Refine
Year of publication
- 2022 (3) (remove)
Document Type
- Article (2)
- Working Paper (1)
Language
- English (3)
Has Fulltext
- yes (3)
Is part of the Bibliography
- no (3)
Keywords
Institute
Sudden cardiac death (SCD) in adolescents and young adults may be the first manifestation of an inherited arrhythmic syndrome. Thus identification of a genetic origin in sudden death cases deemed inconclusive after a comprehensive autopsy and may help to reduce the risk of lethal episodes in the remaining family. Using next-generation sequencing (NGS), a large number of variants of unknown significance (VUS) are detected. In the majority of cases, there is insufficient evidence of pathogenicity, representing a huge dilemma in current genetic investigations. Misinterpretation of such variants may lead to inaccurate genetic diagnoses and/or the adoption of unnecessary and/or inappropriate therapeutic approaches. In our study, we applied current (ACMG) recommendations for variant classification in post-mortem genetic screening of a cohort of 56 SCD victims. We identified a total 53 rare protein-altering variants (MAF < 0.2%) classified as VUS or worse. Twelve percent of the cases exhibited a clinically actionable variant (pathogenic, likely pathogenic or VUS – potentially pathogenic) that would warrant cascade genetic screening in relatives. Most of the variants detected by means of the post-mortem genetic investigations were VUS. Thus, genetic testing by itself might be fairly meaningless without supporting background data. This data reinforces the need for an experienced multidisciplinary team for obtaining reliable and accountable interpretations of variant significance for elucidating potential causes for SCDs in the young. This enables the early identification of relatives at risk or excludes family members as genetic carriers. Also, development of adequate forensic guidelines to enable appropriate interpretation of rare genetic variants is fundamental.
We investigate what statistical properties drive risk-taking in a large set of observational panel data on online poker games (n=4,450,585). Each observation refers to a choice between a safe 'insurance' option and a binary lottery of winning or losing the game. Our setting offers a real-world choice situation with substantial incentives where probability distributions are simple, transparent, and known to the individuals. We find that individuals reveal a strong and robust preference for skewness. The effect of skewness is most pronounced among experienced and losing players but remains highly significant for winning players, in contrast to the variance effect.
Importance: The entry of artificial intelligence into medicine is pending. Several methods have been used for the predictions of structured neuroimaging data, yet nobody compared them in this context.
Objective: Multi-class prediction is key for building computational aid systems for differential diagnosis. We compared support vector machine, random forest, gradient boosting, and deep feed-forward neural networks for the classification of different neurodegenerative syndromes based on structural magnetic resonance imaging.
Design, setting, and participants: Atlas-based volumetry was performed on multi-centric T1-weighted MRI data from 940 subjects, i.e., 124 healthy controls and 816 patients with ten different neurodegenerative diseases, leading to a multi-diagnostic multi-class classification task with eleven different classes.
Interventions: N.A.
Main outcomes and measures: Cohen’s kappa, accuracy, and F1-score to assess model performance.
Results: Overall, the neural network produced both the best performance measures and the most robust results. The smaller classes however were better classified by either the ensemble learning methods or the support vector machine, while performance measures for small classes were comparatively low, as expected. Diseases with regionally specific and pronounced atrophy patterns were generally better classified than diseases with widespread and rather weak atrophy.
Conclusions and relevance: Our study furthermore underlines the necessity of larger data sets but also calls for a careful consideration of different machine learning methods that can handle the type of data and the classification task best.