THESIS PROPOSAL 2026/2027

 

Development of Computational Models of the Respiratory System for Mechanical Ventilation Simulations

Mechanical ventilation is an essential life-support therapy for patients who are unable to maintain adequate spontaneous breathing. However, the interaction between the ventilator and the respiratory system is highly complex and depends on several factors, including airway geometry, tissue mechanical properties, lung heterogeneity, disease conditions, and ventilator settings. Computational models can provide a powerful tool to investigate these interactions, simulate different physiological and pathological conditions, and support the development of strategies aimed at improving ventilation while reducing the risk of ventilator-induced lung injury.

Several computational models of the respiratory system are currently being developed and used within the laboratory. These models range from detailed representations of the bronchial tree and lung tissue to more compact models specifically designed to reproduce particular aspects of respiratory mechanics. They can be used to simulate conventional mechanical ventilation, oscillatory pressure or flow signals, respiratory impedance measurements, and the effects of changes in airway resistance, lung compliance, tissue properties, or regional obstruction.

The thesis project will focus on the further development and application of mathematical and computational models of the respiratory system for mechanical ventilation simulations. Depending on the specific project and on the student’s interests, the activity may involve extending and improving existing MATLAB models, implementing or adapting respiratory models in Python, or exploring alternative simulation environments such as Simscape to develop more targeted models of specific components or phenomena of the respiratory system.

Possible activities include studying the physiological and mechanical principles underlying respiratory models, analysing existing modelling approaches from the scientific literature, implementing equations and numerical simulation algorithms, modifying model structure and parameters, developing simulation pipelines, and comparing model outputs under different physiological, pathological, and ventilatory conditions. 

Students will mainly work on computational activities, including programming, data analysis, model validation, parameter studies, and visualization of simulation results. Depending on the specific thesis topic, experimental or clinical data available in the laboratory may also be used to compare model predictions with real measurements and to guide model refinement.

The expected duration of the thesis is between 10 and 14 months, depending on the student’s level of commitment and on the complexity of the selected modelling activity. Since the project is predominantly computational, most of the work can be carried out remotely. Students are nevertheless welcome to work in the laboratory as much as they wish, allowing regular interaction with tutors and other researchers and facilitating discussions on model development, interpretation of results, and integration with ongoing experimental activities.

For further information about this project, please contact Matteo Mentasti: matteo.mentasti@polimi.it

 

Development and Optimization of a System to Reduce Air Pollution Exposure in Newborns

Newborns are particularly vulnerable to air pollution because of their developing respiratory system, high ventilation rate relative to body size, and prolonged time spent in confined environments such as cribs, bassinets, and neonatal care areas. Pollutants such as particulate matter, volatile organic compounds, and other airborne contaminants may therefore represent an important source of exposure during the first stages of life. Technologies capable of locally improving air quality around the newborn could help reduce this exposure without requiring the treatment of the entire surrounding environment.

A dedicated experimental system has already been developed in the laboratory to investigate this problem. The setup includes a newborn-sized mannequin equipped with environmental sensors, a breathing simulation system, and a prototype device designed to filter and deliver cleaner air in the region surrounding the infant. The system can be used to reproduce realistic exposure conditions and to measure how changes in air quality, airflow patterns, device configuration, and environmental conditions affect the air effectively reaching the newborn.

The thesis project will focus on the further development, optimization, and experimental evaluation of this system, with particular attention to improving the performance of the clean-air delivery and filtration device. Possible activities include redesigning components of the air distribution system, optimizing airflow rates and outlet geometries, selecting and integrating filtering technologies, improving the positioning of the device relative to the newborn, and evaluating different operating strategies to maximize pollutant reduction while maintaining suitable environmental and comfort conditions.

Students will work on activities such as studying sensors and air filtration technologies, designing and prototyping mechanical components using CAD and 3D printing, integrating electronic components and actuators, developing control and data acquisition software, and improving the overall experimental setup. A significant part of the thesis will be devoted to experimental campaigns in which the system is tested under different pollution and ventilation conditions, measuring variables such as particulate matter, gaseous pollutants, temperature, humidity, and airflow in order to quantify the reduction in exposure achieved by different configurations.

The collected data will be analysed to compare the effectiveness of alternative designs and operating conditions, identify the parameters that most strongly affect the newborn’s exposure, and guide the optimization of the system. Depending on the specific direction of the project, the work may also include the development of control strategies that automatically adapt the operation of the device based on measured environmental conditions.

The expected duration of the thesis is between 10 and 14 months, depending on the student’s level of commitment and the inherent unpredictability of experimental work. Since the project involves sensors, prototypes, and experimental measurements, a substantial part of the activity will be carried out in the laboratory. Students will work in close contact with tutors and other researchers during system development, experimental planning, and data analysis.

For further information about this project, please contact Matteo Mentasti: matteo.mentasti@polimi.it

 

Development of a System to Evaluate the Health Impacts of Beddings and Sleeping Environments for Newborns

Mattresses and sleeping environments for infants and young children can pose potential health risks, particularly when their geometry, materials, or interaction with the infant’s face interfere with normal breathing. For example, soft or poorly ventilated surfaces may contribute to partial airway obstruction or create confined spaces in which exhaled CO2 accumulates, increasing the likelihood of rebreathing. Mattress materials, shapes, covers, and environmental conditions can all influence these effects. Therefore, experimental systems capable of reproducing the interaction between an infant and the sleeping surface are important for objectively assessing the safety and performance of different mattresses and bedding configurations.

A dedicated experimental setup is currently available in the laboratory. It includes 3D-printed infant heads derived from MRI scans, a neonatal breathing simulator, and multiple sensors for monitoring variables such as respiratory flow and CO2 exchanged with the surrounding environment. The system is already functional, but further development is required to improve its robustness, integration, and measurement capabilities before it can be systematically used for mattress characterization.

The thesis activity will therefore involve both hardware and software development and, importantly, the experimental use of the completed setup to evaluate different mattresses and sleeping configurations. Possible developments include the integration of additional sensors and hardware components, improvement and centralization of data acquisition and control on a single microcontroller, development of graphical interfaces for real-time monitoring and data logging, and design of mechanical components such as chambers, supports, and connectors using CAD and 3D printing.

Students will work on activities such as studying sensor and component datasheets, writing and testing firmware, designing and assembling electronic circuits, integrating new sensors and actuators, developing graphical interfaces for data visualization, and improving the overall experimental setup. Once the system has been consolidated, a significant part of the thesis will be devoted to planning and performing experimental measurements on different mattresses and bedding configurations, analysing variables such as airflow and CO2 rebreathing, and comparing the results in order to identify differences related to mattress geometry, materials, and test conditions.

The expected duration of the thesis is between 10 and 14 months, depending on the student’s level of commitment and the inherent unpredictability of experimental work. Students will be required to carry out most of their activities in the laboratory, as the project involves experimental equipment and hardware available only on-site. Working in the laboratory will also allow continuous interaction with tutors and other researchers, facilitating troubleshooting, experimental planning, and teamwork.

For further information about this project, please contact Matteo Mentasti: matteo.mentasti@polimi.it

 

Development of a Mechanical Ventilation System Based on Surface Electromyography of the Diaphragm for Preterm Infants

Prematurity, defined as birth before the 37th week of gestation, is a critical issue in neonatal health. Being born too early disrupts the normal maturation of vital organs, particularly the lungs, often leading to severe complications. Preterm infants face a much higher risk of respiratory problems compared to full-term infants. These include conditions like Infant Respiratory Distress Syndrome (iRDS) and Bronchopulmonary Dysplasia (BPD), which frequently require mechanical ventilation (MV) to support breathing. For preterm infants, MV must not only provide effective respiratory support but also synchronize with the infant’s spontaneous breathing to avoid unnecessary strain and complications. Most ventilators currently rely on flow and pressure sensors to detect the infant’s respiratory effort. However, these methods can be inaccurate due to factors like air leaks in the circuit or changes in the mechanical properties of the infant’s lungs. To address these issues, researchers in Toronto developed Neurally Adjusted Ventilator Assist (NAVA). NAVA uses signals from the diaphragm, the primary respiratory muscle, to control the ventilator. These signals are captured via a special esophageal catheter equipped with electrodes. While NAVA improves synchronization significantly, it has drawbacks: the catheter is invasive, expensive, works only with specific ventilators, and increases the risk of infection.

This raises a key research question: how can we achieve the precision of NAVA while eliminating its invasiveness? This thesis explores the use of surface electromyography of the diaphragm as a non-invasive alternative for detecting spontaneous breathing in preterm infants. The objective is to develop and optimize a real-time system that uses surface electromyography signals to trigger mechanical ventilation, leveraging machine learning or traditional signal processing methods.

The thesis workflow includes four main steps: developing a real-time algorithm for signal processing, implementing basic hardware to run the algorithm in real-time, validating the system in vitro at Techres Lab, Politecnico di Milano, conducting a pilot clinical study in collaboration with Amsterdam UMC.  The expected duration of the thesis is between 10 and 14 months. Laboratory attendance is strongly recommended. Additionally, there may be opportunities to travel for short to medium periods as part of project activities or collaborations.

For further information, please contact Ilaria Girimonte at ilaria.girimonte@polimi.it.

 

 

OLD MASTER THESIS

View on POLITESI