AI4GMA

The early diagnosis of cerebral palsy is an area which has recently seen significant multi-disciplinary research. Diagnostic tools such as the General Movements Assessment (GMA), have produced some very promising results. However, the prospect of automating these processes may improve accessibility of the assessment and also enhance the understanding of movement development of infants. Previous works have established the viability of using pose-based features extracted from RGB video sequences to undertake classification of infant body movements based upon the GMA. In this project, we propose a series of new and improved features, and a feature fusion pipeline for this classification task. We also introduce the RVI-38 dataset, a series of videos captured as part of routine clinical care. By utilising this challenging dataset we establish the robustness of several motion features for classification, subsequently informing the design of our proposed feature fusion framework based upon the GMA. We also implement several other methods from the literature for direct comparison using two independent datasets.

While machine learning-based frameworks have obtained excellent performance in a wide range of visual understanding tasks, most of the existing frameworks can be considered black-box approaches since most of the classification frameworks only output the predicted label without specifying exactly what influences the classification decision. Whilst this is acceptable in typical computer vision tasks, it is less preferable in healthcare applications, since it is essential for the clinicians to verify the prediction as well. In order to make our proposed framework more interpretable, we include a visualization module as well as making deep learning models more interpretable by highlighting the body-parts that are contributing to the classification decision.