In image captioning models, the main challenge in describing an image is identifying all the objects by precisely considering the relationships between the objects and producing various captions.... Show moreIn image captioning models, the main challenge in describing an image is identifying all the objects by precisely considering the relationships between the objects and producing various captions. Over the past few years, many methods have been proposed, from an attribute-to-attribute comparison approach to handling issues related to semantics and their relationships. Despite the improvements, the existing techniques suffer from inadequate positional and geometrical attributes concepts. The reason is that most of the abovementioned approaches depend on Convolutional Neural Networks (CNNs) for object detection. CNN is notorious for failing to detect equivariance and rotational invariance in objects. Moreover, the pooling layers in CNNs cause valuable information to be lost. Inspired by the recent successful approaches, this paper introduces a novel framework for extracting meaningful descriptions based on a parallelized capsule network that describes the content of images through a high level of understanding of the semantic contents of an image. The main contribution of this paper is proposing a new method that not only overrides the limitations of CNNs but also generates descriptions with a wide variety of words by using Wikipedia. In our framework, capsules focus on the generation of meaningful descriptions with more detailed spatial and geometrical attributes for a given set of images by considering the position of the entities as well as their relationships. Qualitative experiments on the benchmark dataset MS-COCO show that our framework outperforms state-of-the-art image captioning models when describing the semantic content of the images. Show less
Leeuw, A.W. de; Baar, R. van.; Knobbe, A.J.; Zwaard, S. van der 2022
In this study, we investigated the relationships between training load, perceived wellness and match performance in professional volleyball by applying the machine learning techniques XGBoost,... Show moreIn this study, we investigated the relationships between training load, perceived wellness and match performance in professional volleyball by applying the machine learning techniques XGBoost, random forest regression and subgroup discovery. Physical load data were obtained by manually logging all physical activities and using wearable sensors. Daily wellness of players was monitored using questionnaires. Match performance was derived from annotated actions by a video scout during matches. We identified conditions of predictor variables that related to attack and pass performance (p < 0.05). Better attack performance is related to heavy weights of lower-body strength training exercises in the preceding four weeks. However, worse attack performance is linked to large variations in weights of full-body strength training exercises, excessively heavy upper-body strength training, low jump heights and small variations in the number of high jumps in the four weeks prior to competition. Lower passing performance was associated with small variations in the number of high jumps in the preceding week and an excessive amount of high jumps performed, on average, in the two weeks prior to competition. Differences in findings with respect to passing and attack performance suggest that elite volleyball players can improve their performance if training schedules are adapted to the position of a player. Show less
Current assays for Clostridioides difficile in nonhospital settings are outsourced and time-intensive, resulting in both delayed diagnosis and quarantining of infected individuals. We designed a... Show moreCurrent assays for Clostridioides difficile in nonhospital settings are outsourced and time-intensive, resulting in both delayed diagnosis and quarantining of infected individuals. We designed a more rapid point-of-care assay featuring a "turn-on " bioluminescent readout of a C. difficile-specific protease, PPEP-1. NanoLuc, a bright and stable luciferase, was "caged " with a PPEP-1-responsive peptide tail that inhibited luminescence. Upon proteolytic cleavage, the peptide was released and NanoLuc activity was restored, providing a visible readout. The bioluminescent sensor detected PPEP-1 concentrations as low as 10 nM. Sensor uncaging was achieved within minutes, and signal was captured using a digital camera. Importantly, the sensor was also functional at ambient temperature and compatible with fecal material, suggesting that it can be readily deployed in a variety of settings. Show less
Feijen, M.; Egorova, A.D.; Beeres, S.L.M.A.; Treskes, R.W. 2021
Heart failure (HF) hospitalisations due to decompensation are associated with shorter lifeexpectancy and lower quality of life. These hospitalisations pose a significant burden on the patients... Show moreHeart failure (HF) hospitalisations due to decompensation are associated with shorter lifeexpectancy and lower quality of life. These hospitalisations pose a significant burden on the patients,doctors and healthcare resources. Early detection of an upcoming episode of decompensationmay facilitate timely optimisation of the ambulatory medical treatment and thereby prevent heartfailure-related hospitalisations. The HeartLogicTM algorithm combines data from five sensors ofcardiac implantable electronic devices into a cumulative index value. It has been developed for earlydetection of fluid retention in heart failure patients. This review aims to provide an overview of thecurrent literature and experience with the HeartLogicTM algorithm, illustrate how the index can beimplemented in daily clinical practice and discuss ongoing studies and potential future developmentsof interest. Show less
Measuring psychophysiological signals of adolescents using unobtrusive wearable sensors may contribute to understanding the development of emotional disorders. This study investigated the... Show moreMeasuring psychophysiological signals of adolescents using unobtrusive wearable sensors may contribute to understanding the development of emotional disorders. This study investigated the feasibility of measuring high quality physiological data and examined the validity of signal processing in a school setting. Among 86 adolescents, a total of more than 410 h of electrodermal activity (EDA) data were recorded using a wrist-worn sensor with gelled electrodes and over 370 h of heart rate data were recorded using a chest-strap sensor. The results support the feasibility of monitoring physiological signals at school. We describe specific challenges and provide recommendations for signal analysis, including dealing with invalid signals due to loose sensors, and quantization noise that can be caused by limitations in analog-to-digital conversion in wearable devices and be mistaken as physiological responses. Importantly, our results show that using toolboxes for automatic signal preprocessing, decomposition, and artifact detection with default parameters while neglecting differences between devices and measurement contexts yield misleading results. Time courses of students’ physiological signals throughout the course of a class were found to be clearer after applying our proposed preprocessing steps. Show less
In the critical setting of a trauma team activation, team composition is crucial information that should be accessible at a glance. This calls for a technological solution, which are widely... Show moreIn the critical setting of a trauma team activation, team composition is crucial information that should be accessible at a glance. This calls for a technological solution, which are widely available, that allows access to the whereabouts of personnel. This diversity presents decision makers and users with many choices and considerations. The aim of this review is to give a comprehensive overview of available real-time person identification techniques and their respective characteristics. A systematic literature review was performed to create an overview of identification techniques that have been tested in medical settings or already have been implemented in clinical practice. These techniques have been investigated on a total of seven characteristics: costs, usability, accuracy, response time, hygiene, privacy, and user safety. The search was performed on 11 May 2020 in PubMed and the Web of Science Core Collection. PubMed and Web of Science yielded a total n = 265 and n = 228 records, respectively. The review process resulted in n = 23 included records. A total of seven techniques were identified: (a) active and (b) passive Radio-Frequency Identification (RFID) based systems, (c) fingerprint, (d) iris, and (e) facial identification systems and infrared (IR) (f) and ultrasound (US) (g) based systems. Active RFID was largely documented in the included literature. Only a few could be found about the passive systems. Biometric (c, d, and e) technologies were described in a variety of applications. IR and US techniques appeared to be a niche, as they were only spoken of in few (n = 3) studies. Show less
Marinucci, D.; Sbrollini, A.; Marcantoni, I.; Morettini, M.; Swenne, C.A.; Burattini, L. 2020
Atrial fibrillation (AF) is a common cardiac disorder that can cause severe complications. AF diagnosis is typically based on the electrocardiogram (ECG) evaluation in hospitals or in clinical... Show moreAtrial fibrillation (AF) is a common cardiac disorder that can cause severe complications. AF diagnosis is typically based on the electrocardiogram (ECG) evaluation in hospitals or in clinical facilities. The aim of the present work is to propose a new artificial neural network for reliable AF identification in ECGs acquired through portable devices. A supervised fully connected artificial neural network (RSL_ANN), receiving 19 ECG features (11 morphological, 4 on F waves and 4 on heart-rate variability (HRV)) in input and discriminating between AF and non-AF classes in output, was created using the repeated structuring and learning (RSL) procedure. RSL_ANN was created and tested on 8028 (training: 4493; validation: 1125; testing: 2410) annotated ECGs belonging to the "AF Classification from a Short Single Lead ECG Recording" database and acquired with the portable KARDIA device by AliveCor. RSL_ANN performance was evaluated in terms of area under the curve (AUC) and confidence intervals (CIs) of the received operating characteristic. RSL_ANN performance was very good and very similar in training, validation and testing datasets. AUC was 91.1% (CI: 89.1-93.0%), 90.2% (CI: 86.2-94.3%) and 90.8% (CI: 88.1-93.5%) for the training, validation and testing datasets, respectively. Thus, RSL_ANN is a promising tool for reliable identification of AF in ECGs acquired by portable devices. Show less
The rapid evolution of technology, sensors and personal digital devices offers an opportunity to acquire health related data seamlessly, unobtrusively and in real time. In this opinion piece, we... Show moreThe rapid evolution of technology, sensors and personal digital devices offers an opportunity to acquire health related data seamlessly, unobtrusively and in real time. In this opinion piece, we discuss the relevance and opportunities for using digital sensing in dermatology, taking eczema as an exemplar. Show less
Chilipirea, C.; Baratchi, M.; Dobre, C.; Steen, M. van 2018
The Kinect v2 sensor may be a cheap and easy to use sensor to quantify gait in clinical settings, especially when applied in set-ups integrating multiple Kinect sensors to increase the measurement... Show moreThe Kinect v2 sensor may be a cheap and easy to use sensor to quantify gait in clinical settings, especially when applied in set-ups integrating multiple Kinect sensors to increase the measurement volume. Reliable estimates of foot placement locations are required to quantify spatial gait parameters. This study aimed to systematically evaluate the effects of distance from the sensor, side and step length on estimates of foot placement locations based on Kinect’s ankle body points. Subjects (n = 12) performed stepping trials at imposed foot placement locations distanced 2 m or 3 m from the Kinect sensor (distance), for left and right foot placement locations (side), and for five imposed step lengths. Body points’ time series of the lower extremities were recorded with a Kinect v2 sensor, placed frontoparallelly on the left side, and a gold-standard motion-registration system. Foot placement locations, step lengths, and stepping accuracies were compared between systems using repeated-measures ANOVAs, agreement statistics and two one-sided t-tests to test equivalence. For the right side at the 2mdistance from the sensor we found significant between-systems differences in foot placement locations and step lengths, and evidence for nonequivalence. This distance by side effect was likely caused by differences in body orientation relative to the Kinect sensor. It can be reduced by using Kinect’s higher-dimensional depth data to estimate foot placement locations directly from the foot’s point cloud and/or by using smaller inter-sensor distances in the case of a multi-Kinect v2 set-up to estimate foot placement locations at greater distances from the sensor. Show less
Rood, M.T.M.; Raspe, M.; Hove, J.B. ten; Jalink, K.; Velders, A.H.; Leeuwen, F.W.B. van 2015
