progression-free survival. In conclusion, newly diagnosed glioma patients have globally altered functional connectomic profiles, which mainly affect hub connectivity and relate to clinical phenotypes. These findings underscore the promise of using connectomics as a future biomarker in this patient population.

Objective Research on pharmacovigilance from social media data has focused on mining adverse drug events (ADEs) using annotated datasets, with publications generally focusing on 1 of 3 tasks: ADE classification, named entity recognition for identifying the span of ADE mentions, and ADE mention normalization to standardized terminologies. While the common goal of such systems is to detect ADE signals that can be used to inform public policy, it has been impeded largely by limited end-to-end solutions for large-scale analysis of social media reports for different drugs. Materials and Methods We present a dataset for training and evaluation of ADE pipelines where the ADE distribution is closer to the average ‘natural balance’ with ADEs present in about 7% of the tweets. The deep learning architecture involves an ADE extraction pipeline with individual components for all 3 tasks. Results The system presented achieved state-of-the-art performance on comparable datasets and scored a classification performance of F1 = 0.63, span extraction performance of F1 = 0.44 and an end-to-end entity resolution performance of F1 = 0.34 on the presented dataset. Discussion The performance of the models continues to highlight multiple challenges when deploying pharmacovigilance systems that use social media data. We discuss the implications of such models in the downstream tasks of signal detection and suggest future enhancements. Conclusion Mining ADEs from Twitter posts using a pipeline architecture requires the different components to be trained and tuned based on input data imbalance in order to ensure optimal performance on the end-to-end resolution task.

In the medical domain, user-generated social media text is increasingly used as a valuablecomplementary knowledge source to scientific medical literature. The extraction of this knowledge iscomplicated by colloquial language use and misspellings. However, lexical normalization of suchdata has not been addressed effectively. This paper presents a data-driven lexical normalizationpipeline with a novel spelling correction module for medical social media. Our method significantlyoutperforms state-of-the-art spelling correction methods and can detect mistakes with an F1 of 0.63despite extreme imbalance in the data. We also present the first corpus for spelling mistake detectionand correction in a medical patient forum.

Transfer learning is promising for many NLP applications, especially in tasks with limited labeled data. This paper describes the methods developed by team TMRLeiden for the 2019 Social Media Mining for Health Applications (SMM4H) Shared Task. Our methods use state-of-the-art transfer learning methods to classify, extract and normalise adverse drug effects (ADRs) and to classify personal health mentions from health-related tweets. The code and fine-tuned models are publicly available. 1