Advancements in deep learning speech representations have facilitated the effective use of extensive unlabeled speech datasets for Parkinson's disease (PD) modeling with minimal annotated data. This study employs the non-fine-tuned wav2vec 1.0 architecture to develop machine learning models for PD speech diagnosis tasks, such as cross-database classification and regression to predict demographic and articulation characteristics. The primary aim is to analyze overlapping components within the embeddings on both classification and regression tasks, investigating whether latent speech representations in PD are shared across models, particularly for related tasks. Firstly, evaluation using three multi-language PD datasets showed that wav2vec accurately detected PD based on speech, outperforming feature extraction using mel-frequency cepstral coefficients in the proposed cross-database classification scenarios. In cross-database scenarios using Italian and English-read texts, wav2vec demonstrated performance comparable to intra-dataset evaluations. We also compared our cross-database findings against those of other related studies. Secondly, wav2vec proved effective in regression, modeling various quantitative speech characteristics related to articulation and aging. Ultimately, subsequent analysis of important features examined the presence of significant overlaps between classification and regression models. The feature importance experiments discovered shared features across trained models, with increased sharing for related tasks, further suggesting that wav2vec contributes to improved generalizability. The study proposes wav2vec embeddings as a next promising step toward a speech-based universal model to assist in the evaluation of PD.
- MeSH
- Databases, Factual * MeSH
- Deep Learning MeSH
- Middle Aged MeSH
- Humans MeSH
- Parkinson Disease * physiopathology MeSH
- Speech * physiology MeSH
- Aged MeSH
- Machine Learning MeSH
- Check Tag
- Middle Aged MeSH
- Humans MeSH
- Male MeSH
- Aged MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
PURPOSE: The purpose of this study was to analyze pressure injury (PI) occurrence upon admission and at any time during the hospital course inpatients care facilities in the Czech Republic. Secondary aims were to evaluate demographic and clinical data of patients with PI and the impact of a PI on length of stay (LOS) in the hospital. DESIGN: Retrospective, cross-sectional analysis. SETTING AND SUBJECTS: The sample comprised data of hospitalized patients entered into the National Register of Hospitalized Patients (NRHOSP) database of the Czech Republic between 2007 and 2014 with a diagnosis L89 (pressure ulcer of unspecified site based on the International Classification of Diseases, Tenth Revision, ICD-10). Electronic records of 17,762,854 hospitalizations were reviewed. METHOD: Data from the NRHOSP from all acute and non-acute care hospitals in the Czech Republic were analyzed. Specifically, we analyzed patients admitted to acute and non-acute care facilities with a primary or secondary diagnosis of PI. RESULTS: The NRHOSP database included 17,762,854 cases, of which 46,224 cases (33,342 cases in acute care hospitals; 12,882 in non-acute care hospitals) had the L89 diagnosis (0.3%). The mean age of patients admitted with a PI was 73.8 ± 15.3 years (mean ± SD), and their average LOS was 33.2 ± 76.9 days. The mean LOS of patients hospitalized with L89 code as a primary diagnosis (n = 6877) was significantly longer compared to those patients for whom L89 code was a secondary diagnosis (25.8 vs 20.2 days, P < .001) in acute care facilities. In contrast, we found no difference in the mean LOS for patients hospitalized in non-acute care facility (58.7 days vs 65.1 days; P = .146) with ICD code L89. CONCLUSION: Pressure injuries were associated with significant LOS in both acute and non-acute care settings in the Czech Republic. Despite the valuable insights we obtained from the analysis of NRHOSP data, we advocate creation of a more valid and reliable electronic reporting system that enables policy makers to evaluate the quality and safety concerning PI and its impact on patients and the healthcare system.
- MeSH
- Databases, Factual standards statistics & numerical data MeSH
- Pressure Ulcer classification epidemiology nursing MeSH
- Inpatients statistics & numerical data MeSH
- Middle Aged MeSH
- Humans MeSH
- Hospitals standards statistics & numerical data MeSH
- Statistics, Nonparametric MeSH
- Cross-Sectional Studies MeSH
- Retrospective Studies MeSH
- Aged, 80 and over MeSH
- Aged MeSH
- Check Tag
- Middle Aged MeSH
- Humans MeSH
- Male MeSH
- Aged, 80 and over MeSH
- Aged MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
- Geographicals
- Czech Republic MeSH
IMPORTANCE: Surgeon-scientists are uniquely positioned to facilitate translation between the laboratory and clinical settings to drive innovation in patient care. However, surgeon-scientists face many challenges in pursuing research, such as increasing clinical demands that affect their competitiveness to apply for National Institutes of Health (NIH) funding compared with other scientists. OBJECTIVE: To examine how NIH funding has been awarded to surgeon-scientists over time. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used publicly available data from the NIH RePORTER (Research Portfolio Online Reporting Tools Expenditures and Results) database for research project grants awarded to departments of surgery between 1995 and 2020. Surgeon-scientists were defined as NIH-funded faculty holding an MD or MD-PhD degree with board certification in surgery; PhD scientists were NIH-funded faculty holding a PhD degree. Statistical analysis was performed from April 1 to August 31, 2022. MAIN OUTCOME: National Institutes of Health funding to surgeon-scientists compared with PhD scientists, as well as NIH funding to surgeon-scientists across surgical subspecialties. RESULTS: Between 1995 and 2020, the number of NIH-funded investigators in surgical departments increased 1.9-fold from 968 to 1874 investigators, corresponding to a 4.0-fold increase in total funding (1995, $214 million; 2020, $861 million). Although the total amount of NIH funding to both surgeon-scientists and PhD scientists increased, the funding gap between surgeon-scientists and PhD scientists increased 2.8-fold from a $73 million difference in 1995 to a $208 million difference in 2020, favoring PhD scientists. National Institutes of Health funding to female surgeon-scientists increased significantly at a rate of 0.53% (95% CI, 0.48%-0.57%) per year from 4.8% of grants awarded to female surgeon-scientists in 1995 to 18.8% in 2020 (P < .001). However, substantial disparity remained, with female surgeon-scientists receiving less than 20% of NIH grants and funding dollars in 2020. In addition, although there was increased NIH funding to neurosurgeons and otolaryngologists, funding to urologists decreased significantly from 14.9% of all grants in 1995 to 7.5% in 2020 (annual percent change, -0.39% [95% CI, -0.47% to -0.30%]; P < .001). Despite surgical diseases making up 30% of the global disease burden, representation of surgeon-scientists among NIH investigators remains less than 2%. CONCLUSION AND RELEVANCE: This study suggests that research performed by surgeon-scientists continues to be underrepresented in the NIH funding portfolio, highlighting a fundamental need to support and fund more surgeon-scientists.
- MeSH
- Biomedical Research * MeSH
- Surgeons * economics MeSH
- Databases, Factual MeSH
- Humans MeSH
- National Institutes of Health (U.S.) economics MeSH
- Cross-Sectional Studies MeSH
- Check Tag
- Humans MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
- Research Support, N.I.H., Extramural MeSH
- Research Support, U.S. Gov't, Non-P.H.S. MeSH
- Geographicals
- United States MeSH
BACKGROUND: Recent short-term clinical trials in patients with Duchenne Muscular Dystrophy (DMD) have indicated greater disease variability in terms of progression than expected. In addition, as average life-expectancy increases, reliable data is required on clinical progression in the older DMD population. OBJECTIVE: To determine the effects of corticosteroids on major clinical outcomes of DMD in a large multinational cohort of genetically confirmed DMD patients. METHODS: In this cross-sectional study we analysed clinical data from 5345 genetically confirmed DMD patients from 31 countries held within the TREAT-NMD global DMD database. For analysis patients were categorised by corticosteroid background and further stratified by age. RESULTS: Loss of ambulation in non-steroid treated patients was 10 years and in corticosteroid treated patients 13 years old (p = 0.0001). Corticosteroid treated patients were less likely to need scoliosis surgery (p < 0.001) or ventilatory support (p < 0.001) and there was a mild cardioprotective effect of corticosteroids in the patient population aged 20 years and older (p = 0.0035). Patients with a single deletion of exon 45 showed an increased survival in contrast to other single exon deletions. CONCLUSIONS: This study provides data on clinical outcomes of DMD across many healthcare settings and including a sizeable cohort of older patients. Our data confirm the benefits of corticosteroid treatment on ambulation, need for scoliosis surgery, ventilation and, to a lesser extent, cardiomyopathy. This study underlines the importance of data collection via patient registries and the critical role of multi-centre collaboration in the rare disease field.
- MeSH
- Databases as Topic MeSH
- Child MeSH
- Adult MeSH
- Muscular Dystrophy, Duchenne epidemiology genetics therapy MeSH
- Adrenal Cortex Hormones therapeutic use MeSH
- Infant MeSH
- Humans MeSH
- Adolescent MeSH
- Young Adult MeSH
- Infant, Newborn MeSH
- Child, Preschool MeSH
- Cross-Sectional Studies MeSH
- Treatment Outcome MeSH
- Check Tag
- Child MeSH
- Adult MeSH
- Infant MeSH
- Humans MeSH
- Adolescent MeSH
- Young Adult MeSH
- Male MeSH
- Infant, Newborn MeSH
- Child, Preschool MeSH
- Publication type
- Journal Article MeSH
... -- 5.4.5 Other protein domain databases 112 -- 5.4.6 Towards a unified database of protein families ... ... and domains 114 -- 5.5 2-D PAGE databases 114 -- 5.5.1 SWISS-2DPAGE 115 -- 5.5.2 Other 2-D PAGE databases ... ... 120 -- 5.8 Genomic databases 120 -- 5.8.1 OMIM 121 -- 5.9 Metabolic databases 123 -- 5.9.1 Some specific ... ... What future for protein databases? ... ... databases and analytical methods 166 -- 6.5 The future of proteome database integration 168 -- 6.5.1 ...
Principles and practice
1st ed. xviii, 243 s.
1st ed. viii, I-1 - V-66 s. + CD-ROM
197Au, 209Bi, 59Co, natFe and 169Tm samples were irradiated several times with quasi-monoenergetic neutrons from the p+7Li reaction in the energy range of 18-34 MeV. The activities of the samples were measured with the HPGe detector and the reaction rates were calculated. The cross sections were extracted using the SAND-II code with the reference cross sections from the IRDFF database.
While various QRS detection and classification methods were developed in the past, the Holter ECG data acquired during daily activities by wearable devices represent new challenges such as increased noise and artefacts due to patient movements. Here, we present a deep-learning model to detect and classify QRS complexes in single-lead Holter ECG. We introduce a novel approach, delivering QRS detection and classification in one inference step. We used a private dataset (12,111 Holter ECG recordings, length of 30 s) for training, validation, and testing the method. Twelve public databases were used to further test method performance. We built a software tool to rapidly annotate QRS complexes in a private dataset, and we annotated 619,681 QRS complexes. The standardised and down-sampled ECG signal forms a 30-s long input for the deep-learning model. The model consists of five ResNet blocks and a gated recurrent unit layer. The model's output is a 30-s long 4-channel probability vector (no-QRS, normal QRS, premature ventricular contraction, premature atrial contraction). Output probabilities are post-processed to receive predicted QRS annotation marks. For the QRS detection task, the proposed method achieved the F1 score of 0.99 on the private test set. An overall mean F1 cross-database score through twelve external public databases was 0.96 ± 0.06. In terms of QRS classification, the presented method showed micro and macro F1 scores of 0.96 and 0.74 on the private test set, respectively. Cross-database results using four external public datasets showed micro and macro F1 scores of 0.95 ± 0.03 and 0.73 ± 0.06, respectively. Presented results showed that QRS detection and classification could be reliably computed in one inference step. The cross-database tests showed higher overall QRS detection performance than any of compared methods.
- MeSH
- Algorithms MeSH
- Artifacts MeSH
- Electrocardiography, Ambulatory methods MeSH
- Electrocardiography methods MeSH
- Ventricular Premature Complexes * MeSH
- Humans MeSH
- Wearable Electronic Devices * MeSH
- Signal Processing, Computer-Assisted MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
- Research Support, Non-U.S. Gov't MeSH
Spiders are a highly diversified group of arthropods and play an important role in terrestrial ecosystems as ubiquitous predators, which makes them a suitable group to test a variety of eco-evolutionary hypotheses. For this purpose, knowledge of a diverse range of species traits is required. Until now, data on spider traits have been scattered across thousands of publications produced for over two centuries and written in diverse languages. To facilitate access to such data, we developed an online database for archiving and accessing spider traits at a global scale. The database has been designed to accommodate a great variety of traits (e.g. ecological, behavioural and morphological) measured at individual, species or higher taxonomic levels. Records are accompanied by extensive metadata (e.g. location and method). The database is curated by an expert team, regularly updated and open to any user. A future goal of the growing database is to include all published and unpublished data on spider traits provided by experts worldwide and to facilitate broad cross-taxon assays in functional ecology and comparative biology. Database URL:https://spidertraits.sci.muni.cz/.
- MeSH
- Arthropods * MeSH
- Databases, Factual MeSH
- Ecosystem MeSH
- Phenotype MeSH
- Spiders * genetics MeSH
- Animals MeSH
- Check Tag
- Animals MeSH
- Publication type
- Journal Article MeSH
- Research Support, Non-U.S. Gov't MeSH
