Accuracy Feedback Research Articles

Design and implementation of positioning and control systems for capsule endoscopes

Wireless capsule endoscope (WCE) is a new diagnostic device that can be utilised for evaluating the whole digestive tract if effectively actuated. In this research, a new one degree of freedom (1DOF) actuation system based on a magnetic levitation concept is proposed for the capsule endoscope navigation system. The proposed system is used to move a permanent magnet inserted in WCE body by an arrangement of the current controlled electromagnetic actuator placed on a 3DOF movable frame. The aim of this study is to design a proportional–integral–derivative (PID) controller to suspend the capsule and keep it at a demand test region. DC and AC magnetic field-based positioning systems using the Hall effect sensor and the coil sensor, respectively are used to provide the controller by capsule position feedback. Improvement of the position feedback accuracy based on AC magnetic field using discrete Fourier transform is presented. Realistic simulation design of the system is implemented using Matlab/Simulink environment to validate the PID controller. The navigation scheme is implemented practically utilszing digital signal processor to verify the effectiveness of the controller. Finally, simulation and experimental results of the capsule navigation system are presented to show the performance of the proposed controller.

The ethics of online screening for mental health in South Africa: A systematic review

Mental health care in many low to middle-income countries like South Africa is often under resourced. Many individuals in these countries have access to the internet and often use this as a first source to search for information. Hence the possibility for online screening of mental health holds the promise of early detection and intervention. However, there are no clear guidelines for the ethics of online screening for mental health. This study aims to explore the literature in an attempt to explore existing guidelines on online screening with a view to formulating recommendations for a more universally applicable standard for online screening of mental health. A systematic review method was used to explore the literature. Articles were obtained from six electronic databases from 1970 to 2019. Through the use of the PRISMA method 12 articles met the inclusion criteria. There was only one article that made specific reference to ethical considerations regarding online screening instruments, whilst the remaining 11 made reference to some pertinent ethical considerations for online screening of mental health. Using a thematic analysis, six core themes were identified across the articles which covered online screening, namely modes of tests, psychometric properties of the tool, issues of consent, accuracy of results, feedback and test security. These were aspects that needed to be addressed over those already in place for pen-and-paper screening tools. The results are used to provide recommendations for ethical guidelines for the online screening of mental health.

DS-NLCsiNet: Exploiting Non-Local Neural Networks for Massive MIMO CSI Feedback

Channel state information (CSI) feedback plays an important part in frequency division duplex (FDD) massive multiple-input multiple-output (MIMO) systems. However, it is still facing many challenges, e.g., excessive feedback overhead, low feedback accuracy and a large number of training parameters. In this letter, to address these practical concerns, we propose a deep learning (DL)-based CSI feedback scheme, named DS-NLCsiNet. By taking advantage of non-local blocks, DS-NLCsiNet can capture long-range dependencies efficiently. In addition, dense connectivity is adopted to strengthen the feature refinement module. Simulation results demonstrate that DS-NLCsiNet achieves higher CSI feedback accuracy and better reconstruction quality for the same compression ratio, when compared to state-of-the-art compression schemes.

Maxillofacial surgical simulation system with haptic feedback

<p style='text-indent:20px;'>Due to the complexity of the maxillofacial surgery, the novice should be sufficiently trained before one is qualified to carry on the surgery. To reduce the training costs and improve the training efficiency, a virtual mandible surgical system with haptic feedback is proposed. This surgical simulation system offers users the haptic feedback while simulating maxillofacial surgery. An integrated model is introduced to optimize the system simulation process, which includes force output to a six-degree-of-freedom haptic device. Based on the anatomy structure of the bone tissue, a two-layer mechanism model is designed to balance the requirement of real-time response and the force feedback accuracy. Collision detection, force rendering, and grinding function are studied to simulate some essential operations: open reduction, osteotomy, and palate fixation. The proposed simulation platform can assist in the training and planning of these oral and maxillofacial surgeries. The fast response feature enables surgeons to design a patient-specific guide plate in real-time. Ten stomatology surgeons evaluated this surgical simulation system from the following four indexes: the level of immersion, user-friendliness, stability, and the effect of surgical training. The evaluation score is eight out of ten.</p>

Respiratory-Swallow Training Methods: Accuracy of Automated Detection of Swallow Onset, Respiratory Phase, Lung Volume at Swallow Onset, and Real-Time Performance Feedback Tested in Healthy Adults.

Background Preliminary studies have shown that respiratory-swallow training (RST) is a successful treatment for oropharyngeal head and neck cancer patients with refractory dysphagia. Refining the RST protocol with automated analysis software to provide real-time performance feedback has the potential to improve accessibility, reproducibility, and translation to diverse clinical settings. Method An automated software program for data acquisition and analysis developed to detect swallows, determine respiratory phase, calculate lung volume at the onset of the swallow, and provide real-time performance feedback was tested for feasibility in a small cohort of healthy adults. Outcome Measures Percent difference in swallow detection and accuracy of real-time performance feedback of respiratory phase and lung volume at swallowing onset between the automated software and the manual gold standard method were determined. Results The automated software program accurately detected the onset of the swallow on 91% of the swallows completed during the training trials. Feedback of respiratory phase and lung volume was accurate on 94% of the trials in which the swallow was accurately detected. Conclusions This novel, automated, and real-time RST software successfully detected the onset of the swallow, respiratory phase, and lung volume at swallow onset and provided appropriate real-time performance feedback with a high degree of accuracy in healthy adults. The software has the potential to improve the accessibility, efficiency, and translation of RST to diverse patient populations.

Overprecision increases subsequent surprise.

Overconfident people should be surprised that they are so often wrong. Are they? Three studies examined the relationship between confidence and surprise in order to shed light on the psychology of overprecision in judgment. Participants reported ex-ante confidence in their beliefs, and after receiving accuracy feedback, they then reported ex-post surprise. Results show that more ex-ante confidence produces less ex-post surprise for correct answers; this relationship reverses for incorrect answers. However, this sensible pattern only holds for some measures of confidence; it fails for confidence-interval measures. The results can help explain the robust durability of overprecision in judgment.

Using EEM fluorescence to characterize the membrane integrity of membrane bioreactor (MBR)

In this study, a novel method for EEM fluorescence spectra examination of the membrane integrity in MBRs was investigated. Dissolved organic matter (DOM) was measured to determine the performance of membrane retention. The sensitivity of membrane integrity tests which reported as a log reduction value (LRV) was recalculated. The results show that, two protein-like species, one humic-like species and one fulvic-like species were found in MBRs using parallel factor analysis (PARAFAC). When the membrane pore size was 0.2 μm, the majority of macromolecular protein-like species could be effectively rejected. The fluorescence regional integration (FRI) results for the tryptophan species and the aromatic protein species showed a positive correlation with the different breakage rates of fibers, the correlation coefficients were 0.980 and 0.987, respectively. In comparison with particle counting detection, EEM fluorescence spectral analysis exhibits good practicability with regards to response time and feedback accuracy. With an LRV of 4.26, it can detect the water quality and membrane integrity without having to add other substances as tracers. In addition to expanding the membrane integrity technologies, this study also provides a method for real-time membrane filtration detection and water quality analysis in MBRs.

Deep learning reservoir porosity prediction based on multilayer long short-term memory network

The cost of obtaining a complete porosity value using traditional coring methods is relatively high, and as the drilling depth increases, the difficulty of obtaining the porosity value also increases. Nowadays, the prediction of fine reservoir parameters for oil and gas exploration is becoming more and more important. Therefore, high-efficiency and low-cost prediction of porosity based on logging data is necessary. We have developed a machine-learning method based on the traditional long short-term memory (LSTM) model, called multilayer LSTM (MLSTM), to perform the porosity prediction task. We used three different wells in a block in southern China for the prediction task, including a training well and two test wells. One test well has the same logging data type as the training well, whereas the other test well differs from the training well in the logging depth and parameter types. Two different types of test data sets are used to detect the generalization ability of the network. A set of data was used to train the MLSTM network, and the hyperparameters of the network were adjusted through experimental accuracy feedback. We also tested the performance of the network using two sets of log data from different regions, including generalization and sensitivity of the network. During the training phase of the porosity prediction model, the developed MLSTM establishes a minimized objective function, uses the Adam optimization algorithm to update the weight of the network, and adjusts the network hyperparameters to select the best target according to the feedback of the network accuracy. Compared with conventional sequence neural networks, such as the gated recurrent unit and recurrent neural network, the logging data experiments show that MLSTM has better robustness and accuracy in depth sequence prediction. Especially, the porosity value at the depth inflection point can be better predicted when the trend of the depth sequence was predicted. This framework is expected to reduce the porosity prediction errors when data are insufficient and log depths are different.

Exploring Context-Free Languages via Planning: The Case for Automating Machine Learning

While Machine Learning has achieved considerable success in recent years, this success crucially relies on human experts to select appropriate features, workflows, algorithms with their hyper-parameters, etc. Automating the role of the human expert has seen some attention from the Machine Learning community, with a dedicated workshop running since 2014 at one of the top Machine Learning conferences. % In this work, we propose to exploit multiple AI Planning tools for automating the human expert, generating Machine Learning pipelines automatically. We start from a knowledge about possible valid pipelines encoded as Context-Free Grammar, translate the problem of generating the corresponding language into Hierarchical Task Network (HTN) Planning model, further translate the HTN Planning model into a classical planning model. We use existing planners to produce multiple plans for the classical planning task, translate these plans into Machine Learning pipelines, train and evaluate these pipelines. Based on pipelines' accuracy feedback we update the classical planning model to improve the quality of pipelines obtained in next iterations. Using planning tools allows us to exploit the flexibility of model update instead of solution modification. We present an application that helps users to focus pipelines' exploration process by allowing to encode additional constraints on desired pipelines. Our experimental evaluation shows the feasibility of using planning techniques in this context.

0306 Corticotropin-Releasing Hormone Receptor 1 Gene Polymorphism Modulates Cognitive Flexibility Following Acute Stress and Total Sleep Deprivation

Abstract Introduction Dynamic decision processes requiring flexible updating of information are impaired by stress and sleep loss, both of which activate the hypothalamic-pituitary-adrenal (HPA) stress response. Corticotropin-releasing hormone (CRH) initiates the HPA pathway. The CRH receptor (CRHR1) gene contains a single nucleotide polymorphism that modulates this response. We investigated whether cognitive flexibility is affected by CRHR1 polymorphism following a night of acute stress and total sleep deprivation (TSD). Methods N=46 healthy, young adults (21.8±3.4y; 21 females) participated in an in-laboratory 31h sleep deprivation study. Beginning at 19:30 until 07:30, the Maastricht Acute Stress Test (MAST) was administered every 4h. The MAST alternates a cold pressor task with an oral subtraction task five times in a single bout. At 29h wakefulness, subjects performed a novel go/no-go reversal learning task. Stimulus-response rules were presented at the beginning of the task, and subjects were asked to either respond or withhold a response to the presented stimuli while receiving accuracy feedback. Halfway through the task, the stimulus-response rules were reversed. Performance was assessed by discriminability index (d’), hit rate (HR), and false alarm rate (FAR). Saliva samples were collected immediately prior, immediately after, and 30min after each MAST and assayed for cortisol. One saliva sample from each subject was assayed for CRHR1 genotype. Results CRHR1 genotypes were in Hardy-Weinberg equilibrium (χ 2=2.97, p=0.08). Mixed effects ANOVA with fixed effects of CRHR1 genotype, pre/post-reversal, and their interaction found a significant CRHR1 by reversal interaction for d’ (F2,319=3.88, p=0.022) and HR (F2,319=3.16, p=0.044) following a night of stress and TSD. No such interaction was found at well-rested baseline (d’: F2,319=2.51, p=0.083; HR: F2,319=1.55, p=0.213). Subjects homozygous for the T allele had higher mean post-MAST cortisol levels (0.40±0.06 µg/dL) with better pre-reversal performance, but worse post-reversal performance compared to heterozygous and homozygous G allele carriers. Conclusion CRHR1 genotype modulates dynamic decision making following a night of acute stress and TSD. A higher cortisol stress response (T/T genotype) is beneficial to maintaining task relevant information (stability), but significantly impairs the ability to update task-relevant information following a change in situational demands (flexibility). Support CDMRP grant W81XWH-17-C-0088

Analysis of Surgical Forces Required to Gain Access Using a Probe for Minimally Invasive Spine Surgery via Cadaveric-Based Experiments Towards Use in Training Simulators.

Virtual Reality haptic-based surgical simulators for training purposes have recently been receiving increased traction within the medical field. However, its future adoption is contingent on the accuracy and reliability of the haptic feedback. This study describes and analyzes the implementation of a set of haptic-tailored experiments to extract the force feedback of a medical probe used in minimally invasive spinal lumbar interbody fusion surgeries. Experiments to extract linear, lateral and rotational insertion, relaxation and extraction of the tool within the spinal muscles, intervertebral discs and lumbar nerve on two cadaveric torsos were conducted. Notably, mean force-displacement and torque-angular displacement curves describing the different tool-tissue responses were reported with a maximum force of 6.87 (±1.79) N at 40 mm in the muscle and an initial rupture force through the Annulus Fibrosis of 20.550 (±7.841) N at 6.441 mm in the L4/L5 disc. The analysis showed that increasing the velocity of the probe slightly reduced and delayed depth of the muscle punctures but significantly lowered the force reduction due to relaxation. Decreasing probe depth resulted with a reduction to the force relaxation drop. However, varying the puncturing angle of attack resulted with a significant effect on increasing force intensities. Finally, not resecting the thoracolumbar fascia prior to puncturing the muscle resulted with a significant increase in the force intensities. These results present a complete characterization of the input required for probe access for spinal surgeries to provide an accurate haptic response in training simulators.

Behind the therapy door: what is “usual care” aphasia therapy in acute stroke management?

ABSTRACT Background: Usual care is the term used to describe everyday practice in the management of a client within a profession. The knowledge of the tasks used in therapy and key therapeutic processes used within these treatments, provides critical information about if and how the therapy works. The Very Early Rehabilitation in SpEech Randomised Controlled Trial (VERSE RCT) had three arms with therapists within the intensive Usual Care-Plus arm (UC-Plus) providing daily direct aphasia therapy at their discretion for 20 sessions. Aims: To describe usual care aphasia treatment provided in the Usual Care-Plus arm of VERSE RCT. Methods and Procedures: One in four intensive Usual Care-Plus treatment sessions were video-recorded (N = 187) within the main trial. Twenty-five of these (13%) were transcribed, coded, and analysed for therapeutic inputs to describe usual care aphasia therapy using the Template for Intervention Description and Replication (TIDieR) checklist as an overriding framework. Outcomes and Results: Therapy predominantly took place in an inpatient setting (52%) with an average session duration of 51 minutes (SD 7.8). Across the sessions, 96 different tasks were used and 57% of these focused on verbal expression at the single word level. Visual materials were most frequently used compared to the use of technology during sessions. Therapists (n = 16) did the majority of the talking during sessions and most frequently provided models as cues or problem-solving accuracy feedback. Models (55%), sentence completion (51%), and orthographic cues (44%) were the most successful at eliciting the target response. Conclusions: Considerable variability in task selection was seen in the sample which may be a hallmark of usual care. Therapists may have a preference for single word tasks and appear to produce the majority of verbal utterances during sessions, potentially creating an unequal communication environment. This study provided a comprehensive description from the Usual Care-Plus data of the VERSE RCT and may establish a baseline of therapy type for future research.

Short‐term load forecasting utilizing wavelet transform and time series considering accuracy feedback

<scp>Short‐term</scp> load forecasting utilizing wavelet transform and time series considering <scp>accuracy feedback</scp>

“You're wrong!”: The impact of accuracy feedback on the bat-and-ball problem

The popular bat-and-ball problem is a relatively simple math riddle on which people are easily biased by intuitive or heuristic thinking. In two studies we tested the impact of a simple but somewhat neglected manipulation – the impact of minimal accuracy feedback – on bat-and-ball performance. Participants solved a total of 15 standard and 15 control versions of the bat-and-ball problem in three consecutive blocks. Half of the participants received accuracy feedback in the intermediate block. Results of both studies indicated that the feedback had, on average, no significant effect on bat-and-ball accuracy over and above mere repeated presentation. We did observe a consistent improvement for a small number of individual participants. Explorative analyses indicated that this improved group showed a more pronounced conflict detection effect (i.e., latency increase) at the pretest and took more deliberation time after receiving the negative feedback compared to the unimproved group.

Student engagement with automated written corrective feedback (AWCF) provided by Grammarly: A multiple case study

Abstract Despite the increased use of automated writing evaluation (AWE) systems and similar programs for assessment purposes in second language (L2) writing classrooms, research on student engagement with automated feedback is scarce. This naturalistic case study explored two ESL college students’ engagement with automated written corrective feedback (AWCF) provided by Grammarly when revising a final draft. Following previous research, student engagement was operationalized using three interconnected dimensions: behavioral, cognitive, and affective. Behavioral engagement was explored through the analysis of QuickTime-based screencasts of students’ Grammarly usage. Cognitive and affective engagement were measured through the analysis of students’ comments during stimulated recall of the aforementioned screencasts and semi-structured interview. Findings suggest that students had different levels of engagement with AWCF. One showed greater cognitive engagement through his questioning of AWCF. However, he did little to verify the accuracy of feedback which resulted in moderate changes to his draft. The other’s overreliance on AWCF indicated more limited cognitive engagement which led to feedback’s blind acceptance. Nevertheless, this also resulted in moderate changes to her draft.

Training attenuates the influence of sensory uncertainty on confidence estimation.

Confidence is typically correlated with perceptual sensitivity, but these measures are dissociable. For example, confidence judgements are disproportionately affected by the variability of sensory signals. Here, in a preregistered study we investigate whether this signal variability effect on confidence can be attenuated with training. Participants completed five sessions where they viewed pairs of motion kinematograms and performed comparison judgements on global motion direction, followed by confidence ratings. In pre- and post-training sessions, the range of direction signals within each stimulus was manipulated across four levels. Participants were assigned to one of three training groups, differing as to whether signal range was varied or fixed during training, and whether or not trial-by-trial accuracy feedback was provided. The effect of signal range on confidence was reduced following training, but this result was invariant across the training groups, and did not translate to improved metacognitive insight. These results suggest that the influence of suboptimal heuristics on confidence can be mitigated through experience, but this shift in confidence estimation remains coarse, without improving insight into confidence estimation at the level of individual decisions.

Examining the Effects of Feedback Accuracy and Timing on Skill Acquisition

ABSTRACT Feedback involves providing information about past performance and indicating how to improve future performance. Although the literature contains numerous examples of feedback as an effective method for improving performance across a range of organizational settings, much remains unknown about the specifics of how feedback acts to change behavior. This study evaluated the combined effects of feedback accuracy (100%, 80%, 20%) and timing (feedback following each trial or after a block of 25 trials) on skill acquisition in undergraduate students when presented with a computerized match-to-sample task that required participants to learn the names of shapes. Results reveal that feedback accuracy had a significantly greater effect on performance than the timing of the feedback.

Precise verbal feedback may expedite the attainment of standing long jump accuracy in women

The purpose of this study was to determine whether the precision of verbal feedback affected the accuracy of a standing long jump performance. Sixty-eight female participants (M age=21.44±1.01) were randomly assigned to two experimental groups. In total, each subject performed nine jumps at three distances, i.e. 50 cm (3 trials), 75 cm (3 trials) and 100 cm (3 trials). Group 2 performed jumps in a reverse order. During the task execution, the study participants received verbal feedback at different levels of precision (no feedback; information that the jump was either: too far, close, or correct; information about the distance in centimetres). The results of statistical analysis indicated that the speed of changes in jump accuracy depended on the precision of verbal feedback. The difference between an actual jump result and a target result increased proportionally with jump length. With a large number of trials, verbal feedback accuracy does not affect significantly the precision of women’s long jump performance. However, when the high accuracy of task performance needs to be obtained quickly, feedback ought to define the exact difference between the target and the actual jump results. The current findings also indicate that providing no feedback does not contribute to improving motor task performance accuracy.

A Novel CSI Feedback Approach for Massive MIMO Using LSTM-Attention CNN

In this paper, a novel mechanism is studied to improve the performance of the channel state information (CSI) feedback in massive multiple-input multiple-output (MIMO) systems. The proposed mechanism encompasses convolutional neural network (CNN)-based CSI compression and reconstruction structure. In this structure, the long-short term memory (LSTM) is adopted to learn temporal correlation of channels, and then, an attention mechanism is developed to perceive local information and automatically weight feature information. In addition, the CNN framework is further adjusted to reduce the number of training parameters and accelerate CSI recovery. The CNN structure with optimal training parameters can be achieved via offline iterative training and learning based on various training datasets. Comparative experimental studies demonstrate the effectiveness of the proposed approach that the trained CNN can obtain the higher feedback accuracy and better system performance in massive MIMO CSI online feedback reconstruction. Moreover, the proposed scheme in the less parameters-based neural network owns a higher performance with lower computational complexity compared to the conventional algorithms.

The Effect of Unfocused Direct and Indirect Written Corrective Feedback on Rewritten Texts and New Texts: Looking into Feedback for Accuracy and Feedback for Acquisition

AbstractThis study attempts to add new empirical evidence on the language learning (operationalized and measured in terms of several dimensions of accuracy) that may derive from accessing and processing written corrective feedback (WCF) on one's own writing. The research questions examined potential interactions between type of WCF (direct vs. indirect), type of errors (grammar vs. nongrammar), and the perspective of feedback (accuracy vs. acquisition) in a single research design. To this end, 46 English majors at a Spanish university participated in a pretest–posttest design, with 2 intervention groups (those receiving direct or indirect WCF and asked to process it via written languaging) and a control group (who wrote and rewrote their texts without the help of WCF but also engaged in languaging). The analyses conducted show (a) limited appropriation of the WCF received, (b) positive short‐term and longer term benefits resulting from the combined effect of WCF and written languaging, and (c) differential effects of type of WCF on error types. These findings are discussed from the perspective of current theory and research on the manner in which WCF may contribute to language learning in terms of the dual distinction between feedback for accuracy and feedback for acquisition.