Key result
A fully convolutional network trained to classify intracranial pressure waveforms based on cerebral compliance and arterial blood pressure morphology achieved 93.1% accuracy on a hold-out test set.
Why the study?
Analysis of intracranial pressure waveforms could detect cerebral pathophysiologies, but large variation in the generating arterial blood pressure signal presents a major obstacle.
Can fully convolutional networks and extended PCA accurately classify intracranial pressure waveforms to detect cerebral compliance states?
Can fully convolutional networks and extended PCA accurately classify intracranial pressure waveforms to detect cerebral compliance states?
Answers are grounded in this paper and cite their sources.
Fully convolutional networks and extended PCA can effectively classify intracranial pressure waveforms to assess cerebral compliance, even with pathological arterial blood pressure inputs such as heart failure.
It is generally assumed that the analysis of intracranial pressure (ICP) waveforms could be used for the detection of multiple cerebral pathophysiologies. A main obstacle for the analysis of ICP waveforms is given by the large variation of their generating signal, the arterial blood pressure (ABP). Using extended principal component analysis (PCA) we show that it is possible to distinguish between ICP waveforms generated by pathological ABP waveforms, e.g. in the case of a heart failure, without loosing information about the state of the cerebral compliance. We also create a dataset for ICP pulse classification that can be used to train models to distinguish between an intact and diminished cerebral compliance, as a function of the ICP-generating ABP pulse. As a baseline for classification we build a fully convolutional network (FCN) that reaches high performance with relatively few data.
No takes yet. Share an insight, caveat, or question.
Vrabie et al. (2020) studied Traumatic Brain Injury (TBI) and Subarachnoid Hemorrhage (SAH) (n=7). Fully Convolutional Network (FCN) for ICP waveform classification was evaluated on Classification accuracy on a hold-out test set. A fully convolutional network trained to classify intracranial pressure waveforms based on cerebral compliance and arterial blood pressure morphology achieved 93.1% accuracy on a hold-out test set.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: