RESUMO
BACKGROUND: Computational methods of predicting protein stability changes upon missense mutations are invaluable tools in high-throughput studies involving a large number of protein variants. However, they are limited by a wide variation in accuracy and difficulty of assessing prediction uncertainty. Using a popular computational tool, FoldX, we develop a statistical framework that quantifies the uncertainty of predicted changes in protein stability. RESULTS: We show that multiple linear regression models can be used to quantify the uncertainty associated with FoldX prediction for individual mutations. Comparing the performance among models with varying degrees of complexity, we find that the model precision improves significantly when we utilize molecular dynamics simulation as part of the FoldX workflow. Based on the model that incorporates information from molecular dynamics, biochemical properties, as well as FoldX energy terms, we can generally expect upper bounds on the uncertainty of folding stability predictions of ± 2.9 kcal/mol and ± 3.5 kcal/mol for binding stability predictions. The uncertainty for individual mutations varies; our model estimates it using FoldX energy terms, biochemical properties of the mutated residue, as well as the variability among snapshots from molecular dynamics simulation. CONCLUSIONS: Using a linear regression framework, we construct models to predict the uncertainty associated with FoldX prediction of stability changes upon mutation. This technique is straightforward and can be extended to other computational methods as well.
Assuntos
Mutação de Sentido Incorreto , Dobramento de Proteína , Incerteza , Mutação , Simulação de Dinâmica Molecular , Estabilidade Proteica , Ligação ProteicaRESUMO
Despite the advancement in deep learning-based semantic segmentation methods, which have achieved accuracy levels of field experts in many computer vision applications, the same general approaches may frequently fail in 3D medical image segmentation due to complex tissue structures, noisy acquisition, disease-related pathologies, as well as the lack of sufficiently large datasets with associated annotations. For expeditious diagnosis and quantitative image analysis in large-scale clinical trials, there is a compelling need to predict segmentation quality without ground truth. In this paper, we propose a deep learning framework to locate erroneous regions on the boundary surfaces of segmented objects for quality control and assessment of segmentation. A Convolutional Neural Network (CNN) is explored to learn the boundary related image features of multi-objects that can be used to identify location-specific inaccurate segmentation. The predicted error locations can facilitate efficient user interaction for interactive image segmentation (IIS). We evaluated the proposed method on two data sets: Osteoarthritis Initiative (OAI) 3D knee MRI and 3D calf muscle MRI. The average sensitivity scores of 0.95 and 0.92, and the average positive predictive values of 0.78 and 0.91 were achieved, respectively, for erroneous surface region detection of knee cartilage segmentation and calf muscle segmentation. Our experiment demonstrated promising performance of the proposed method for segmentation quality assessment by automated detection of erroneous surface regions in medical images.
Assuntos
Articulação do Joelho , Osteoartrite , Humanos , Redes Neurais de Computação , Controle de Qualidade , SemânticaRESUMO
To better solve the problem of thermal error of computerized numerical control machining equipment (CNCME), a thermal error prediction model based on the sparrow search algorithm and long short-term memory neural network (SSA-LSTMNN) is proposed. Firstly, the Fuzzy C-means clustering algorithm (FCMCA) is used to screen the key temperature-sensitive points of the CNCME. Secondly, by taking the temperature rise data of key temperature-sensitive points as input and the corresponding time thermal error data as output, we established the SSA-LSTMNN thermal error prediction model. The SSA is used to optimize the parameters of LSTMNN and make its performance play the best. Taking the VMC1060 vertical machining center as the research object, we carried out the experiment. Finally, the prediction effect of the proposed model is compared with the article swarm optimized algorithm and LSTM neural network (PSOA-LSTMNN), the LSTMNN, and the traditional recurrent neural network (TRNN) model. The results show that the average values of the predicted residual fluctuations of the SSA-LSTMNN model are all more than 44% lower than those of the other three models under different operating conditions, which has a strong practicality.
RESUMO
OBJECTIVE: To identify delivery error type and predict associated error magnitude by image-based features using machine learning (ML). METHODS: In this study, a total of 40 thoracic plans (including 208 beams) were selected, and four error types with different magnitudes were introduced into the original plans, including 1) collimator misalignment (COLL), 2) monitor unit (MU) variation, 3) systematic multileaf collimator misalignment (MLCS), and 4) random MLC misalignment (MLCR). These dose distributions of portal dose predictions for the original plans were defined as the reference dose distributions (RDD), while those for the error-introduced plans were defined as the error-introduced dose distributions (EDD). Both distributions were calculated for all beams with portal dose image prediction (PDIP). Besides, 14 image-based features were extracted from RDD and EDD of portal dose predictions to obtain the feature vectors. In addition, a random forest was adopted for the multiclass classification task, and regression prediction for error magnitude. RESULTS: The top five features extracted with the highest weight included 1) the relative displacement in the x direction, 2) the ratio of the absolute minimum residual error to the maximal RDD value, 3) the product of the maximum and minimum residuals, 4) the ratio of the absolute maximum residual error to the maximal RDD value, and 5) the ratio of the absolute mean residual value to the maximal RDD value. The relative displacement in the x direction had the highest weight. The overall accuracy of the five-class classification model was 99.85% for the validation set and 99.30% for the testing set. This model could be applied to the classification of the error-free plan, COLL, MU, MLCS, and MLCR with an accuracy of 100%, 98.4%, 99.9%, 98.0%, and 98.3%, respectively. MLCR had the worst performance in error magnitude prediction (70.1-96.6%), while others had better performance in error magnitude prediction (higher than 93%). In the error magnitude prediction, the mean absolute error (MAE) between predicted error magnitude and actual error ranged from 0.03 to 0.33, with the root mean squared error (RMSE) varying from 0.17 to 0.56 for the validation set. The MAE and RMSE ranged from 0.03 to 0.50 and 0.44 to 0.59 for the test set, respectively. CONCLUSION: It could be demonstrated in this study that the image-based features extracted from RDD and EDD can be employed to identify different types of delivery errors and accurately predict error magnitude with the assistance of ML techniques. They can be used to associate traditional gamma analysis with clinically based analysis for error classification and magnitude prediction in patient-specific IMRT quality assurance.
Assuntos
Radioterapia de Intensidade Modulada , Humanos , Radioterapia de Intensidade Modulada/métodos , Planejamento da Radioterapia Assistida por Computador/métodos , Aprendizado de Máquina , Dosagem RadioterapêuticaRESUMO
BACKGROUND: Intraoperative adverse events lead to patient injury and death, and are increasing. Early warning systems (EWSs) have been used to detect patient deterioration and save lives. However, few studies have used EWSs to monitor surgical performance and caution about imminent technical errors. Previous (non-surgical) research has investigated neural activity to predict future motor errors using electroencephalography (EEG). The present proof-of-concept cohort study investigates whether EEG could predict technical errors in surgery. METHODS: In a large academic hospital, three surgical fellows performed 12 elective laparoscopic general surgeries. Audiovisual data of the operating room and the surgeon's neural activity were recorded. Technical errors and epochs of good surgical performance were coded into events. Neural activity was observed 40 s prior and 10 s after errors and good events to determine how far in advance errors were detected. A hierarchical regression model was used to account for possible clustering within surgeons. This prospective, proof-of-concept, cohort study was conducted from July to November 2021, with a pilot period from February to March 2020 used to optimize the technique of data capture and included participants who were blinded from study hypotheses. RESULTS: Forty-five technical errors, mainly due to too little force or distance (n = 39), and 27 good surgical events were coded during grasping and dissection. Neural activity representing error monitoring (p = .008) and motor uncertainty (p = .034) was detected 17 s prior to errors, but not prior to good surgical performance. CONCLUSIONS: These results show that distinct neural signatures are predictive of technical error in laparoscopic surgery. If replicated with low false-alarm rates, an EEG-based EWS of technical errors could be used to improve individualized surgical training by flagging imminent unsafe actions-before errors occur and cause patient harm.
Assuntos
Competência Clínica , Laparoscopia , Humanos , Estudos de Coortes , Estudos Prospectivos , Laparoscopia/efeitos adversos , EletroencefalografiaRESUMO
Summary: Ultrawide field fundus images could be applied in deep learning models to predict the refractive error of myopic patients. The predicted error was related to the older age and greater spherical power. Purpose: To explore the possibility of predicting the refractive error of myopic patients by applying deep learning models trained with ultrawide field (UWF) images. Methods: UWF fundus images were collected from left eyes of 987 myopia patients of Eye and ENT Hospital, Fudan University between November 2015 and January 2019. The fundus images were all captured with Optomap Daytona, a 200° UWF imaging device. Three deep learning models (ResNet-50, Inception-v3, Inception-ResNet-v2) were trained with the UWF images for predicting refractive error. 133 UWF fundus images were also collected after January 2021 as an the external validation data set. The predicted refractive error was compared with the "true value" measured by subjective refraction. Mean absolute error (MAE), mean absolute percentage error (MAPE) and coefficient (R 2) value were calculated in the test set. The Spearman rank correlation test was applied for univariate analysis and multivariate linear regression analysis on variables affecting MAE. The weighted heat map was generated by averaging the predicted weight of each pixel. Results: ResNet-50, Inception-v3 and Inception-ResNet-v2 models were trained with the UWF images for refractive error prediction with R 2 of 0.9562, 0.9555, 0.9563 and MAE of 1.72(95%CI: 1.62-1.82), 1.75(95%CI: 1.65-1.86) and 1.76(95%CI: 1.66-1.86), respectively. 29.95%, 31.47% and 29.44% of the test set were within the predictive error of 0.75D in the three models. 64.97%, 64.97%, and 64.47% was within 2.00D predictive error. The predicted MAE was related to older age (P < 0.01) and greater spherical power(P < 0.01). The optic papilla and macular region had significant predictive power in the weighted heat map. Conclusions: It was feasible to predict refractive error in myopic patients with deep learning models trained by UWF images with the accuracy to be improved.
RESUMO
The prediction of the sensory consequences of physical movements is a fundamental feature of the human brain. This function is attributed to a forward model, which generates predictions based on sensory and efferent information. The neural processes underlying such predictions have been studied using the error-related negativity (ERN) as a fronto-central event-related potential in electroencephalogram (EEG) tracings. In this experiment, 16 participants practiced a novel motor task for 4000 trials over ten sessions. Neural correlates of error processing were recorded in sessions one, five, and ten. Along with significant improvements in task performance, the ERN amplitude increased over the sessions. Simultaneously, the feedback-related negativity (FRN), a neural marker corresponding to the processing of movement-outcome feedback, attenuated with learning. The findings suggest that early in learning, the motor control system relies more on information from external feedback about terminal outcome. With increasing task performance, the forward model is able to generate more accurate outcome predictions, which, as a result, increasingly contributes to error processing. The data also suggests a complementary relationship between the ERN and the FRN over motor learning.
Assuntos
Eletroencefalografia , Potenciais Evocados , Encéfalo , Humanos , Aprendizagem , Movimento , Desempenho PsicomotorRESUMO
PURPOSE: The accuracy of the CyberKnife Synchrony Respiratory Tracking System is dependent on the breathing pattern of a patient. Therefore, the tracking error in each patient must be determined. Support vector regression (SVR) can be used to easily identify the tracking error in each patient. This study aimed to develop a system with SVR that can predict tracking error according to a patient's respiratory waveform. METHODS: Datasets of the respiratory waveforms of 93 patients were obtained. The feature variables were variation in respiration amplitude, tumor velocity, and phase shift between tumor and the chest wall, and the target variable was tracking error. A learning model was evaluated with tenfold cross-validation. We documented the difference between the predicted and actual tracking errors and assessed the correlation coefficient and coefficient of determination. RESULTS: The average difference and maximum difference between the actual and predicted tracking errors were 0.57 ± 0.63 mm and 2.1 mm, respectively. The correlation coefficient and coefficient of determination were 0.86 and 0.74, respectively. CONCLUSION: We developed a system for obtaining tracking error by using SVR. The accuracy of such a system is clinically useful. Moreover, the system can easily evaluate tracking error. We developed a system that can be used to predict the tracking error of SRTS in the CyberKnife Robotic Radiosurgery System using machine learning. The feature variables were the breathing parameters, and the target variable was the tracking error. We used support vector regression algorithm.
Assuntos
Radiocirurgia , Robótica , Algoritmos , Humanos , Respiração , Sistema RespiratórioRESUMO
Reliable uncertainty quantification for statistical models is crucial in various downstream applications, especially for drug design and discovery where mistakes may incur a large amount of cost. This topic has therefore absorbed much attention and a plethora of methods have been proposed over the past years. The approaches that have been reported so far can be mainly categorized into two classes: distance-based approaches and Bayesian approaches. Although these methods have been widely used in many scenarios and shown promising performance with their distinct superiorities, being overconfident on out-of-distribution examples still poses challenges for the deployment of these techniques in real-world applications. In this study we investigated a number of consensus strategies in order to combine both distance-based and Bayesian approaches together with post-hoc calibration for improved uncertainty quantification in QSAR (Quantitative Structure-Activity Relationship) regression modeling. We employed a set of criteria to quantitatively assess the ranking and calibration ability of these models. Experiments based on 24 bioactivity datasets were designed to make critical comparison between the model we proposed and other well-studied baseline models. Our findings indicate that the hybrid framework proposed by us can robustly enhance the model ability of ranking absolute errors. Together with post-hoc calibration on the validation set, we show that well-calibrated uncertainty quantification results can be obtained in domain shift settings. The complementarity between different methods is also conceptually analyzed.
RESUMO
There are many monitoring environments, such as railway control, in which lapses of attention can have tragic consequences. Problematically, sustained monitoring for rare targets is difficult, with more misses and longer reaction times over time. What changes in the brain underpin these 'vigilance decrements'? We designed a multiple-object monitoring (MOM) paradigm to examine how the neural representation of information varied with target frequency and time performing the task. Behavioural performance decreased over time for the rare target (monitoring) condition, but not for a frequent target (active) condition. This was mirrored in neural decoding using magnetoencephalography: coding of critical information declined more during monitoring versus active conditions along the experiment. We developed new analyses that can predict behavioural errors from the neural data more than a second before they occurred. This facilitates pre-empting behavioural errors due to lapses in attention and provides new insight into the neural correlates of vigilance decrements.
Assuntos
Atenção/fisiologia , Encéfalo/fisiologia , Tempo de Reação/fisiologia , Vigília/fisiologia , Adulto , Feminino , Humanos , Masculino , New South Wales , Adulto JovemRESUMO
Instead of a flag valid/non-valid usually proposed in the quality control (QC) processes of air quality (AQ), we proposed a method that predicts the p-value of each observation as a value between 0 and 1. We based our error predictions on three approaches: the one proposed by the Working Group on Guidance for the Demonstration of Equivalence (European Commission (2010)), the one proposed by Wager (Journal of MachineLearningResearch, 15, 1625-1651 (2014)) and the one proposed by Lu (Journal of MachineLearningResearch, 22, 1-41 (2021)). Total Error framework enables to differentiate the different errors: input, output, structural modeling and remnant. We thus theoretically described a one-site AQ prediction based on a multi-site network using Random Forest for regression in a Total Error framework. We demonstrated the methodology with a dataset of hourly nitrogen dioxide measured by a network of monitoring stations located in Oslo, Norway and implemented the error predictions for the three approaches. The results indicate that a simple one-site AQ prediction based on a multi-site network using Random Forest for regression provides moderate metrics for fixed stations. According to the diagnostic based on predictive qq-plot and among the three approaches used in this study, the approach proposed by Lu provides better error predictions. Furthermore, ensuring a high precision of the error prediction requires efforts on getting accurate input, output and prediction model and limiting our lack of knowledge about the "true" AQ phenomena. We put effort in quantifying each type of error involved in the error prediction to assess the error prediction model and further improving it in terms of performance and precision.
RESUMO
The Error-related negativity (Ne/ERN) and the feedback-related negativity (FRN), two event-related potentials in electroencephalogram tracings, have been used to examine error processing in conscious actions. In the classical terminology the Ne/ERN and the FRN are differentiated with respect to whether internal (Ne/ERN) or external (FRN) error information is processed. In motor tasks, however, errors of different types can be made: A wrong action can be selected that is not adequate to achieve the task goal (or action effect), or the correctly selected action can be mis-performed such that the task goal might be missed (movement error). Depending on the motor task and the temporal sequences of these events, internal and external error information can coincide. Hence, a clear distinction of the information source is difficult, and the classical terminology that differentiates the Ne/ERN and the FRN with respect to internal and external error information becomes ambiguous. But, a stronger focus on the characteristics of the definition of "task" and the cause of "errors", as well as on temporal characteristics of event-related potentials with respect to the task action allows separate examination of the processing of movement errors, the processing of the prediction of action effect errors, or the processing of the detection of action effect errors. The present article gives an overview of example studies investigating the Ne/ERN and the FRN in motor tasks, classifies them with respect to action effect errors or movement errors, and proposes updated terminology.
RESUMO
The mismatch negativity (MMN) is an event related brain potential (ERP) elicited by unpredicted sounds presented in a sequence of repeated auditory stimuli. The neural sources of the MMN have been previously attributed to a fronto-temporo-parietal network which crucially overlaps with the so-called auditory dorsal stream, involving inferior and middle frontal, inferior parietal, and superior and middle temporal regions. These cortical areas are structurally connected by the arcuate fasciculus (AF), a three-branch pathway supporting the feedback-feedforward loop involved in auditory-motor integration, auditory working memory, storage of acoustic templates, as well as comparison and update of those templates. Here, we characterized the individual differences in the white-matter macrostructural properties of the AF and explored their link to the electrophysiological marker of passive change detection gathered in a melodic multifeature MMN-EEG paradigm in 26 healthy young adults without musical training. Our results show that left fronto-temporal white-matter connectivity plays an important role in the pre-attentive detection of rhythm modulations within a melody. Previous studies have shown that this AF segment is also critical for language processing and learning. This strong coupling between structure and function in auditory change detection might be related to life-time linguistic (and possibly musical) exposure and experiences, as well as to timing processing specialization of the left auditory cortex. To the best of our knowledge, this is the first time in which the relationship between neurophysiological (EEG) and brain white-matter connectivity indexes using DTI-tractography are studied together. Thus, the present results, although still exploratory, add to the existing evidence on the importance of studying the constraints imposed on cognitive functions by the underlying structural connectivity.
Assuntos
Atenção/fisiologia , Córtex Cerebral/diagnóstico por imagem , Córtex Cerebral/fisiologia , Individualidade , Música/psicologia , Substância Branca/diagnóstico por imagem , Substância Branca/fisiologia , Estimulação Acústica/métodos , Eletroencefalografia/métodos , Feminino , Humanos , Imageamento por Ressonância Magnética/métodos , Masculino , Rede Nervosa/diagnóstico por imagem , Rede Nervosa/fisiologia , Adulto JovemRESUMO
Tripping hazards on the sidewalk cause many falls annually, and the inspection and repair of these hazards cost cities millions of dollars. Currently, there is not an efficient and cost-effective method to monitor the sidewalk to identify any possible tripping hazards. In this paper, a new portable device is proposed using an Intel RealSense D415 RGB-D camera to monitor the sidewalks, detect the hazards, and extract relevant features of the hazards. This paper first analyzes the effects of environmental factors contributing to the device's error and compares different regression techniques to calibrate the camera. The Gaussian Process Regression models yielded the most accurate predictions with less than 0.09 mm Mean Absolute Errors (MAEs). In the second phase, a novel segmentation algorithm is proposed that combines the edge detection and region-growing techniques to detect the true tripping hazards. Different examples are provided to visualize the output results of the proposed method.
Assuntos
Acidentes por Quedas , Algoritmos , Medição de RiscoRESUMO
Various indoor positioning methods have been developed to solve the "last mile on Earth". Ultra-wideband positioning technology stands out among all indoor positioning methods due to its unique communication mechanism and has a broad application prospect. Under non-line-of-sight (NLOS) conditions, the accuracy of this positioning method is greatly affected. Unlike traditional inspection and rejection of NLOS signals, all base stations are involved in positioning to improve positioning accuracy. In this paper, a Long Short-Term Memory (LSTM) network is used while maximizing the use of positioning equipment. The LSTM network is applied to process the raw Channel Impulse Response (CIR) to calculate the ranging error, and combined with the improved positioning algorithm to improve the positioning accuracy. It has been verified that the accuracy of the predicted ranging error is up to centimeter level. Using this prediction for the positioning algorithm, the average positioning accuracy improved by about 62%.
RESUMO
Accurate attitude and heading reference system (AHRS) play an essential role in navigation applications and human body tracking systems. Using low-cost microelectromechanical system (MEMS) inertial sensors and having accurate orientation estimation, simultaneously, needs optimum orientation methods and algorithms. The error of attitude estimation may lead to imprecise navigation and motion capture results. This paper proposed a novel intermittent calibration technique for MEMS-based AHRS using error prediction and compensation filter. The method, inspired from the recognition of gyroscope's error and by a proportional integral (PI) controller, can be regulated to increase the accuracy of the prediction. The experimentation of this study for the AHRS algorithm, aided by the proposed prediction filter, was tested with real low-cost MEMS sensors consists of accelerometer, gyroscope, and magnetometer. Eventually, the error compensation was performed by post-processing the measurements of static and dynamic tests. The experimental results present about 35% accuracy improvement in attitude estimation and demonstrate the explicit performance of proposed method.
Assuntos
Sistemas Microeletromecânicos , Algoritmos , Calibragem , Humanos , Orientação , Orientação EspacialRESUMO
In this paper, we present an architecture of a personalized glucose monitoring system (PGMS). PGMS consists of both invasive and non-invasive sensors on a single device. Initially, blood glucose is measured invasively and non-invasively, to train the machine learning models. Then, paired data and corresponding errors are divided scientifically into six different clusters based on blood glucose ranges as per the patient's diabetic conditions. Each cluster is trained to build the unique error prediction model using an adaptive boosting (AdaBoost) algorithm. Later, these error prediction models undergo personalized calibration based on the patient's characteristics. Once, the errors in predicted non-invasive values are within the acceptable error range, the device gets personalized for a patient to measure the blood glucose non-invasively. We verify PGMS on two different datasets. Performance analysis shows that the mean absolute relative difference (MARD) is reduced exceptionally to 7.3% and 7.1% for predicted values as compared to 25.4% and 18.4% for measured non-invasive glucose values. The Clarke error grid analysis (CEGA) plot for non-invasive predicted values shows 97% data in Zone A and 3% data in Zone B for dataset 1. Moreover, for dataset 2 results echoed with 98% and 2% in Zones A and B, respectively.
RESUMO
In this work, we propose an adaptive face tracking scheme that compensates for possible face tracking errors during its operation. The proposed scheme is equipped with a tracking divergence estimate, which allows to detect early and minimize the face tracking errors, so the tracked face is not missed indefinitely. When the estimated face tracking error increases, a resyncing mechanism based on Constrained Local Models (CLM) is activated to reduce the tracking errors by re-estimating the tracked facial features' locations (e.g., facial landmarks). To improve the Constrained Local Model (CLM) feature search mechanism, a Weighted-CLM (W-CLM) is proposed and used in resyncing. The performance of the proposed face tracking method is evaluated in the challenging context of driver monitoring using yawning detection and talking video datasets. Furthermore, an improvement in a yawning detection scheme is proposed. Experiments suggest that our proposed face tracking scheme can obtain a better performance than comparable state-of-the-art face tracking methods and can be successfully applied in yawning detection.
Assuntos
Face/diagnóstico por imagem , Face/fisiologia , Processamento de Imagem Assistida por Computador/métodos , Bocejo/fisiologia , Adolescente , Adulto , Algoritmos , Feminino , Humanos , Masculino , Gravação em Vídeo , Adulto JovemRESUMO
Recent studies have highlighted that the observation of hand-object interactions can influence perceptual weight judgments made by an observer. Moreover, observing salient motor errors during object lifting allows individuals to update their internal sensorimotor representation about object weight. Embodying observed visuomotor cues for the planning of a motor command further enables individuals to accurately scale their fingertip forces when subsequently lifting the same object. However, it is still unknown whether the observation of a skilled lift is equally able to mediate predictive motor control in the observer. Here, we tested this hypothesis by asking participants to grasp and lift a manipulandum after observing an actor's lift. The object weight changed unpredictably (light or heavy) every fourth to sixth trial performed by the actor. Participants were informed that they would always lift the same weight as the actor and that, based on the experimental condition, they would have to observe skilled or erroneously performed lifts. Our results revealed that the observation of both skilled and erroneously performed lifts allows participants to update their internal sensorimotor object representation, in turn enabling them to predict force scaling accurately. These findings suggest that the observation of salient motor errors, as well as subtle features of skilled motor performance, are embodied in the observer's motor repertoire and can drive changes in predictive motor control.
RESUMO
Detecting and evaluating errors in action execution is essential for learning. Through complex interactions of the inverse and the forward model, the human motor system can predict and subsequently adjust ongoing or subsequent actions. Inputs to such a prediction are efferent and afferent signals from various sources. The aim of the current study was to examine the impact of visual as well as a combination of efferent and proprioceptive input signals to error prediction in a complex motor task. Predicting motor errors has been shown to be correlated with a neural signal known as the error-related negativity (Ne/ERN). Here, we tested how the Ne/ERN amplitude was modulated by the availability of different sensory signals in a semi-virtual throwing task where the action outcome (hit or miss of the target) was temporally delayed relative to movement execution allowing participants to form predictions about the outcome prior to the availability of knowledge of results. 19 participants practiced the task and electroencephalogram was recorded in two test conditions. In the Visual condition, participants received only visual input by passively observing the throwing movement. In the EffProp condition, participants actively executed the task while visual information about the real and the virtual effector was occluded. Hence, only efferent and proprioceptive signals were available. Results show a significant modulation of the Ne/ERN in the Visual condition while no effect could be observed in the EffProp condition. In addition, amplitudes of the feedback-related negativity in response to the actual outcome feedback were found to be inversely related to the Ne/ERN amplitudes. Our findings indicate that error prediction is modulated by the availability of input signals to the forward model. The observed amplitudes were found to be attenuated in comparison to previous studies, in which all efferent and sensory inputs were present. Furthermore, we assume that visual signals are weighted higher than proprioceptive signals, at least in goal-oriented tasks with visual targets.