Processant ...

Hydrocephalus is a pathology characterized by the abnormal accumulation of cerebrospinal fluid (CSF) within the brain's ventricles, leading to alterations in intracranial pressure (ICP). Within this condition, normal-pressure hydrocephalus (iNPH) is a particular case where ICP remains within normal ranges, yet patients may still exhibit symptoms consistent with hydrocephalus. One of the main challenges in diagnosing NPH is determining the exact position of a patient on the pressure-volume curve, as incorporating a second parameter alongside mean ICP can provide valuable additional insights for diagnosis. A key concept in this context is cerebral compliance, which describes the brain's ability to accommodate volume changes with minimal pressure variations. This parameter is closely related to the pulsatile component of ICP, which can be extracted from the cardiac waveform embedded within the ICP signal. Some clinical monitors have recently begun incorporating this parameter, but a critical issue arises: some devices calculate it in the time domain, while others use the frequency domain via the Fourier Transform. This methodological discrepancy results in inconsistent values that are treated as equivalent, which could affect clinical decisions. This TFG focuses on analyzing this issue by comparing the pulsatile component values obtained in both domains for a series of iNPH patients. The practical part involves processing anonymized real ICP data provided by the Neurosurgery Department at Vall d’Hebron Hospital, using Python within the Jupyter Notebook environment in Anaconda, applying signal processing techniques to extract and compare the pulsatile component in both domains. The methodology includes a review of existing approaches, signal pre-processing through filtering techniques to remove artifacts and noise, and extraction of the pulsatile component in the time domain via peak detection and cyclic variations analysis. The frequency domain analysis involves applying Fast Fourier Transform (FFT) to obtain the spectral content and extracting the fundamental harmonic amplitude as an indicator of the pulsatile component, complemented by spectrogram analysis to examine variations over time. The final objective is to quantify the differences between these methods, explore the mathematical causes behind these discrepancies, and assess their potential clinical implications, aiming to contribute to a more standardized approach for evaluating pulsatility in normal-pressure hydrocephalus diagnosis.