Background: Kinesiophobia is a psycho-cognitive factor that hampers recovery after orthopedic surgery. No evidence exists on the influence of kinesiophobia on the short-term recovery of function in patients with knee replacement (KR). Therefore, the aim of the present study is to investigate the impact of kinesiophobia on short-term patient-reported outcomes (PROMs) and performance-based measures (PBMs). Methods: Forty-three KR patients filled in the Tampa scale for kinesiophobia (TSK) at time of discharge. Patients with TSK ≥ 37 were allocated to the kinesiophobia group (n = 24), others to the no-kinesiophobia group (n = 19). Patients were asked to complete PROMs and to execute PBMs at discharge and at 6-weeks follow-up. An independent samples t-test was used to compare group differences for PROMs and PBMs at both measurement sessions. Multiple linear regression analysis models were used to model PBM outcomes from age, pain and TSK scores. Results: Significant differences were observed between groups for PROMs and PBMs. Kinesiophobia significantly contributed to the reduced functional outcomes. Conclusion: At discharge from the hospital, 55.8% of KR patients demonstrated high levels of kinesiophobia (TSK ≥ 37). This may negatively influence short-term recovery of these patients, by putting them at higher risk for falling and reduced functionality.

Effective uncertainty estimation is becoming increasingly attractive for enhancing the reliability of neural networks. This work presents a novel approach, termed Credal-Set Interval Neural Networks (CreINNs), for classification. CreINNs retain the fundamental structure of traditional Interval Neural Networks, capturing weight uncertainty through deterministic intervals. CreINNs are designed to predict an upper and a lower probability bound for each class, rather than a single probability value. The probability intervals can define a credal set, facilitating estimating different types of uncertainties associated with predictions. Experiments on standard multiclass and binary classification tasks demonstrate that the proposed CreINNs can achieve superior or comparable quality of uncertainty estimation compared to variational Bayesian Neural Networks (BNNs) and Deep Ensembles. Furthermore, CreINNs significantly reduce the computational complexity of variational BNNs during inference. Moreover, the effective uncertainty quantification of CreINNs is also verified when the input data are intervals.

Effectively estimating the uncertainty attached to neural network predictions thus becomes essential to improve robustness, reliability, and trustworthiness. This paper provides an overview of various methodologies for representing, quantifying, and distinguishing two major types of uncertainties (namely, ‘aleatoric’ and ‘epistemic’ uncertainty) in neural networks. The review covers classical probabilistic techniques such as Bayesian neural networks and deep ensembles, methods from generalized probability that leverage uncertainty representations such as Dirichlet distributions, belief functions, random sets, probability intervals, and credal sets, among others. Additionally, interval-based approaches employing interval models are also examined. We discuss the strengths and limitations of various methodologies and identify promising research directions for potential future exploration.

Traditional instrumented seat posts determine context-induced seat loads to analyze damping properties. This paper presents an enhanced instrumented seat post able to measure all six load components to resolve user-induced seat loads. User-induced cycling loads consist of all loads the user applies to the bicycle during cycling and is measured at the steer stem, the seat post, and the pedals. Seat loads are essentially uncharted territory, as most studies only address pedal loading to study cycling technique. In this paper, a conventional seat post is redesigned by equipping it with a u-shaped component and strain gauges. The instrumented seat post is straightforward thanks to (i) the simple design, (ii) the gravitational calibration method, and (iii) the permitted clearance on the strain gauge alignment. Analyzing mean seat loading in function of the pedal cycle can provide extra insights into cycling technique and the related injuries. It is an interesting addition to the universally adopted method of utilizing singular pedal loads.

This paper considers a constrained optimisation problem under uncertainty with at least one element modelled as a probability box uncertainty. The uncertainty is expressed in the coefficient matrices of constraints and/or coefficients of goal function. In our previous work, such problems were studied under interval, fuzzy sets, and ε-contamination uncertainty models. Our aim here is to give theoretical solutions to the problem under more advanced and informative (p-box) uncertainty model and generalise the approach to calculate the theoretical solutions for linear programming problems. The approach is to convert the optimisation problem under uncertainty to a decision problem using imprecise decision theory where the uncertainty is eliminated. We investigate what theoretical results can be obtained for probability box type of uncertainty model and compare them to classical cases for two different optimality criteria: maximinity and maximality.