Development and application of quantitative physical methodologies for non-destructive diagnostic characterisation in the fields of cultural heritage and medicine

The present doctoral thesis focuses on the development and application of quantitative physical methodologies for non-destructive diagnostic characterisation. These methodologies operate within two distinct fields that require objective and reproducible information: cultural heritage and medicine. The common thread uniting the research is the design and implementation of data analysis pipelines that translate physical signals into quantifiable diagnostic parameters. Part I focuses on portable low-field Nuclear Magnetic Resonance (NMR) using single-sided devices such as the NMR-MOUSE. In Cultural Heritage, low-field NMR relaxometry characterizes the porous structure of stone and wood non-invasively. Measurements of moisture content and penetration of conservation products enabled the assessment of treatment effectiveness and conservation status, demonstrating the potential of NMR relaxometric profiling for in situ diagnostics and monitoring. In the medical field of articular cartilage, a multi-parametric NMR analysis was performed on ex vivo bovine samples. The combination of different NMR parameters allowed a quantitative discrimination of cartilage zones (i.e., superficial, middle, and deep) and reflected variations in water mobility, proteoglycan content, and collagen structure, serving as indicators of osteoarthritis severity. Part II addresses quantitative analysis of medical Computed Tomography (CT) images. In acute ischemic stroke, a methodology was developed to refine the definition of the ischemic core (irreversibly damaged tissue), through a multi-parametric combination of CT perfusion maps. The results suggest that integrating multiple parameters improves differentiation between core and penumbra (salvageable tissue). A radiomic approach was developed for patients with post-thrombectomy CT hyperdensity. Radiomic features demonstrated their sensitivity in distinguishing hyperdensities due to contrast retention and hemorrhagic transformation, and revealed the role of clinical factors (e.g., renal function) in the hyperdensity formation. In summary, this thesis validates the effectiveness of a quantitative, physics-based approach for extracting accurate and objective diagnostic information, thereby contributing to progress in both Cultural Heritage preservation and advanced medical diagnostics.