Robust flight navigation out of distribution with liquid neural networks.

Autonomous robots can learn to perform visual navigation tasks from offline human demonstrations and generalize well to online and unseen scenarios within the same environment they have been trained on. It is challenging for these agents to take a step further and robustly generalize to new environments with drastic scenery changes that they have never encountered. Here, we present a method to create robust flight navigation agents that successfully perform vision-based fly-to-target tasks beyond their training environment under drastic distribution shifts. To this end, we designed an imitation learning framework using liquid neural networks, a brain-inspired class of continuous-time neural models that are causal and adapt to changing conditions. We observed that liquid agents learn to distill the task they are given from visual inputs and drop irrelevant features. Thus, their learned navigation skills transferred to new environments. When compared with several other state-of-the-art deep agents, experiments showed that this level of robustness in decision-making is exclusive to liquid networks, both in their differential equation and closed-form representations.

Identifying and Mitigating Potential Biases in Predicting Drug Approvals.

INTRODUCTION: Machine learning models are increasingly applied to predict the drug development outcomes based on intermediary clinical trial results. A key challenge to this task is to address various forms of bias in the historical drug approval data. OBJECTIVE: We aimed to identify and mitigate the bias in drug approval predictions and quantify the impacts of debiasing in terms of financial value and drug safety. METHODS: We instantiated the Debiasing Variational Autoencoder, the state-of-the-art model for automated debiasing. We trained and evaluated the model on the Citeline dataset provided by Informa Pharma Intelligence to predict the final drug development outcome from phase II trial results. RESULTS: The debiased Debiasing Variational Autoencoder model achieved better performance (measured by the [Formula: see text] score 0.48) in predicting the drug development outcomes than its un-debiased baseline ([Formula: see text] score 0.25). It had a much higher true-positive rate than baseline (60% vs 15%), while its true-negative rate was slightly lower (88% vs 99%). The Debiasing Variational Autoencoder distinguished between drugs developed by large pharmaceutical firms and those by small biotech companies. The model prediction is strongly influenced by multiple factors such as prior approval of the drug for another indication, whether the trial meets the positive/negative endpoints, and the year when the trial is completed. We estimate that the debiased model generates financial value for the drug developer in six major therapeutic areas, with a range of US$763-1,365 million. CONCLUSIONS: Our analysis shows that debiasing improves the financial efficiency of late-stage drug development. From the pharmacovigilance perspective, the debiased model is more likely to identify drugs that are both safe and effective. Meanwhile, it may predict a higher probability of success for drugs with potential adverse effects (because of its lower true-negative rate), thus it must be used with caution to predict the development outcomes of drug candidates currently in the pipeline.

Evidential Deep Learning for Guided Molecular Property Prediction and Discovery.

While neural networks achieve state-of-the-art performance for many molecular modeling and structure-property prediction tasks, these models can struggle with generalization to out-of-domain examples, exhibit poor sample efficiency, and produce uncalibrated predictions. In this paper, we leverage advances in evidential deep learning to demonstrate a new approach to uncertainty quantification for neural network-based molecular structure-property prediction at no additional computational cost. We develop both evidential 2D message passing neural networks and evidential 3D atomistic neural networks and apply these networks across a range of different tasks. We demonstrate that evidential uncertainties enable (1) calibrated predictions where uncertainty correlates with error, (2) sample-efficient training through uncertainty-guided active learning, and (3) improved experimental validation rates in a retrospective virtual screening campaign. Our results suggest that evidential deep learning can provide an efficient means of uncertainty quantification useful for molecular property prediction, discovery, and design tasks in the chemical and physical sciences.

Results using the AO hook plate for dislocations of the acromioclavicular joint.

The AO Hook plate has been used for stabilization of acromioclavicular joint dislocations. We present our experience of this newly introduced device in a general hospital, since there are not many papers in the literature reporting on this. A total of 16 patients were treated with the AO Hook plate between November 2001 and November 2003 at Princess Alexandra Hospital in Harlow, UK. For functional assessment 6 months after removal of the plate, the constant score and the pain visual analogue score were used. The pain visual analogue score ranged from 0 to 6 (mean: 0.87) and the constant score ranged from 78 to 100 (mean: 96.4). In one instance, a patient developed instability after removal of the plate. The use of this device results in excellent functional outcome for the treatment of acromioclavicular joint dislocations.

Fixation of proximal humerus fractures using the PHILOS plate: early experience.

UNLABELLED: Proximal humeral fractures, especially in elderly patients, remain a challenging problem for the surgeon because the complication rate for these fractures still remains high. The internal locked system (PHILOS) plate is a new device used for proximal humerus fracture fixation is designed to decrease the high complication rate. We prospectively evaluated our early experience using this system. Twenty patients with fractures of the proximal humerus were treated with a PHILOS plate from September 2001 to January 2004 at Princess Alexandra Hospital in Harlow, UK. Functional assessment was done using the Constant shoulder score. Two patients who had brachial plexus injury were evaluated only with the visual analogue score because we thought that the Constant objective assessment would be unreliable. Complications were monitored. The mean Constant score was found to be 76.1% (range, 30-100%). The preliminary results seem to be satisfactory. According to our experience, the plate design provides stable fixation with a good functional outcome and eliminates most hardware problems such as failure and impingement syndrome. The PHILOS plate is suitable for the majority of fractures providing that the correct surgical technique is used. LEVEL OF EVIDENCE: Therapeutic study, level IV (case series). See the Guidelines for Authors for a complete description of levels of evidence.