Giulia Acquaviva
I believe I deserve the KIVI EngD Award because my EngD project combines technological
innovation, evidence-driven engineering and a clear potential for societal and clinical impact.
My research focused on designing an automated tool for detecting uterine contractions from
electrohysterographic (EHG) signals recorded through electrodes placed on the maternal
abdomen. Using an ensemble of machine learning models, I developed a method capable of
identifying uterine contractions during both pre-labor and labor, with the aim of contributing
to more accurate and accessible maternal monitoring.
The innovation of my work lies not only in the application of signal processing techniques,
such as machine learning, to uterine activity, but particularly in addressing the pre-labor
phase, which remains relatively underexplored in the literature. While most existing
approaches focus on labor, understanding uterine activity before the onset of labor could
provide valuable information for improving monitoring and supporting earlier and more
informed clinical decisions. I developed and trained the algorithm using a publicly available
dataset and then validated it on an independent private dataset, achieving performance
comparable to existing methods reported in the literature. This additional validation
strengthened the robustness and practical relevance of the proposed approach.
The potential impact of this work extends beyond the algorithm itself. Throughout my EngD, I
also collaborated with Sencure B.V. (R&D) within the larger NewLife project, which aims to
improve the safety of new lives while supporting maternal well-being. As part of my work
package, the objective was to contribute to the design and development of a maternal
abdominal patch capable of simultaneously acquiring different types of physiological and
imaging data, including ECG, EHG, ultrasound images and IMU measurements. Within this
multidisciplinary effort, I worked on evaluating the quality of innovative electrodes with
different morphologies and textile substrates, assessing their potential for integration into the
patch. This experience allowed me to connect algorithm development with the practical
challenges of biomedical hardware and to collaborate closely with an industrial partner
towards a common technological goal.
Beyond the technical results, I am most proud of the personal commitment, curiosity, and
dedication I brought to this project. I invested significant effort in understanding complex
physiological data, developing and evaluating the algorithms, and critically interpreting their
performance. At the same time, I actively engaged with researchers and industry partners,
learning how different perspectives can contribute to solving complex biomedical challenges.
Most importantly, I genuinely enjoyed the research process and became deeply passionate
about the possibility of using engineering to address real-world healthcare needs.
For these reasons, I believe my project reflects the values of the KIVI EngD Award: innovation
through a novel application of ensemble-based machine learning, impact through the pursuit
of more accurate, less invasive and accessible maternal monitoring, and leadership and
collaboration through interdisciplinary research connecting academia, engineering, and
industry.

