Entity-enhanced BERT for medical specialty prediction based on clinical questionnaire data.

A medical specialty prediction system for remote diagnosis can reduce the unexpected costs incurred by first-visit patients who visit the wrong hospital department for their symptoms. To develop medical specialty prediction systems, several researchers have explored clinical predictive models using real medical text data. Medical text data include large amounts of information regarding patients, which increases the sequence length.

Enhanced high-energy proton radiation hardness of ZnO thin-film transistors with a passivation layer.

Metal-oxide thin-film semiconductors have been highlighted as next-generation space semiconductors owing to their excellent radiation hardness based on their dimensional advantages of very low thickness and insensitivity to crystal structure. However, thin-film transistors (TFTs) do not exhibit intrinsic radiation hardness owing to the chemical reactions at the interface exposed to ambient air. In this study, significantly enhanced radiation hardness of AlO-passivated ZnO TFTs against high-energy protons with energies of up to 100 MeV is obtained owing to the passivation layer blocking interactions with external reactants, thereby maintaining the chemical stability of the thin-film semiconductor.

Fall rates in bifocal, trifocal, and progressive addition lens glasses wearers.

Purpose: There is evidence that wearing multifocal glasses increases the risk of falls in older people, especially when walking on stairs and in unfamiliar settings. However, it is not clear whether all types of multifocal glasses are equally associated with falls. This study comprised a secondary analysis of data from the VISIBLE randomized controlled trial to determine whether fall rates differ among older bifocal, trifocal, and progressive addition lens glasses wearers.

Angular motion of the thorax during the golf swing: a comparison of two orientation angle sequences.

Orientation angles are commonly used to describe complex angular motions of the body. Selecting the most appropriate rotation sequence for a given segment's motion is crucial. The purpose of this study was to develop a set of generalisable, primary axis-centric sequence selection strategies and to compare the lean direction-lean-rotation (LDLR) sequence, selected for thoracic motion during golf driving based on the strategies, with the conventional rotation-bend-side bend (RBSB) sequence in assessing the level of inter-angle cross-talk.

Superloaded Multiplexed scRNA-seq Data Preserves Primary Immune Cell Heterogeneity but Necessitates Stringent Doublet Removal.

Background: Single-cell RNA sequencing (scRNA-seq) has improved our ability to characterize rare cell populations. In practice, cells from different tissues or donors are simultaneously loaded onto the instrument (multiplexed) at the recommended (standard loading) or higher (superloading) numbers to save time and money. Although cost-effective, superloading can stymie computational analyses owing to high multiplet rates and sample complexity.