Refine
Document Type
- Article (3)
Has Fulltext
- yes (3)
Is part of the Bibliography
- no (3)
Keywords
- COVID (1)
- Comparative analysis (1)
- Deep neural network (1)
- Gradient boosting (1)
- Multi-syndrome classification (1)
- Neurodegenerative syndromes (1)
- Random forest (1)
- Support vector machine (1)
- accident (1)
- injury (1)
- lockdown (1)
- pediatric intensive care (1)
- trauma (1)
Institute
- Medizin (2)
- Präsidium (1)
- Rechtswissenschaft (1)
Es wäre eine bessere Welt, würde es diese Bilder nicht geben: Die Rede ist von Darstellungen, die sexuellen Missbrauch von und sexualisierte Gewalt an Kindern und Jugendlichen zeigen. Die physischen und psychischen Verletzungen, die durch den Missbrauch, aber auch durch dessen Perpetuierung in Bildern verursacht werden, sind unermesslich. Daher greift die Gesellschaft zu einem ihrer schärfsten Schwerter – dem Strafrecht.
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.
Children’s and adolescents’ lives drastically changed during COVID lockdowns worldwide. To compare accident- and injury-related admissions to pediatric intensive care units (PICU) during the first German COVID lockdown with previous years, we conducted a retrospective multicenter study among 37 PICUs (21.5% of German PICU capacities). A total of 1444 admissions after accidents or injuries during the first lockdown period and matched periods of 2017–2019 were reported and standardized morbidity ratios (SMR) were calculated. Total PICU admissions due to accidents/injuries declined from an average of 366 to 346 (SMR 0.95 (CI 0.85–1.05)). Admissions with trauma increased from 196 to 212 (1.07 (0.93–1.23). Traffic accidents and school/kindergarten accidents decreased (0.77 (0.57–1.02 and 0.26 (0.05–0.75)), whereas household and leisure accidents increased (1.33 (1.06–1.66) and 1.34 (1.06–1.67)). Less neurosurgeries and more visceral surgeries were performed (0.69 (0.38–1.16) and 2.09 (1.19–3.39)). Non-accidental non-suicidal injuries declined (0.73 (0.42–1.17)). Suicide attempts increased in adolescent boys (1.38 (0.51–3.02)), but decreased in adolescent girls (0.56 (0.32–0.79)). In summary, changed trauma mechanisms entailed different surgeries compared to previous years. We found no evidence for an increase in child abuse cases requiring intensive care. The increase in suicide attempts among boys demands investigation.