Model-based method in understanding patient-ventilator interaction

Mechanically ventilated respiratory failure patients risk experiencing desynchronisation with ventilator support. This desynchronisation is known as patient-ventilator asynchrony (PVA). Frequent occurrence of PVA during MV results in poor patient-ventilator interaction, leading to further lung injury, increased patient dependency on MV, and increased mortality. There are various types of PVA, and it is difficult to detect them individually. Thus, there is a need for a method to automatically classify types of PVA to improve the understanding of PVA and its effect towards patient-ventilator interaction.

In this research, we seek to develop novel model-based methods to classify different PVA types and quantify their severity. Machine learning or model-based methods potentially provide a real-time assessment of the patient’s PVA and how the patient interacts with the ventilator. A benchmark PVA dataset will be curated from our retrospective clinical trials consisting of more than 7 million breathing cycles. This dataset will provide a standardised platform for asynchrony analysis. We will develop models to detect different types of PVA and quantify their severity. Finally, we will evaluate the severity level of PVA in existing patient data to develop an adjustment MV protocol.

This research will utilise big data collected from our system to continuously monitor the PVA during MV to further elucidate PVA’s role in patient-ventilator interaction. This research will establish a unique and new solution to better understand the effect of PVA, and improve patient-ventilator interaction. Importantly, better management of MV and reduced patient dependency on MV would enable savings of up to RM2,800 per patient day, a significant economic incentive to pursue this research. Finally, this proposal addresses a major problem, taking advantage of the unique available dataset and existing, strong collaboration ties. The knowledge and collaborative experiences gained can be extended to other biomedical research addressing costly, complex clinical problems.

To learn more, please visit: https://lnkd.in/gSf2sEU9

In this research, we seek to recruit potential postgraduate candidates (Master and/ or PhD) with the following task:

Key Responsibilities:

1) Real-time data process device development and PVA data curation

2) Development of a model-based and/ or machine-learning model and incorporating them in the device

3) Conduct experimental trial to evaluate PVA severity and development of MV adjustment protocols.

Collaborate with a multidisciplinary team of engineers, clinicians, and researchers.

Qualifications:

A degree (or with a Master) in either Biomedical Engineering, Mechanical Engineering, Mechatronics Engineering, or other related field.

Strong background in physiological modelling, machine learning, and data analysis.

Proficiency in programming languages such as Python or MATLAB

Excellent problem-solving skills and attention to detail.

Strong written and verbal communication skills.

Benefits:

Opportunity to work on cutting-edge research with significant real-world impact.

Access to state-of-the-art facilities and resources.

Collaboration with leading experts in the field.

Competitive stipend and tuition coverage.

Interested candidates should submit the following to Chiew.yeong.shiong@monash.edu:

A cover letter detailing your research interests and relevant experience.

A CV highlighting your academic and professional achievements.

Contact information for three academic or professional references.

Join us in revolutionizing ICU treatment through innovative model-based engineering and machine-learning approaches!