Hot spot detection based on feature space representation of visual search in medical imaging

This paper presents a new framework of analysing saccadic eye movements for capturing intrinsic visual search behaviours of different observers in medical image understanding. The method is based on the information theory for capturing salient image features from which the diagnoses are performed. We demonstrate how to obtain feature space fixation density functions along the scan paths that are normalised to the image content. This allows a reliable identification of salient image features, which can be mapped to spatial space to highlight regions of interest for directing visual attention. The processing steps involved are elaborated with an example for CT image understanding, and we demonstrate how the proposed framework can be used as a basis for decision support in medical image understanding.