End-to-End Automatic Morphological Classification of Intracranial Pressure Pulse Waveforms Using Deep Learning.

Cyprian Mataczynski,Magdalena Kasprowicz,Agnieszka Uryga,Malgorzata Burzynska,Agnieszka Kazimierska,Andrzej Rusiecki

doi:10.1109/jbhi.2021.3088629

Abstract

Mean intracranial pressure (ICP) is commonly used in the management of patients with intracranial pathologies. However, the shape of the ICP signal over a single cardiac cycle, called ICP pulse waveform, also contains information on the state of the craniospinal space. In this study we aimed to propose an end-to-end approach to classification of ICP waveforms and assess its potential clinical applicability. ICP pulse waveforms obtained from long-term ICP recordings of 50 neurointensive care unit (NICU) patients were manually classified into four classes ranging from normal to pathological. An additional class was introduced to simultaneously identify artifacts. Several deep learning models and data representations were evaluated. An independent testing dataset was used to assess the performance of final models. Occurrence of different waveform types was compared with the patients' clinical outcome. Residual Neural Network using 1-D ICP signal as input was identified as the best performing model with accuracy of 93% in the validation and 82% in the testing dataset. Patients with unfavorable outcome exhibited significantly lower incidence of normal waveforms compared to the favorable outcome group even at ICP levels below 20 mm Hg (median [first-third quartile]: 9 [1-36]% vs. 63 [52-88] %, p = 0.002). Results of this study confirm the possibility of analyzing ICP pulse waveform morphology in long-term recordings of NICU patients. Proposed approach could potentially be used to provide additional information on the state of patients with intracranial pathologies beyond mean ICP.

Highlights

I NTRACRANIAL pressure (ICP) is frequently monitored in patients with brain pathologies as elevated ICP puts the patient at risk of cerebral ischemia or herniation of structures within the cranial vault
Residual Neural Network using 1-D ICP signal as input was identified as the best performing model with accuracy of 93% in the validation and 82% in the testing dataset
Results of this study confirm the possibility of analyzing ICP pulse waveform morphology in long-term recordings of neurointensive care unit (NICU) patients

Summary

Introduction

I NTRACRANIAL pressure (ICP) is frequently monitored in patients with brain pathologies as elevated ICP puts the patient at risk of cerebral ischemia or herniation of structures within the cranial vault. The clinical state of the patient cannot be fully characterized by mean ICP alone as the changes in intracranial volume which influence ICP can be buffered to a certain degree [1]. As long as the compensatory mechanisms for adapting to increased volume are intact and the compliance of the system is normal, small increases in intracranial volume result in small increases in ICP. When brain compliance is decreased and the compensatory mechanisms are exhausted, small increases in volume lead to disproportionately large increases in ICP. The pressure–volume curve, i.e., the exponential relationship between pressure and volume in the intracranial space, has long been regarded as a potential source of useful information on the state of the craniospinal system [3]. Despite promising results published on that subject, direct measurement of compliance has, proven difficult to implement in clinical practice on a larger scale [2]

Objectives

Results

Discussion

Conclusion