# Respiratory pressure and flow data collection device providing a framework for closed-loop mechanical ventilation

## TL;DR

This paper introduces a sensor device for mechanical ventilation that supports personalized treatment by integrating with digital twin models and closed-loop systems.

## Contribution

The novel contribution is a low-error sensor system that enables personalized mechanical ventilation through integration with digital twin models and BLE communication.

## Key findings

- The device has a mean flow data error of 3.2% and an operating range of ±50.8 cmH2O.
- It uses BLE communication to connect with ventilators and digital twin models for closed-loop control.
- The system supports future development of personalized ventilation treatment in ICU settings.

## Abstract

This article details a pressure and flow sensor system device which enables a framework for the research and development of personalized mechanical ventilator support in a closed-loop or semi-closed-loop control system, where the measurements from this device could be hooked to digital twin models and any ventilator allowing open control.

In current practice, patient response to mechanical ventilation is highly variable. Furthermore, current weaning best-practice relies on clinical experience which can lead to variability and inequality in both care and health outcomes. Personalized care can improve these inequalities in care due to patient variability when combined with digital twin models, which simulate physiology based on patient specific data, by improving the level of care possible in the ICU (Intensive Care Unit), regardless of clinician experience and/or patient variability.

The device consists of two 3D printed custom Venturis and a Y-piece, with differential pressure sensors measuring gauge, inhalation, and exhalation pressure at the patient. The sensor system has an operating range of ±50.8 cmH2O and a mean error in flow data of 3.2%. The system uses BLE (Bluetooth Low Energy) communication between ESP32-S3 development boards to facilitate the closed loop framework. Within this loop, pressure data is sent to a digital beside sheet, which runs digital twin protocols and sends commands to a BLE controlled ventilator. Overall, this device allows the future development and validation of personalized mechanical ventilation treatment through integration with digital twin models.