Abnormal Infant Movements Classification With Deep Learning on Pose-Based Features

  title={Abnormal Infant Movements Classification With Deep Learning on Pose-Based Features},
  author={Kevin D. McCay and Edmond S. L. Ho and Hubert P. H. Shum and Gerhard Fehringer and Claire Marcroft and Nicholas D. Embleton},
  journal={IEEE Access},
The pursuit of early diagnosis of cerebral palsy has been an active research area with some very promising results using tools such as the General Movements Assessment (GMA). In our previous work, we explored the feasibility of extracting pose-based features from video sequences to automatically classify infant body movement into two categories, normal and abnormal. The classification was based upon the GMA, which was carried out on the video data by an independent expert reviewer. In this… 
A Pose-Based Feature Fusion and Classification Framework for the Early Prediction of Cerebral Palsy in Infants
A series of new and improved features are proposed, and a feature fusion pipeline for this classification task based upon the General Movements Assessment, and the proposed pose-based method performs well across both datasets.
Novel AI driven approach to classify infant motor functions
It is shown 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%.
RGB-D Videos-Based Early Prediction of Infant Cerebral Palsy via General Movements Complexity
A novel method for early prediction of infant cerebral palsy based on General Movements Assessment (GMA) theory with RGB-D videos that achieved state-of-the-art with sensitivity of 100, specificity of 87.5%, and accuracy of 91.7%.
Writhing Movement Detection in Newborns on the Second and Third Day of Life Using Pose-Based Feature Machine Learning Classification
A machine learning algorithm is evaluated in writhing movements’ detection in leave-one-out cross-validation for different feature extraction time windows and overlapping time, and results make it possible to indicate the optimal parameters for which 80% accuracy was achieved.
A Spatio-Temporal Attention-Based Model for Infant Movement Assessment From Videos
This work develops and validate a new method for fidgety movement assessment from consumer-grade videos using human pose extracted from short clips, significantly outperforming existing competing methods with better interpretability.
Towards Explainable Abnormal Infant Movements Identification: A Body-part Based Prediction and Visualisation Framework
This paper proposes a new framework for the automated classification of infant body movements, based upon the General Movements Assessment, which unlike previous methods, also incorporates a visualization framework to aid with interpretability.
Cerebral Palsy Prediction with Frequency Attention Informed Graph Convolutional Networks
A frequency attention informed graph convolutional network is proposed and validates it on two consumer-grade RGB video datasets, namely MINI-RGBD and RVI-38 datasets and provides a way for supporting the early diagnosis of CP in the resource-limited regions where the clinical resources are not abundant.
Interpreting Deep Learning based Cerebral Palsy Prediction with Channel Attention
This paper proposes a channel attention module for deep learning models to predict cerebral palsy from infants’ body movements, which highlights the key features the model identifies as important, thereby indicating why certain diagnostic results are found.
Video-Based Automatic Baby Motion Analysis for Early Neurological Disorder Diagnosis: State of the Art and Future Directions
The most promising techniques in computer vision, machine learning and pattern recognition which could be profitably exploited for children motion analysis in videos are given.
Deep learning-based quantitative analyses of spontaneous movements and their association with early neurological development in preterm infants
Quantitative assessments of spontaneous movements in preterm infants are feasible using a deep learning algorithm and sample entropy and 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 pre term infants.


Establishing Pose Based Features Using Histograms for the Detection of Abnormal Infant Movements
A pilot study on extracting important information from video sequences to classify the body movement into two categories, normal and abnormal, and compared the results provided by an independent expert reviewer based on GMA is conducted.
Fully automated image-based estimation of postural point-features in children with cerebral palsy using deep learning
This study demonstrates, for the first time, technical feasibility to automate the identification of a sitting segmental posture including individual trunk segments, changes away from that posture, and support from the upper limb required for the clinical SATCo.
A Deep Learning Frame-Work for Recognizing Developmental Disorders
A novel framework to detect developmental disorders from facial images based on Deep Convolutional Neural Networks for feature extraction and results indicate that the model performs better than average human intelligence in terms of differentiating amongst different disabilities.
Detection of Atypical and Typical Infant Movements using Computer-based Video Analysis
  • S. Orlandi, K. Raghuram, T. Chau
  • Medicine, Psychology
    2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
  • 2018
A computerbased assessment would provide clinicians with an objective tool for early diagnosis of CP, to facilitate early intervention and improve functional outcomes.
Computer Vision for Medical Infant Motion Analysis: State of the Art and RGB-D Data Set
The Moving INfants In RGB-D (MINI-RGBD) data set is released, created using the recently introduced Skinned Multi-Infant Linear body model (SMIL) to foster research in medical infant motion analysis to get closer to an automated system for early detection of neurodevelopmental disorders.
Learning an Infant Body Model from RGB-D Data for Accurate Full Body Motion Analysis
A non-intrusive, low-cost, lightweight acquisition system that captures the shape and motion of infants, and provides a new tool and a step towards a fully automatic system for GMA.
Movement Recognition Technology as a Method of Assessing Spontaneous General Movements in High Risk Infants
Recent translational studies using movement recognition technology as a method of assessing movement in high risk infants are identified and may lead to a greater understanding of the development of the nervous system in infants at high risk of motor impairment.
Preterm Infants’ Pose Estimation With Spatio-Temporal Features
This article significantly enhances the state of art in automatic assessment of preterm infants’ health status by introducing the use of spatio-temporal features for limb detection and tracking, and by being the first study to use depth videos acquired in the actual clinical practice for limb-pose estimation.
Machine Learning of Infant Spontaneous Movements for the Early Prediction of Cerebral Palsy: A Multi-Site Cohort Study
The CIMA model, developed and tested on video recordings from a cohort of 377 high-risk infants at 9–15 weeks corrected age, had sensitivity and specificity comparable to observational GMA or neonatal cerebral imaging for the prediction of CP.
View invariant human action recognition using histograms of 3D joints
This paper presents a novel approach for human action recognition with histograms of 3D joint locations (HOJ3D) as a compact representation of postures and achieves superior results on the challenging 3D action dataset.