Search

Background: Patients with mood disorders usually require repeated and prolonged hospitalization, resulting in a heavy burden on healthcare resources. This study aims to identify variables associated with length of stay(LOS) of repeatedly hospitalized patients with mood disorders and to provide information for optimizing psychiatry management and healthcare resource allocation. Methods: Electronic medical records (EMRs) of repeatedly hospitalized patients with mood disorders from January 2010 to December 2018 were collected and retrospectively analyzed. Chi-square and t-test were adopted to investigate the differences in characteristics between the two groups of short LOS and long LOS. Generalized estimating equation (GEE) was conducted to investigate potential factors influencing LOS. Results: A total of 2,009 repeatedly hospitalized patients with mood disorders were enrolled, of which 797 (39.7%) had a long LOS and 1,212 (60.3%) had a short LOS. Adverse effects of treatment, continuous clinical manifestation, chronic onset type, suicide attempt, comorbidity and use of antidepressants were positively associated with long LOS among all repeatedly hospitalized patients with mood disorders (P < 0.050). For patients with depression, factors associated with long LOS consisted of age, monthly income, adverse effects of treatment, continuous clinical manifestation, suicide attempt and comorbidity (P < 0.050). Whereas, for patients with bipolar disorder (BD), adverse effects of treatment, four or more hospitalizations and use of antidepressants contributed to the long LOS (P < 0.050). Influencing factors of LOS also vary among patients with different effectiveness of treatment. Conclusion: The LOS in repeatedly hospitalized patients with mood disorders was influenced by multiple factors. There were discrepancies in the factors affecting LOS in patients with different diagnoses and effectiveness of treatment, and specific factors should be addressed when evaluating the LOS. [ABSTRACT FROM AUTHOR]

Simple Summary: Electroencephalograms provide valuable insights into brain activity with diverse medical utility that encompasses diagnosis, monitoring, drug discovery, and treatment evaluation. We propose an artificial intelligence model that is uniquely optimized to analyze electroencephalogram signals by directly processing raw data. This model captures intricate spatial and temporal patterns in electroencephalograms through dedicated components like spatial channel attention and sparse transformer encoding. Through being evaluated extensively, our model demonstrates high accuracy in detecting brain diseases and classifying psychotropic drugs. By automatically learning the representations from raw electroencephalogram data, it adapts well across diseases, subjects, and tasks. The model's end-to-end learning aptitude and task versatility constitutes a robust and widely applicable automated electroencephalogram analytics solution. We believe it has the potential to significantly advance electroencephalogram-based diagnosis and personalized medicine. Electroencephalogram (EEG) analysis plays an indispensable role across contemporary medical applications, which encompasses diagnosis, monitoring, drug discovery, and therapeutic assessment. This work puts forth an end-to-end deep learning framework that is uniquely tailored for versatile EEG analysis tasks by directly operating on raw waveform inputs. It aims to address the challenges of manual feature engineering and the neglect of spatial interrelationships in existing methodologies. Specifically, a spatial channel attention module is introduced to emphasize the critical inter-channel dependencies in EEG signals through channel statistics aggregation and multi-layer perceptron operations. Furthermore, a sparse transformer encoder is used to leverage selective sparse attention in order to efficiently process long EEG sequences while reducing computational complexity. Distilling convolutional layers further concatenates the temporal features and retains only the salient patterns. As it was rigorously evaluated on key EEG datasets, our model consistently accomplished a superior performance over the current approaches in detection and classification assignments. By accounting for both spatial and temporal relationships in an end-to-end paradigm, this work facilitates a versatile, automated EEG understanding across diseases, subjects, and objectives through a singular yet customizable architecture. Extensive empirical validation and further architectural refinement may promote broader clinical adoption prospects. [ABSTRACT FROM AUTHOR]

Soil-free assays that induce water stress are routinely used to investigate drought responses in the plant Arabidopsis thaliana . Due to their ease of use, the research community often relies on polyethylene glycol (PEG), mannitol, and salt (NaCl) treatments to reduce the water potential of agar media, and thus induce drought conditions in the laboratory. However, while these types of stress can create phenotypes that resemble those of water deficit experienced by soil-grown plants, it remains unclear how these treatments compare at the transcriptional level. Here, we demonstrate that these different methods of lowering water potential elicit both shared and distinct transcriptional responses in Arabidopsis shoot and root tissue. When we compared these transcriptional responses to those found in Arabidopsis roots subject to vermiculite drying, we discovered many genes induced by vermiculite drying were repressed by low water potential treatments on agar plates (and vice versa). Additionally, we also tested another method for lowering water potential of agar media. By increasing the nutrient content and tensile strength of agar, we show the 'hard agar' (HA) treatment can be leveraged as a high-throughput assay to investigate natural variation in Arabidopsis growth responses to low water potential., Competing Interests: SG, JS, AY, JX, CM, NI, JN, WB, YZ, JE No competing interests declared, (© 2023, Gonzalez, Swift et al.)