ALIVE

Advanced Lung Image processing for personalized mechanical VEntilation

Reconstructed electrical impedance tomography (EIT) image of lung aeration, with pixel-level impedance-time signals reflecting tidal ventilation at each location. Post-processing of signals varies depending on location, e.g. relative to the heart (A close to heart, B further away). Copyright, Annemijn Jonkman.

Mechanical ventilation is a double-edged sword: while life-saving for patients with acute respiratory failure, it may also worsen lung injury and induce respiratory muscle trauma. Therefore, we need simple and reliable bedside techniques to provide lung-protective ventilation tailored to the individual patient’s needs. A very promising technology is Electrical Impedance Tomography (EIT). EIT is gaining popularity worldwide as non-invasive bedside tool to image the lungs in real-time. However, methods to implement EIT information in clinical practice are yet lacking due to technological challenges: clinicians and researchers are dependent on equipment vendors with closed-source software, which limits their ability to manage diverse datasets, leading to poor interoperability and reproducibility of research findings. In this project, we have developed a robust EIT data processing and analysis workflow. This will allow us to fully exploit the clinical benefits of EIT and will promote (inter-)national research projects aimed at optimizing mechanical ventilation in the ICU. The Erasmus MC team has extensive experience in EIT research and advanced analyses; the e-Science center is complementary to our expertise and crucial for knowledge translation into a robust and reusable software workflow.

The core and main output of the ALIVE project is eitprocessing (https://github.com/EIT-ALIVE/eitprocessing), a Python package that enables loading, processing and analyzing different datasets: EIT data from different vendors as well as mechanical ventilation and respiratory waveforms. The software includes filters and analysis methods that are used for current and future research. eitprocessing is modular, which makes expansion with own methods and algorithms, but also expansion to other domains, as easy as possible. Second, eit_dash (https://github.com/EIT-ALIVE/eit_dash) is a concept GUI implementation that, after further development, will make the software usable for a wider audience of EIT users without programming experience. With eitprocessing we have strongly improved the EIT analysis workflow of our team. We have been able to bundle, expand, professionalize and make available to everyone the code used by our research group. We have also published a paper on new EIT signal filtering methods that we developed and compared (Wisse JJ et al., Improved filtering methods to suppress cardiovascular contamination in electrical impedance tomography recordings. Physiol Meas. 2024 May 21;45(5). doi: 10.1088/1361-6579/ad46e3) and a paper in the Journal of Open Source Software is expected early 2025. Our project and output enables not only us, but the entire research community working with EIT, to take the quality, efficiency and effectiveness of data analysis to a higher level. Beyond the ALIVE team, many Technical Medicine students have contributed to development of the software through their internships within our department, thereby promoting open science and digital research skills.

We have a strong (inter)national network, with whom we have initiated community-building around the software tool and stimulated reproducible research. We have organized a Lorentz Center workshop (https://www.lorentzcenter.nl/eit-during-mechanical-ventilation-a-standardized-innovative-workflow.html), where 28 participants from seven major Dutch medical centers/universities and several renowned institutes abroad participated. We have full support from this community and received endorsements from private partners (EIT vendors). With this workshop, we have gathered recommendations for (future) developments and published a consensus paper on the standardized use of EIT in a clinical journal (Scaramuzzo G et al., Electrical impedance tomography monitoring in adult ICU patients: state-of-the-art, recommendations for standardized acquisition, processing, and clinical use, and future directions. Crit Care. 2024 Nov 19;28(1):377. doi: 10.1186/s13054-024-05173-x). Furthermore, we performed a qualitative study (survey and focused group discussions) on how to better implement advanced respiratory monitoring with EIT in clinical practice (Wisse JJ et al., Clinical implementation of advanced respiratory monitoring with esophageal pressure and electrical impedance tomography: results from an international survey and focus group discussion. Intensive Care Med Exp. 2024 Oct 21;12(1):93. doi: 10.1186/s40635-024-00686-9).

We have certainly met our initial project’s objectives but have not finished the software (especially the GUI eit_dash) within the dedicated time, due to the complexity of the data, processing steps and software infrastructure. The collaboration with RSEs was key to developing a modular framework for sustainable software that we could build upon in future projects. Our next step is to continue using and improving eitprocessing within our research team and to provide assistance to analyzing EIT data in international collaborations.

In December 2024, during a festive ceremony of the Erasmus University Rotterdam, our team was awarded the EUR Open and Responsible Science award for the ALIVE software, in the category Open Research. The jury members commended our project for “its exemplary commitment to open science through the creation of freely accessible tools, fostering international collaboration, and advancing consensus in the field, setting a new standard for open, reproducible, and responsible research in healthcare innovation.” This recognition celebrates the success of the project and outstanding contribution of our team to open science and creating impact.

Full list of deliverables: