Abstract:
Objective. Intracranial pathologies are treated using mean intracranial pressure (ICP). The ICP pulse waveform, the shape of the ICP signal over one cardiac cycle, also indicates craniospinal space state. This study proposed an end-to-end ICP waveform classification method and evaluated its clinical applicability. Methods. Long-term ICP recordings of 50 neurointensive care unit (NICU) patients were manually classified into four classes from normal to pathological. An additional class identified artifacts simultaneously. Deep learning models and data representations were assessed. Final models were tested on an independent dataset. Waveform types and clinical outcomes were compared. Results. Residual Neural Network with 1-D ICP signal input performed best with 93% accuracy in validation and 82% in testing dataset. Even at ICP levels below 20 mm Hg, patients with unfavorable outcomes had a lower incidence of normal waveforms than those with favorable outcomes (9 [1–36]% vs. 63 [52–88]%, p = 0.002). Conclusions. This study shows that ICP pulse waveform morphology can be analyzed in long-term NICU patient recordings. The proposed method may provide additional information on intracranial pathology patients beyond mean ICP.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
