Poster

  • PS01.5
  • ePoster

Automating intracranial pressure transducer position identification

Beitrag in

Joint Meeting

Posterthemen

Mitwirkende

Samir Ruxmohan (Dallas, TX / US), Michael Alec Gomba (Dallas, TX / US), Dawit Measho (Dallas, TX / US), Emerson Nairon (Dallas, TX / US), Lindsay McEver (Dallas, TX / US), Brittany Doyle (Dallas, TX / US), Jade Marshall (Dallas, TX / US), Professor DaiWai Olson (Dallas, TX / US), Dick Moberg (Dallas, TX / US)

Abstract

Abstract-Text (inkl. Referenzen und Bildunterschriften)

OBJECTIVE: The position of an external ventricular drain (EVD) stopcock transducer is a determinant of the accuracy of intracranial pressure (ICP). This study validates the feasibility of a stopcock position sensor (SPS) and annotation hub to provide continuous data acquisition (CDA) for accurate ICP data.

BACKGROUND: The impact of environmental and contextual factors on physiological data such as ICP is well known. For example, when the stopcock is in the open-to-drain position, the ICP reading is neither reliable nor accurate. In patients with acute acquired brain injury this data is often recorded manually. Automating the capture of contextual data will improve the quality of the data for post hoc analysis including machine learning. With respect to ICP, identifying the invalid segments of data where the stopcock is positioned to "drain CSF" vs "monitor ICP" is essential.

METHODS: We created an SPS that collects contextual data and generates time-synchronized annotations linked to physiologic data. In a prospective observational trial we evaluate the efficacy of an SPS to accurately identify when a transducer is open to drain vs monitoring ICP. After consent, eligible subjects with an EVD in situ were monitored for up to 24 hours using the SPS.

RESULTS: Phase 1 and 2 prototypes of the SPS were evaluated between January 2022 and April 2023. Between April 2023 and September 2023 the SPS was placed on 12 patients in the Neuroscience Intensive Care Unit. Patients were admitted for subarachnoid hemorrhage (5), intraventricular hemorrhage (2), mass lesion (4), and congenital Hydrocephalus (1). All patients completed the trial and provided linked readings to ICP data and there have been zero adverse events or complications during the study.

CONCLUSIONS: Automated CDA is feasible in the NSICU for patients with acute brain injury. The use of annotated CDA data will enhance the clinical utility of ICP monitoring by providing higher accuracy and reliability of ICP readings. Additional phases of this study hope to further validate initial findings.

    • v1.20.0
    • © Conventus Congressmanagement & Marketing GmbH
    • Impressum
    • Datenschutz