Explore the vast range of titles available.

Aim: This study assesses the predictive ability of the full Work Ability Index (WAI) as well as its individual items in the general population. Methods: The Work, Health and Retirement Study (WHRS) is a stratified random national sample of 25–75-year-olds living in Sweden in 2000 that received a postal questionnaire (n = 6637, response rate = 53%). Current and subsequent sickness absence was obtained from registers. The ability of the WAI to predict long-term sickness absence (LTSA; ⩾ 90 consecutive days) during a period of four years was analysed by logistic regression, from which the Area Under the Receiver Operating Characteristic curve (AUC) was computed. Results: There were 313 incident LTSA cases among 1786 employed individuals. The full WAI had acceptable ability to predict LTSA during the 4-year follow-up (AUC = 0.79; 95% CI 0.76 to 0.82). Individual items were less stable in their predictive ability. However, three of the individual items: current work ability compared with lifetime best, estimated work impairment due to diseases, and number of diagnosed current diseases, exceeded AUC > 0.70. Excluding the WAI item on number of days on sickness absence did not result in an inferior predictive ability of the WAI. Conclusions: The full WAI has acceptable predictive validity, and is superior to its individual items. For public health surveys, three items may be suitable proxies of the full WAI; current work ability compared with lifetime best, estimated work impairment due to diseases, and number of current diseases diagnosed by a physician.

In unsupervised domain adaptation (UDA), knowledge is transferred from label-rich source domains to relevant but unlabeled target domains. Current most popular state-of-the-art works suggest that performing domain alignment from the class perspective can alleviate domain shift. However, most of them based on domain adversarial which is hard to train and converge. In this paper, we propose a novel contrastive learning to improve diversity and discriminability for domain adaptation, dubbed as IDD_ICL, which improve the discriminativeness of the model while increasing the sample diversity. To be precise, we first design a novel implicits contrastive learning loss at sample-level by implicit augment sample of the source domain. While augmenting the diversity of the source domain, we can cluster the samples of the same category in the source domain together, and disperse the samples of different categories, thereby improving the discriminative ability of the model. Furthermore, we show that our algorithm is effective by implicitly learning an infinite number of similar samples. Our results demonstrate that our method doesn’t require complex technologies or specialized equipment, making it readily adoptable and applicable in practical scenarios.

We assessed the test–retest reliability and discriminative ability of a somatosensory temporal discrimination (SSTD) assessment tool for fibromyalgia syndrome (FMS) and determined if pain-related variables were associated with SSTD performance. Twenty-five women with FMS and twenty-five asymptomatic women were assessed during two sessions 7 to 10 days apart. The proportion of correct responses (range 0–100) was calculated. Sociodemographic information was collected for both groups. The participants with FMS also completed the widespread pain index and the Brief Pain Inventory. Test–retest reliability was verified by calculating intraclass correlation coefficients. Discriminative ability was verified by a between-group comparison of scores using a t-test. Associations between SSTD score and pain variables were tested using Pearson or Spearman correlation coefficients. The test–retest reliability of the SSTD score was excellent (ICC > 0.9, CI: 0.79–0.96) for the asymptomatic group and good for the FMS group (ICC: 0.81, 95% CI: 0.62–0.91). The median (Q1–Q3) test session SSTD score differed significantly between the FMS 84.1 (71–88) and the asymptomatic 91.6 (83.4–96.1) groups (p < 0.001). Only pain duration was associated with the SSTD score. In conclusion, the new SSTD test seems reliable for people with FMS and is discriminative. Further studies should examine its sensitivity to change and correlations with other SSTD tests.

Objective To construct a machine learning diagnostic model integrating feature dimensionality reduction techniques and artificial neural network classifiers to develop the value of clinical routine blood indexes for the auxiliary diagnosis of ovarian cancer. Methods Patients with ovarian cancer clearly diagnosed in our hospital were collected as a case group ( n  = 185), and three groups of patients with other malignant otolaryngology tumors ( n  = 138), patients with benign otolaryngology diseases ( n  = 339) and those with normal physical examination ( n  = 92) were used as an overall control group. In this paper, a fully automated segmentation network for magnetic resonance images of ovarian cancer is proposed to improve the reproducibility of tumor segmentation results while effectively reducing the burden on radiologists. A pre-trained Res Net50 is used to the three edge output modules are fused to obtain the final segmentation results. The segmentation results of the proposed network architecture are compared with the segmentation results of the U-net based network architecture and the effect of different loss functions and region of interest sizes on the segmentation performance of the proposed network is analyzed. Results The average Dice similarity coefficient, average sensitivity, average specificity (specificity) and average hausdorff distance of the proposed network segmentation results reached 83.62%, 89.11%, 96.37% and 8.50, respectively, which were better than the U-net based segmentation method. For ROIs containing tumor tissue, the smaller the size, the better the segmentation effect. Several loss functions do not differ much. The area under the ROC curve of the machine learning diagnostic model reached 0.948, with a sensitivity of 91.9% and a specificity of 86.9%, and its diagnostic efficacy was significantly better than that of the traditional way of detecting CA125 alone. The model was able to accurately diagnose ovarian cancer of different disease stages and showed certain discriminative ability for ovarian cancer in all three control subgroups. Conclusion Using machine learning to integrate multiple conventional test indicators can effectively improve the diagnostic efficacy of ovarian cancer, which provides a new idea for the intelligent auxiliary diagnosis of ovarian cancer.

Objectives: To explore the correlation of the neutrophil-to-lymphocyte ratio (NLR) and the C-reactive protein-to-albumin ratio (CRP/ALB, CAR) with renal anemia in maintenance hemodialysis (MHD) patients. Methodology: Clinical data of 275 MHD patients admitted to the Hemodialysis Center of the First Affiliated Hospital of Qiqihar Medical University from December 2022 to December 2024 were retrospectively analyzed. These patients were divided into an anemia (hemoglobin [Hb] < 110 g/L) and a non-anemia (Hb ≥ 110 g/L) group based on the presence of renal anemia. Patients in the anemia group were further stratified by anemia severity into a mild subgroup (Hb > 90 g/L), a moderate subgroup (60 g/L < Hb ≤ 90 g/L), and a severe subgroup (Hb ≤ 60 g/L). The levels of NLR and CAR were measured and compared across all groups. Receiver operating characteristic (ROC) curves were used to analyze the value of each indicator in evaluating the discriminative ability of each indicator for renal anemia status. Additionally, the relationship between NLR/CAR levels and the severity of anemia was evaluated. Results: A total of 92 patients (33.5%) were diagnosed with renal anemia. The levels of NLR and CAR in the anemia group were significantly higher than those in the non-anemia group (P < 0.05). The ROC analysis showed that the areas under the curves (AUCs) for NLR and CAR in diagnosing renal anemia in MHD patients were 0.825 and 0.894, respectively (P < 0.05). With increasing anemia severity, NLR and CAR levels increased significantly (P < 0.05). Conclusion: NLR and CAR exhibit significant discriminative ability for renal anemia in MHD patients and are correlated with the severity of the condition.

The construction of diverse dictionaries for sparse representation of hyperspectral image (HSI) classification has been a hot topic over the past few years. However, compared with convolutional neural network (CNN) models, dictionary-based models cannot extract deeper spectral information, which will reduce their performance for HSI classification. Moreover, dictionary-based methods have low discriminative capability, which leads to less accurate classification. To solve the above problems, we propose a deep learning-based structure dictionary for HSI classification in this paper. The core ideas are threefold, as follows: (1) To extract the abundant spectral information, we incorporate deep residual neural networks in dictionary learning and represent input signals in the deep feature domain. (2) To enhance the discriminative ability of the proposed model, we optimize the structure of the dictionary and design sharing constraint in terms of sub-dictionaries. Thus, the general and specific feature of HSI samples can be learned separately. (3) To further enhance classification performance, we design two kinds of loss functions, including coding loss and discriminating loss. The coding loss is used to realize the group sparsity of code coefficients, in which within-class spectral samples can be represented intensively and effectively. The Fisher discriminating loss is used to enforce the sparse representation coefficients with large between-class scatter. Extensive tests performed on hyperspectral dataset with bright prospects prove the developed method to be effective and outperform other existing methods.

BackgroundSit-To-Stand (STS) tests are reported as feasible alternatives for the assessment of functional fitness but the reliability of these tests in people with coronary artery disease (CAD) has not been reported. This study explored the test-retest reliability, convergent and known-groups validity of the five times, 30-sec and 1-min sit-to-stand test (FTSTS test, 30-s STS test and 1-min STS test respectively) in patients with CAD. The feasibility of applying these tests to distinguish the level of risk for cardiovascular events in CAD patients was also investigated.MethodsPatients with stable CAD performed a 6MWT and 3 STS tests in random order on the same day. Receiver operating characteristic (ROC) curve analyses were conducted using STS test data to differentiate patients with low or high risk of cardiovascular events based on the risk level determined by distance covered in the 6MWT as > or ≤ 419 m. Thirty patients repeated the 3 STS tests on the following day.Results112 subjects with diagnoses of atherosclerosis or post-percutaneous coronary intervention, or post-acute myocardial infarction (post-AMI) participated in the validity analysis. All 3 STS tests demonstrated moderate and significant correlation with the 6MWT (coefficient values r for the FTSTS, 30-s STS and 1-min STS tests were−0.53, 0.57 and 0.55 respectively). Correlations between left ventricular ejection fraction (LVEF) and all STS tests and between 6MWT and LVEF were only weak ( r values ranged from 0.27 to 0.31). Subgroup analysis showed participants in the post-AMI group performed worse in all tests compared to non-myocardial infarction (non-MI) group. The area under the curve (AUC) was 0.80 for FTSTS (sensitivity: 75.0%, specificity: 73.8%, optimal cut-off: >11.7 sec), and the AUC, sensitivity, specificity and optimal cut-off for 30-s STS and 1-min STS test were 0.83, 75.0%, 76.2%, ≤ 12 repetitions and 0.80, 71.4%, 73.8%, ≤ 23 repetitions respectively. The intraclass correlation coefficients (ICC) for repeated measurements of the FTSTS, 30-s STS and 1-min STS tests were 0.96, 0.95 and 0.96 respectively, with the minimal detectable change (MDC95) computed to be 1.1 sec 1.8 repetitions and 3.9 repetitions respectively.ConclusionsAll STS tests demonstrated good test-retest reliability, convergent and known-groups validity. STS tests may discriminate low from high levels of risk for a cardiovascular event in patients with CAD.

Obesity, a risk factor for obstructive sleep apnea (OSA), is usually estimated by body mass index (BMI). However, other adiposity-estimating equations may better capture variations in fat distribution. This study assessed the relationship between OSA severity and 15 adiposity-estimating equations, compared to BMI, with subgroup analyses by sex and age (<50 vs ≥50). We conducted a cross-sectional cohort study using data from 5021 consecutive adults who underwent a Level 1 polysomnography (2015-2017) in a large academic sleep center in Ottawa, Canada. We assessed correlations between adiposity measures and the apnea-hypopnea index (AHI) and examined discriminative ability for moderate-to-severe (AHI ≥15/h) and severe OSA (AHI >30/h) using univariate logistic regressions. The mean age was 49.5 years, 46.6% were women; the mean BMI was 30.0 kg/m and 12.7% had severe OSA. All adiposity equations showed negligible (Pearson r 0.0 to ±0.3) to low (Pearson r ± 0.30 to 0.50) statistically significant correlations with AHI, with many of the equations having a marginally stronger correlation coefficient than BMI, in total and subgroup analysis. Discriminative ability for severe OSA was generally low, with c-indices ranging from 0.52 to 0.67 in the overall sample. However, in females under 50, several equations (eg, Gallagher 2000, Deurenberg 1991 and 1998, ECORE BF) reached excellent discriminative ability (c-indices 0.81), including BMI (c-index 0.80). This pattern was not observed in other subgroups. In this clinical cohort, BMI was associated poorly with AHI; however, the other equations did not outperform BMI. Moreover, BMI demonstrated poor discriminative ability for moderate/severe and severe OSA, with none of the other equations performing better in this context. Notable subgroup differences-particularly among younger females-suggest that tailoring screening strategies by age and sex may improve risk stratification and support refining obesity-based screening approaches.

Aim Ecological niche modelling requires robust estimation of model performance and significance, but common evaluation approaches often yield biased estimates. Null models provide a solution but are rarely used in this field. We implemented an important modification to existing null model tests, evaluating null models with the same withheld records that were used to evaluate the real model. We built and evaluated models across a range of modelling scenarios and for various performance measures using the algorithm Maxent and the monk parakeet (Myiopsitta monachus). Location Native range in Southern America and global invasions predominantly in North/Central America and Europe. Methods We tested the ability of models built under 15 scenarios (five sets of calibration records and three settings that varied the level of model complexity) to predict spatially independent evaluation data in the invaded range (in effect, testing the models under spatial transfer). We quantified performance with measures of discriminatory ability and overfitting based on area under the receiver operating characteristic curve (AUC) and the omission error rate. We estimated null distributions of these measures and calculated effect size and significance. We determined how these estimates varied across modelling scenarios, comparing with two tests existing in the literature. Results Performance varied starkly across modelling scenarios. As expected, the measures of overfitting agreed with each other and provided different information than that of discriminatory ability. However, high performance per se did not show strong association with high effect size and significance. Main Conclusions Ecological niche models should be assessed with measures of effect size and significance based on appropriate null distributions, in contrast to several approaches existing in the literature. The proposed approach using independent evaluation data, implemented with our accompanying code and R package, allows such estimates for either the same or a different region/time period, and it merits use and continued development.

The Prognostic Nutritional Index (PNI) is widely recognized as a screening tool for nutrition. We retrospectively examined the impact of PNI in patients with chronic liver disease (CLD, n = 319, median age = 71 years, 153 hepatocellular carcinoma (HCC) patients) as an observational study. Factors associated with PNI < 40 were also examined. The PNI correlated well with the albumin–bilirubin (ALBI) score and ALBI grade. The 1-year cumulative overall survival rates in patients with PNI ≥ 40 (n = 225) and PNI < 40 (n = 94) were 93.2% and 65.5%, respectively (p < 0.0001). In patients with (p < 0.0001) and without (p < 0.0001) HCC, similar tendencies were found. In the multivariate analysis, hemoglobin (p = 0.00178), the presence of HCC (p = 0.0426), and ALBI score (p < 0.0001) were independent factors linked to PNI < 40. Receiver operating characteristic (ROC) curve analysis based on survival for the PNI yielded an area under the ROC curve of 0.79, with sensitivity of 0.80, specificity of 0.70, and an optimal cutoff point of 42.35. In conclusion, PNI can be a predictor of nutritional status in CLD patients. A PNI of <40 can be useful in predicting the prognosis of patients with CLD.