AI Approaches Towards Prechtl's Assessment of General Movements: A Systematic Literature Review

Towards novel classification of infants' movement patterns supported by computerized video analysis

BACKGROUND

Positional preferences, asymmetry of body position and movements potentially indicate abnormal clinical conditions in infants. However, a lack of standardized nomenclature hinders accurate assessment and documentation of these preferences over time. Video tools offer a safe and reproducible method to analyze and describe infant movement patterns, aiding in physiotherapy management and goal planning. The study aimed to develop an objective classification system for infant movement patterns with particular emphasis on the specific distribution of muscle tension, using methods of computer analysis of video recordings to enhance accuracy and reproducibility in assessments.

METHODS

The study involved the recording of videos of 51 infants between 6 and 15 weeks of age, born at term, with an Apgar score of at least 8 points. Based on observations of a recording of infant spontaneous movements in the supine position, experts identified postural-motor patterns: symmetry and typical asymmetry linked to the asymmetrical tonic neck reflex. Deviations from the typical postural-motor system were indicated, and subcategories of atypical patterns were distinguished. A computer-based inference system was developed to automatically classify individual patterns.

RESULTS

The following division of motor patterns was used: (1) normal patterns, including (a) typical (symmetrical, asymmetrical: variants 1 and 2); and (b) atypical (variants: 1 to 4), (2) positional preference, and (3) abnormal patterns. The proposed automatic classification method achieved an expert decision mapping accuracy of 84%. For atypical patterns, the high reproducibility of the system's results was confirmed. Lower reproducibility, not exceeding 70%, was achieved with typical patterns.

CONCLUSIONS

Based on the observation of infant spontaneous movements, it is possible to identify movement patterns divided into typical and atypical patterns. Computer-based analysis of infant movement patterns makes it possible to objectify and satisfactorily reproduce diagnostic decisions.

Artificial intelligence detects awareness of functional relation with the environment in 3 month old babies

A recent experiment probed how purposeful action emerges in early life by manipulating infants' functional connection to an object in the environment (i.e., tethering an infant's foot to a colorful mobile). Vicon motion capture data from multiple infant joints were used here to create Histograms of Joint Displacements (HJDs) to generate pose-based descriptors for 3D infant spatial trajectories. Using HJDs as inputs, machine and deep learning systems were tasked with classifying the experimental state from which snippets of movement data were sampled. The architectures tested included k-Nearest Neighbour (kNN), Linear Discriminant Analysis (LDA), Fully connected network (FCNet), 1D-Convolutional Neural Network (1D-Conv), 1D-Capsule Network (1D-CapsNet), 2D-Conv and 2D-CapsNet. Sliding window scenarios were used for temporal analysis to search for topological changes in infant movement related to functional context. kNN and LDA achieved higher classification accuracy with single joint features, while deep learning approaches, particularly 2D-CapsNet, achieved higher accuracy on full-body features. For each AI architecture tested, measures of foot activity displayed the most distinct and coherent pattern alterations across different experimental stages (reflected in the highest classification accuracy rate), indicating that interaction with the world impacts the infant behaviour most at the site of organism~world connection.

The General Movement Optimality Score-Revised (GMOS-R) with Socioeconomically Stratified Percentile Ranks

Background: The general movement optimality score (GMOS) quantifies the details of general movements (GMs). We recently conducted psychometric analyses of the GMOS and developed a revised scoresheet. Consequently, the GMOS-Revised (GMOS-R) instrument necessitated validation using new percentile ranks. This study aimed to provide these percentile ranks for the GMOS-R and to investigate whether sex, preterm birth, or the infant's country of birth and residence affected the GMOS-R distribution. Methods: We applied the GMOS-R to an international sample of 1983 infants (32% female, 44% male, and 24% not disclosed), assessed in the extremely and very preterm period (10%), moderate (12%) and late (22%) preterm periods, at term (25%), and post-term age (31%). Data were grouped according to the World Bank's classification into lower- and upper-middle-income countries (LMICs and UMICs; 26%) or high-income countries (HICs; 74%), respectively. Results: We found that sex and preterm or term birth did not affect either GM classification or the GMOS-R, but the country of residence did. A lower median GMOS-R for infants with normal or poor-repertoire GMs from LMICs and UMICs compared with HICs suggests the use of specific percentile ranks for LMICs and UMICs vs. HICs. Conclusion: For clinical and scientific use, we provide a freely available GMOS-R scoring sheet, with percentile ranks reflecting socioeconomic stratification.

Correlation between pose estimation features regarding movements towards the midline in early infancy

In infants, spontaneous movement towards the midline (MTM) indicates the initiation of anti-gravity ability development. Markerless 2D pose estimation is a cost-effective, time-efficient, and quantifiable alternative to movement assessment. We aimed to establish correlations between pose estimation features and MTM in early-age infants. Ninety-four infant videos were analysed to calculate the percentage and rate of MTM occurrence. 2D Pose estimation processed the videos and determined the distances and areas using wrist and ankle landmark coordinates. We collected data using video recordings from 20 infants aged 8-16 weeks post-term age. Correlations between MTM observations and distance values were evaluated. Differences in areas between groups of videos showing MTM and no MTM in the total, lower-limb, and upper-limb categories were examined. MTM observations revealed common occurrences of hand-to-trunk and foot-to-foot movements. Weak correlations were noted between limb distances to the midbody imaginary line and MTM occurrence values. Lower MTM showed significant differences in the lower part (p = 0.003) and whole area (p = 0.001). Video recording by parents or guardians could extract features using 2D pose estimation, assisting in the early identification of MTM in infants. Further research is required to assess a larger sample size with the diversity of MTM motor behaviour, and later developmental skills, and collect data from at-risk infants.

Machine learning approaches to evaluate infants' general movements in the writhing stage-a pilot study

The goals of this study are to describe machine learning techniques employing computer-vision movement algorithms to automatically evaluate infants' general movements (GMs) in the writhing stage. This is a retrospective study of infants admitted 07/2019 to 11/2021 to a level IV neonatal intensive care unit (NICU). Infant GMs, classified by certified expert, were analyzed in two-steps (1) determination of anatomic key point location using a NICU-trained pose estimation model [accuracy determined using object key point similarity (OKS)]; (2) development of a preliminary movement model to distinguish normal versus cramped-synchronized (CS) GMs using cosine similarity and autocorrelation of major joints. GMs were analyzed using 85 videos from 74 infants; gestational age at birth 28.9 ± 4.1 weeks and postmenstrual age (PMA) at time of video 35.9 ± 4.6 weeks The NICU-trained pose estimation model was more accurate (0.91 ± 0.008 OKS) than a generic model (0.83 ± 0.032 OKS, p < 0.001). Autocorrelation values in the lower limbs were significantly different between normal (5 videos) and CS GMs (5 videos, p < 0.05). These data indicate that automated pose estimation of anatomical key points is feasible in NICU patients and that a NICU-trained model can distinguish between normal and CS GMs. These preliminary data indicate that machine learning techniques may represent a promising tool for earlier CP risk assessment in the writhing stage and prior to hospital discharge.

Automating General Movements Assessment with quantitative deep learning to facilitate early screening of cerebral palsy

The Prechtl General Movements Assessment (GMA) is increasingly recognized for its role in evaluating the integrity of the developing nervous system and predicting motor dysfunctions, particularly in conditions such as cerebral palsy (CP). However, the necessity for highly trained professionals has hindered the adoption of GMA as an early screening tool in some countries. In this study, we propose a deep learning-based motor assessment model (MAM) that combines infant videos and basic characteristics, with the aim of automating GMA at the fidgety movements (FMs) stage. MAM demonstrates strong performance, achieving an Area Under the Curve (AUC) of 0.967 during external validation. Importantly, it adheres closely to the principles of GMA and exhibits robust interpretability, as it can accurately identify FMs within videos, showing substantial agreement with expert assessments. Leveraging the predicted FMs frequency, a quantitative GMA method is introduced, which achieves an AUC of 0.956 and enhances the diagnostic accuracy of GMA beginners by 11.0%. The development of MAM holds the potential to significantly streamline early CP screening and revolutionize the field of video-based quantitative medical diagnostics.

Wearable-based human flow experience recognition enhanced by transfer learning methods using emotion data

BACKGROUND

Flow experience is a specific positive and affective state that occurs when humans are completely absorbed in an activity and forget everything else. This state can lead to high performance, well-being, and productivity at work. Few studies have been conducted to determine the human flow experience using physiological wearable sensor devices. Other studies rely on self-reported data.

METHODS

In this article, we use physiological data collected from 25 subjects with multimodal sensing devices, in particular the Empatica E4 wristband, the Emotiv Epoc X electroencephalography (EEG) headset, and the Biosignalplux RespiBAN - in arithmetic and reading tasks to automatically discriminate between flow and non-flow states using feature engineering and deep feature learning approaches. The most meaningful wearable device for flow detection is determined by comparing the performances of each device. We also investigate the connection between emotions and flow by testing transfer learning techniques involving an emotion recognition-related task on the source domain.

RESULTS

The EEG sensor modalities yielded the best performances with an accuracy of 64.97%, and a macro Averaged F1 (AF1) score of 64.95%. An accuracy of 73.63% and an AF1 score of 72.70% were obtained after fusing all sensor modalities from all devices. Additionally, our proposed transfer learning approach using emotional arousal classification on the DEAP dataset led to an increase in performances with an accuracy of 75.10% and an AF1 score of 74.92%.

CONCLUSION

The results of this study suggest that effective discrimination between flow and non-flow states is possible with multimodal sensor data. The success of transfer learning using the DEAP emotion dataset as a source domain indicates that emotions and flow are connected, and emotion recognition can be used as a latent task to enhance the performance of flow recognition.

Infant movement classification through pressure distribution analysis

BACKGROUND

Aiming at objective early detection of neuromotor disorders such as cerebral palsy, we propose an innovative non-intrusive approach using a pressure sensing device to classify infant general movements. Here we differentiate typical general movement patterns of the "fidgety period" (fidgety movements) vs. the "pre-fidgety period" (writhing movements).

METHODS

Participants (N = 45) were sampled from a typically-developing infant cohort. Multi-modal sensor data, including pressure data from a pressure sensing mat with 1024 sensors, were prospectively recorded for each infant in seven succeeding laboratory sessions in biweekly intervals from 4 to 16 weeks of post-term age. 1776 pressure data snippets, each 5 s long, from the two targeted age periods were taken for movement classification. Each snippet was pre-annotated based on corresponding synchronised video data by human assessors as either fidgety present or absent. Multiple neural network architectures were tested to distinguish the fidgety present vs. fidgety absent classes, including support vector machines, feed-forward networks, convolutional neural networks, and long short-term memory networks.

RESULTS

Here we show that the convolution neural network achieved the highest average classification accuracy (81.4%). By comparing the pros and cons of other methods aiming at automated general movement assessment to the pressure sensing approach, we infer that the proposed approach has a high potential for clinical applications.

CONCLUSIONS

We conclude that the pressure sensing approach has great potential for efficient large-scale motion data acquisition and sharing. This will in return enable improvement of the approach that may prove scalable for daily clinical application for evaluating infant neuromotor functions.

2D Capsule Networks Detect Perceived Changes in Infant∼Environment Relationship Reflected in 3D Movement Dynamics

Can infant exploration and causal discovery be detected using Artificial Intelligence (AI)? A recent experiment probed how purposeful action emerges in early life by manipulating infants' functional connection to an object in the environment (i.e., tethering one foot to a colorful mobile). Vicon motion capture data from multiple infant joints were used here to create Histograms of Joint Displacements (HJDs) to generate pose-based descriptors for 3D infant spatial trajectories. Using HJDs as inputs, machine and deep learning systems were tasked with classifying the experimental state from which snippets of movement data were sampled. The architectures tested included k-Nearest Neighbour (kNN), Linear Discriminant Analysis (LDA), Fully connected network (FCNet), 1D-Convolutional Neural Network (1D-Conv), 1D-Capsule Network (1D-CapsNet), 2D-Conv and 2D-CapsNet. Sliding window scenarios were used for temporal analysis to search for topological changes in infant movement related to functional context. kNN and LDA achieved higher classification accuracy with single joint features, while deep learning approaches, particularly 2D-CapsNet, achieved higher accuracy on full-body features. For each AI architecture tested, measures of foot activity displayed the most distinct and coherent pattern alterations across different experimental stages (reflected in the highest classification accuracy rate), indicating that interaction with the world impacts the infant behaviour most at the site of organism∼world connection. Pairing theory-driven experimentation with AI tools thus opens a path to developing functionally-relevant assessments of infant behaviour that are likely to be useful in clinical settings.

Sensor-Based Classification of Primary and Secondary Car Driver Activities Using Convolutional Neural Networks

To drive safely, the driver must be aware of the surroundings, pay attention to the road traffic, and be ready to adapt to new circumstances. Most studies on driving safety focus on detecting anomalies in driver behavior and monitoring cognitive capabilities in drivers. In our study, we proposed a classifier for basic activities in driving a car, based on a similar approach that could be applied to the recognition of basic activities in daily life, that is, using electrooculographic (EOG) signals and a one-dimensional convolutional neural network (1D CNN). Our classifier achieved an accuracy of 80% for the 16 primary and secondary activities. The accuracy related to activities in driving, including crossroad, parking, roundabout, and secondary activities, was 97.9%, 96.8%, 97.4%, and 99.5%, respectively. The F1 score for secondary driving actions (0.99) was higher than for primary driving activities (0.93-0.94). Furthermore, using the same algorithm, it was possible to distinguish four activities related to activities of daily life that were secondary activities when driving a car.

Mobile Solutions for Clinical Surveillance and Evaluation in Infancy-General Movement Apps

The Prechtl General Movements Assessment (GMA) has become a clinician and researcher toolbox for evaluating neurodevelopment in early infancy. Given that it involves the observation of infant movements from video recordings, utilising smartphone applications to obtain these recordings seems like the natural progression for the field. In this review, we look back on the development of apps for acquiring general movement videos, describe the application and research studies of available apps, and discuss future directions of mobile solutions and their usability in research and clinical practice. We emphasise the importance of understanding the background that has led to these developments while introducing new technologies, including the barriers and facilitators along the pathway. The GMApp and Baby Moves apps were the first ones developed to increase accessibility of the GMA, with two further apps, NeuroMotion and InMotion, designed since. The Baby Moves app has been applied most frequently. For the mobile future of GMA, we advocate collaboration to boost the field's progression and to reduce research waste. We propose future collaborative solutions, including standardisation of cross-site data collection, adaptation to local context and privacy laws, employment of user feedback, and sustainable IT structures enabling continuous software updating.

Automated postural asymmetry assessment in infants neurodevelopmental evaluation using novel video-based features

BACKGROUND AND OBJECTIVE

Neurodevelopmental assessment enables the identification of infant developmental disorders in the first months of life. Thus, the appropriate therapy can be initiated promptly, increasing the chances for correct motor function. Posture asymmetry is one of the crucial aspects evaluated during the diagnosis. Available diagnostic methods are mainly based on qualitative assessment and subjective expert opinion. Current trends in computer-aided diagnosis focus mostly on analyzing infants' spontaneous movement videos using artificial intelligence methods, based primarily on limbs movement. This study aims to develop an automatic method for determining the infant's positional asymmetry in a video recording using computer image processing methods.

METHODS

We made the first attempt to determine positional preferences in a recording automatically. We proposed six quantitative features describing trunk and head position based on pose estimation. As a result of our algorithm, we estimate the percentage of each trunk position in a recording using known machine learning methods. The training and test sets were created from 51 recordings collected during our research and 12 recordings from the benchmark dataset evaluated by five of our experts. The method was assessed using the leave-one-subject-out cross-validation method for ground truth video fragments and different classifiers. Log loss for multiclass classification and ROC AUC were determined to evaluate the results for both our and benchmark datasets.

RESULTS

In a classification of the shortened side, the QDA classifier yields the most accurate results, gaining the lowest log loss of 0.552 and AUC of 0.913. The high accuracy (92.03) and sensitivity (93.26) confirm the method's potential in screening for asymmetry.

CONCLUSIONS

The method allows obtaining quantitative information about positional preference, a valuable extension of basic diagnostics without additional tools and procedures. In combination with an analysis of limbs movement, it may constitute one of the elements of a novelty computer-aided infants' diagnosis system in the future.

Sleep Stage Classification in Children Using Self-Attention and Gaussian Noise Data Augmentation

The analysis of sleep stages for children plays an important role in early diagnosis and treatment. This paper introduces our sleep stage classification method addressing the following two challenges: the first is the data imbalance problem, i.e., the highly skewed class distribution with underrepresented minority classes. For this, a Gaussian Noise Data Augmentation (GNDA) algorithm was applied to polysomnography recordings to seek the balance of data sizes for different sleep stages. The second challenge is the difficulty in identifying a minority class of sleep stages, given their short sleep duration and similarities to other stages in terms of EEG characteristics. To overcome this, we developed a DeConvolution- and Self-Attention-based Model (DCSAM) which can inverse the feature map of a hidden layer to the input space to extract local features and extract the correlations between all possible pairs of features to distinguish sleep stages. The results on our dataset show that DCSAM based on GNDA obtains an accuracy of 90.26% and a macro F1-score of 86.51% which are higher than those of our previous method. We also tested DCSAM on a well-known public dataset-Sleep-EDFX-to prove whether it is applicable to sleep data from adults. It achieves a comparable performance to state-of-the-art methods, especially accuracies of 91.77%, 92.54%, 94.73%, and 95.30% for six-stage, five-stage, four-stage, and three-stage classification, respectively. These results imply that our DCSAM based on GNDA has a great potential to offer performance improvements in various medical domains by considering the data imbalance problems and correlations among features in time series data.

Building Blocks for Deep Phenotyping in Infancy: A Use Case Comparing Spontaneous Neuromotor Functions in Prader-Willi Syndrome and Cerebral Palsy

With the increasing worldwide application of the Prechtl general movements assessment (GMA) beyond its original field of the early prediction of cerebral palsy (CP), substantial knowledge has been gained on early neuromotor repertoires across a broad spectrum of diagnostic groups. Here, we aimed to profile the neuromotor functions of infants with Prader-Willi syndrome (PWS) and to compare them with two other matched groups. One group included infants with CP; the other included patients who were treated at the same clinic and turned out to have inconspicuous developmental outcomes (IOs). The detailed GMA, i.e., the motor optimality score-revised (MOS-R), was used to prospectively assess the infants' (N = 54) movements. We underwent cross-condition comparisons to characterise both within-group similarities and variations and between-group distinctions and overlaps in infants' neuromotor functions. Although infants in both the PWS and the CP groups scored similarly low on MOS-R, their motor patterns were different. Frog-leg and mantis-hand postures were frequently seen in the PWS group. However, a PWS-specific general movements pattern was not observed. We highlight that pursuing in-depth knowledge within and beyond the motor domain in different groups has the potential to better understand different conditions, improve accurate diagnosis and individualised therapy, and contribute to deep phenotyping for precision medicine.

Semi-supervised body parsing and pose estimation for enhancing infant general movement assessment

General movement assessment (GMA) of infant movement videos (IMVs) is an effective method for early detection of cerebral palsy (CP) in infants. We demonstrate in this paper that end-to-end trainable neural networks for image sequence recognition can be applied to achieve good results in GMA, and more importantly, augmenting raw video with infant body parsing and pose estimation information can significantly improve performance. To solve the problem of efficiently utilizing partially labeled IMVs for body parsing, we propose a semi-supervised model, termed SiamParseNet (SPN), which consists of two branches, one for intra-frame body parts segmentation and another for inter-frame label propagation. During training, the two branches are jointly trained by alternating between using input pairs of only labeled frames and input of both labeled and unlabeled frames. We also investigate training data augmentation by proposing a factorized video generative adversarial network (FVGAN) to synthesize novel labeled frames for training. FVGAN decouples foreground and background generation which allows for generating multiple labeled frames from one real labeled frame. When testing, we employ a multi-source inference mechanism, where the final result for a test frame is either obtained via the segmentation branch or via propagation from a nearby key frame. We conduct extensive experiments for body parsing using SPN on two infant movement video datasets; on these partially labeled IMVs, we show that SPN coupled with FVGAN achieves state-of-the-art performance. We further demonstrate that our proposed SPN can be easily adapted to the infant pose estimation task with superior performance. Last but not least, we explore the clinical application of our method for GMA. We collected a new clinical IMV dataset with GMA annotations, and our experiments show that our SPN models for body parsing and pose estimation trained on the first two datasets generalize well to the new clinical dataset and their results can significantly boost the convolutional recurrent neural network (CRNN) based GMA prediction performance when combined with raw video inputs.

Design and Construct Validity of a Postural Control Test for Pre-Term Infants

A review of the literature indicated that the greatest prognostic value for predicting motor impairment in high-risk infants is the absence of fidgety movements (FMs) at 3 months of post-term age. The purpose of the present study was to characterize a new posturometric test (PT) based on a center-of-pressure (CoP) movement analysis, in terms of design and construct validity, for the detection of postural control disturbances in pre-term infants. The comparative studies were carried out between pre-term infants who presented normal FMs (18 participants) and infants with absent FMs (19 participants), which consisted of the analysis of the CoP trajectory and CoP area in supine and prone positions using the force platform. New PT was performed simultaneously with GMs recorded using a force platform. Statistical analyses revealed significant differences between the groups of infants who presented absent FMs and normal FMs for almost all CoP parameters describing spontaneous sway in the supine position. Based on these preliminary results, it can be concluded, that the application of PT based on the analysis of CoP trajectory, area, and velocity in the supine position has been demonstrated to be valid for the detection of postural control disturbances in pre-term infants.

SenseHunger: Machine Learning Approach to Hunger Detection Using Wearable Sensors

The perception of hunger and satiety is of great importance to maintaining a healthy body weight and avoiding chronic diseases such as obesity, underweight, or deficiency syndromes due to malnutrition. There are a number of disease patterns, characterized by a chronic loss of this perception. To our best knowledge, hunger and satiety cannot be classified using non-invasive measurements. Aiming to develop an objective classification system, this paper presents a multimodal sensory system using associated signal processing and pattern recognition methods for hunger and satiety detection based on non-invasive monitoring. We used an Empatica E4 smartwatch, a RespiBan wearable device, and JINS MEME smart glasses to capture physiological signals from five healthy normal weight subjects inactively sitting on a chair in a state of hunger and satiety. After pre-processing the signals, we compared different feature extraction approaches, either based on manual feature engineering or deep feature learning. Comparative experiments were carried out to determine the most appropriate sensor channel, device, and classifier to reliably discriminate between hunger and satiety states. Our experiments showed that the most discriminative features come from three specific sensor modalities: Electrodermal Activity (EDA), infrared Thermopile (Tmp), and Blood Volume Pulse (BVP).

Development and Validation of a Deep Learning Method to Predict Cerebral Palsy From Spontaneous Movements in Infants at High Risk

IMPORTANCE

Early identification of cerebral palsy (CP) is important for early intervention, yet expert-based assessments do not permit widespread use, and conventional machine learning alternatives lack validity.

OBJECTIVE

To develop and assess the external validity of a novel deep learning-based method to predict CP based on videos of infants' spontaneous movements at 9 to 18 weeks' corrected age.

DESIGN, SETTING, AND PARTICIPANTS

This prognostic study of a deep learning-based method to predict CP at a corrected age of 12 to 89 months involved 557 infants with a high risk of perinatal brain injury who were enrolled in previous studies conducted at 13 hospitals in Belgium, India, Norway, and the US between September 10, 2001, and October 25, 2018. Analysis was performed between February 11, 2020, and September 23, 2021. Included infants had available video recorded during the fidgety movement period from 9 to 18 weeks' corrected age, available classifications of fidgety movements ascertained by the general movement assessment (GMA) tool, and available data on CP status at 12 months' corrected age or older. A total of 418 infants (75.0%) were randomly assigned to the model development (training and internal validation) sample, and 139 (25.0%) were randomly assigned to the external validation sample (1 test set).

EXPOSURE

Video recording of spontaneous movements.

MAIN OUTCOMES AND MEASURES

The primary outcome was prediction of CP. Deep learning-based prediction of CP was performed automatically from a single video. Secondary outcomes included prediction of associated functional level and CP subtype. Sensitivity, specificity, positive and negative predictive values, and accuracy were assessed.

RESULTS

Among 557 infants (310 [55.7%] male), the median (IQR) corrected age was 12 (11-13) weeks at assessment, and 84 infants (15.1%) were diagnosed with CP at a mean (SD) age of 3.4 (1.7) years. Data on race and ethnicity were not reported because previous studies (from which the infant samples were derived) used different study protocols with inconsistent collection of these data. On external validation, the deep learning-based CP prediction method had sensitivity of 71.4% (95% CI, 47.8%-88.7%), specificity of 94.1% (95% CI, 88.2%-97.6%), positive predictive value of 68.2% (95% CI, 45.1%-86.1%), and negative predictive value of 94.9% (95% CI, 89.2%-98.1%). In comparison, the GMA tool had sensitivity of 70.0% (95% CI, 45.7%-88.1%), specificity of 88.7% (95% CI, 81.5%-93.8%), positive predictive value of 51.9% (95% CI, 32.0%-71.3%), and negative predictive value of 94.4% (95% CI, 88.3%-97.9%). The deep learning method achieved higher accuracy than the conventional machine learning method (90.6% [95% CI, 84.5%-94.9%] vs 72.7% [95% CI, 64.5%-79.9%]; P < .001), but no significant improvement in accuracy was observed compared with the GMA tool (85.9%; 95% CI, 78.9%-91.3%; P = .11). The deep learning prediction model had higher sensitivity among infants with nonambulatory CP (100%; 95% CI, 63.1%-100%) vs ambulatory CP (58.3%; 95% CI, 27.7%-84.8%; P = .02) and spastic bilateral CP (92.3%; 95% CI, 64.0%-99.8%) vs spastic unilateral CP (42.9%; 95% CI, 9.9%-81.6%; P < .001).

CONCLUSIONS AND RELEVANCE

In this prognostic study, a deep learning-based method for predicting CP at 9 to 18 weeks' corrected age had predictive accuracy on external validation, which suggests possible avenues for using deep learning-based software to provide objective early detection of CP in clinical settings.

[Differences in the Distribution of the General Movements Classification Between Neonatal Risk Groups in the Children's Hospital Zurich: An Observational Study]

INTRODUCTION

Neonatal infants are at increased risk for motor development disorders.

OBJECTIVE

To compare General Movements (GMs) classification between three neonatal risk groups, correlate the GMs Assessment (GMA) with a standardized developmental neurological examination (SDNE) and determine risk factors for abnormal GMs.

METHODS

Monocentric observational study with three risk groups (children with operated congenital heart disease (CHD) n=26, with operated congenital gastrointestinal malformations (CGM) n=17 and with fetal operated myelomeningocele (MMC) n=12 who underwent inpatient video-based examination. GMA was evaluated according to Hadders-Algra classification and divided into 4 categories: normal optimal (NO), normal suboptimal (NS), mildly abnormal (MA), definitely abnormal (DA).

RESULTS

The distribution was as follows: CHD 80.8% NS, 19.2% MA, CGM 5.9% NO, 64.7% NS, 29.4% MA, MMC upper extremities 100% NS, lower extremities 33.3% NS, 33.3% MA and 33.3% DA (group comparison Kruskal-Wallis 10.729, p=0.003). GMA correlated significantly with SDNE (Spearman r _s=0.869, p<0.001). Binary logistic regression analysis showed that only gestational age (Chi²=11.93, p<0.001) correlated with abnormal GMs.

CONCLUSION

The majority of children showed normal GMs. Children with MMC and those with lower gestational age showed an increased risk of abnormal GMs. The GMA and SDNE represent complementary "bedside tools" to detect early motor abnormalities.

Deep learning-based quantitative analyses of spontaneous movements and their association with early neurological development in preterm infants

This study aimed to develop quantitative assessments of spontaneous movements in high-risk preterm infants based on a deep learning algorithm. Video images of spontaneous movements were recorded in very preterm infants at the term-equivalent age. The Hammersmith Infant Neurological Examination (HINE) was performed in infants at 4 months of corrected age. Joint positional data were extracted using a pretrained pose-estimation model. Complexity and similarity indices of joint angle and angular velocity in terms of sample entropy and Pearson correlation coefficient were compared between the infants with HINE < 60 and ≥ 60. Video images of spontaneous movements were recorded in 65 preterm infants at term-equivalent age. Complexity indices of joint angles and angular velocities differed between the infants with HINE < 60 and ≥ 60 and correlated positively with HINE scores in most of the joints at the upper and lower extremities (p < 0.05). Similarity indices between each joint angle or joint angular velocity did not differ between the two groups in most of the joints at the upper and lower extremities. Quantitative assessments of spontaneous movements in preterm infants are feasible using a deep learning algorithm and sample entropy. The results indicated that complexity indices of joint movements at both the upper and lower extremities can be potential candidates for detecting developmental outcomes in preterm infants.

Video-Based Automatic Baby Motion Analysis for Early Neurological Disorder Diagnosis: State of the Art and Future Directions

Neurodevelopmental disorders (NDD) are impairments of the growth and development of the brain and/or central nervous system. In the light of clinical findings on early diagnosis of NDD and prompted by recent advances in hardware and software technologies, several researchers tried to introduce automatic systems to analyse the baby's movement, even in cribs. Traditional technologies for automatic baby motion analysis leverage contact sensors. Alternatively, remotely acquired video data (e.g., RGB or depth) can be used, with or without active/passive markers positioned on the body. Markerless approaches are easier to set up and maintain (without any human intervention) and they work well on non-collaborative users, making them the most suitable technologies for clinical applications involving children. On the other hand, they require complex computational strategies for extracting knowledge from data, and then, they strongly depend on advances in computer vision and machine learning, which are among the most expanding areas of research. As a consequence, also markerless video-based analysis of movements in children for NDD has been rapidly expanding but, to the best of our knowledge, there is not yet a survey paper providing a broad overview of how recent scientific developments impacted it. This paper tries to fill this gap and it lists specifically designed data acquisition tools and publicly available datasets as well. Besides, it gives a glimpse of the most promising techniques in computer vision, machine learning and pattern recognition which could be profitably exploited for children motion analysis in videos.

Technology-assisted quantification of movement to predict infants at high risk of motor disability: A systematic review

AIM

To systematically review the scientific literature to determine the predictive validity of technology-assisted measures of observable infant movement in infants less than six months of corrected age (CA) to identify high-risk of motor disability.

METHOD

A comprehensive search for randomised and non-randomised controlled trials, cohort studies and cross-comparison trials was performed on five electronic databases up to Feb 2021. Studies were included if they quantified infant movement before 6 months CA using some method of technology-assistance and compared the instrumented measure to a diagnostic clinical measure of neurodevelopment. Studies were excluded if they did not report a technology-assisted measure of infant movement. Methodological quality of the included studies was assessed using the Downs and Black scale.

RESULTS

23 studies met the full inclusion and exclusion criteria. Methodological quality of the included papers ranged from 9 to 24 (out of 26) on the Downs and Black scale. Infant movement assessments included the General Movements Assessment (GMA) and domains of the Hammersmith Infant Neurological Assessment (HINE). Studies used 2D video recordings, RGB-Depth recordings, accelerometry, and electromagnetic motion tracking technologies to quantify movement. Analytical approaches and movement features of interest were individual and varied. Technology assisted quantitative assessments identified cases of later diagnosed CP with sensitivity 44-100 %, specificity 59-95 %, Area under the ROC Curve 82-93 %; and typical development with sensitivity range 30-46 %, specificity 88-95 %, Area under the ROC Curve 68 %.

INTERPRETATION

Technology-assisted assessments of movement in infants less than 6 months CA using current technologies are feasible. Validation of measurement tools are limited. Although methods and results appear promising clinical uptake of technology-assisted assessments remains limited.

The Symphony of Team Flow in Virtual Teams. Using Artificial Intelligence for Its Recognition and Promotion

More and more teams are collaborating virtually across the globe, and the COVID-19 pandemic has further encouraged the dissemination of virtual teamwork. However, there are challenges for virtual teams – such as reduced informal communication – with implications for team effectiveness. Team flow is a concept with high potential for promoting team effectiveness, however its measurement and promotion are challenging. Traditional team flow measurements rely on self-report questionnaires that require interrupting the team process. Approaches in artificial intelligence, i.e., machine learning, offer methods to identify an algorithm based on behavioral and sensor data that is able to identify team flow and its dynamics over time without interrupting the process. Thus, in this article we present an approach to identify team flow in virtual teams, using machine learning methods. First of all, based on a literature review, we provide a model of team flow characteristics, composed of characteristics that are shared with individual flow and characteristics that are unique for team flow. It is argued that those characteristics that are unique for team flow are represented by the concept of collective communication. Based on that, we present physiological and behavioral correlates of team flow which are suitable – but not limited to – being assessed in virtual teams and which can be used as input data for a machine learning system to assess team flow in real time. Finally, we suggest interventions to support team flow that can be implemented in real time, in virtual environments and controlled by artificial intelligence. This article thus contributes to finding indicators and dynamics of team flow in virtual teams, to stimulate future research and to promote team effectiveness.

Novel AI driven approach to classify infant motor functions

The past decade has evinced a boom of computer-based approaches to aid movement assessment in early infancy. Increasing interests have been dedicated to develop AI driven approaches to complement the classic Prechtl general movements assessment (GMA). This study proposes a novel machine learning algorithm to detect an age-specific movement pattern, the fidgety movements (FMs), in a prospectively collected sample of typically developing infants. Participants were recorded using a passive, single camera RGB video stream. The dataset of 2800 five-second snippets was annotated by two well-trained and experienced GMA assessors, with excellent inter- and intra-rater reliabilities. Using OpenPose, the infant full pose was recovered from the video stream in the form of a 25-points skeleton. This skeleton was used as input vector for a shallow multilayer neural network (SMNN). An ablation study was performed to justify the network's architecture and hyperparameters. We show for the first time that the SMNN is sufficient to discriminate fidgety from non-fidgety movements in a sample of age-specific typical movements with a classification accuracy of 88%. The computer-based solutions will complement original GMA to consistently perform accurate and efficient screening and diagnosis that may become universally accessible in daily clinical practice in the future.

The future of General Movement Assessment: The role of computer vision and machine learning - A scoping review

BACKGROUND

The clinical and scientific value of Prechtl general movement assessment (GMA) has been increasingly recognised, which has extended beyond the detection of cerebral palsy throughout the years. With advancing computer science, a surging interest in developing automated GMA emerges.

AIMS

In this scoping review, we focused on video-based approaches, since it remains authentic to the non-intrusive principle of the classic GMA. Specifically, we aimed to provide an overview of recent video-based approaches targeting GMs; identify their techniques for movement detection and classification; examine if the technological solutions conform to the fundamental concepts of GMA; and discuss the challenges of developing automated GMA.

METHODS AND PROCEDURES

We performed a systematic search for computer vision-based studies on GMs.

OUTCOMES AND RESULTS

We identified 40 peer-reviewed articles, most (n = 30) were published between 2017 and 2020. A wide variety of sensing, tracking, detection, and classification tools for computer vision-based GMA were found. Only a small portion of these studies applied deep learning approaches. A comprehensive comparison between data acquisition and sensing setups across the reviewed studies, highlighting limitations and advantages of each modality in performing automated GMA is provided.

CONCLUSIONS AND IMPLICATIONS

A "method-of-choice" for automated GMA does not exist. Besides creating large datasets, understanding the fundamental concepts and prerequisites of GMA is necessary for developing automated solutions. Future research shall look beyond the narrow field of detecting cerebral palsy and open up to the full potential of applying GMA to enable an even broader application.

Writhing Movement Detection in Newborns on the Second and Third Day of Life Using Pose-Based Feature Machine Learning Classification

Observation of neuromotor development at an early stage of an infant’s life allows for early diagnosis of deficits and the beginning of the therapeutic process. General movement assessment is a method of spontaneous movement observation, which is the foundation for contemporary attempts at objectification and computer-aided diagnosis based on video recordings’ analysis. The present study attempts to automatically detect writhing movements, one of the normal general movement categories presented by newborns in the first weeks of life. A set of 31 recordings of newborns on the second and third day of life was divided by five experts into videos containing writhing movements (with occurrence time) and poor repertoire, characterized by a lower quality of movement in relation to the norm. Novel, objective pose-based features describing the scope, nature, and location of each limb’s movement are proposed. Three machine learning algorithms are evaluated in writhing movements’ detection in leave-one-out cross-validation for different feature extraction time windows and overlapping time. The experimental results make it possible to indicate the optimal parameters for which 80% accuracy was achieved. Based on automatically detected writhing movement percent in the video, infant movements are classified as writhing movements or poor repertoire with an area under the ROC (receiver operating characteristics) curve of 0.83.