Biomedical optics provides non-invasive tools for investigating tissue structure and physiology by exploiting the interaction of light with biological media. Among these techniques, time-domain near-infrared spectroscopy (TD NIRS) and diffuse correlation spectroscopy (DCS) enable the quantification of tissue oxygenation and blood flow, respectively. However, their use in heterogeneous tissues such as skeletal muscle is challenged by model-dependent biases and limited signal-to-noise ratio. This thesis presents methodological and technological advances in TD NIRS and DCS aimed at improving the quantitative assessment of skeletal muscle hemodynamics and perfusion in the context of aging. A novel hybrid TD NIRS/DCS device was developed and extensively characterized on calibrated phantoms and in-vivo, demonstrating state-of-the-art performance, excellent stability, and suitability for long-term clinical studies. In parallel, TD NIRS and DCS measurements acquired within the Trajector-Age study using an established optical system were analyzed to investigate muscle oxygenation and blood flow in aging and physically active populations. On top of offering novel insights into the physiology of aging, these measurements also revealed a strong influence of adipose tissue thickness and protocol variability on optical biomarkers. Motivated by these findings, improved TD NIRS data analysis strategies were developed based on Monte Carlo simulations and validated in-vivo, enabling correction of geometric biases in layered tissues and extending the range of reliable measurements. Additionally, a deep-learning-based denoising framework was introduced for DCS, substantially enhancing signal quality and the precision of blood flow estimation while remaining compatible with conventional analysis pipelines. Overall, this work advances the methodological robustness and interpretability of diffuse optical techniques for studying skeletal muscle aging and supports their broader translational application in clinical and physiological research.
L'ottica biomedica fornisce strumenti non invasivi per studiare la struttura e la fisiologia dei tessuti sfruttando l'interazione della luce con i mezzi biologici. Tra le sue tecniche, la spettroscopia nel vicino infrarosso risolta nel tempo (TD NIRS) e la spettroscopia di correlazione diffusa (DCS) permettono di quantificare rispettivamente l'ossigenazione tissutale e il flusso sanguigno. Tuttavia, il loro utilizzo in tessuti eterogenei come i muscoli scheletrici è spesso ostacolato da distorsioni dovute al modello di analisi dei dati e dal limitato rapporto segnale-rumore. Questa tesi presenta avanzamenti metodologici e tecnologici per le tecniche TD NIRS e DCS mirati a migliorare la stima quantitativa dell'emodinamica e della perfusione dei muscoli scheletrici nel contesto dell'invecchiamento. Un innovativo dispositivo ibrido TD NIRS/DCS è stato sviluppato e caratterizzato in maniera estensiva sia su fantocci calibrati che in-vivo, dimostrando performance allo stato dell’arte, stabilità eccellente, e idoneità per studi clinici a lungo termine. Parallelamente, le misure TD NIRS e DCS acquisite nell’ambito dello studio Trajector-Age usando un sistema ottico validato sono state analizzate per investigare l’ossigenazione muscolare e il flusso sanguigno in popolazioni con età e livelli di attività fisica diversi. Oltre che rivelare nuovi aspetti della fisiologia dell’invecchiamento, queste misure hanno anche dimostrato la forte influenza dello spessore dello strato adiposo e della variabilità del protocollo di misura sui biomarker ottici. Motivati da tali scoperte, sono state sviluppate e validate migliori strategie di analisi dati TD NIRS basate su simulazioni Monte Carlo, permettendo di correggere distorsioni di natura geometrica in tessuti stratificati e di estendere il dominio di affidabilità delle misure. Inoltre, è stato introdotto un sistema di rimozione del rumore per dati DCS basato su deep learning, che migliora significativamente la qualità del segnale e la precisione della stima del flusso sanguigno mantenendo la compatibilità con metodi di analisi tradizionali. Complessivamente, il presente lavoro costituisce un avanzamento della robustezza metodologica e dell’interpretabilità per le tecniche di ottica biomedica applicate allo studio dell’invecchiamento dei muscoli scheletrici, e supporta la loro applicazione traslazionale più ampia alla ricerca clinica e fisiologica.
Diffuse optics for the clinic: techniques, analysis and device engineering in muscle aging
NABACINO, MARCO
2025/2026
Abstract
Biomedical optics provides non-invasive tools for investigating tissue structure and physiology by exploiting the interaction of light with biological media. Among these techniques, time-domain near-infrared spectroscopy (TD NIRS) and diffuse correlation spectroscopy (DCS) enable the quantification of tissue oxygenation and blood flow, respectively. However, their use in heterogeneous tissues such as skeletal muscle is challenged by model-dependent biases and limited signal-to-noise ratio. This thesis presents methodological and technological advances in TD NIRS and DCS aimed at improving the quantitative assessment of skeletal muscle hemodynamics and perfusion in the context of aging. A novel hybrid TD NIRS/DCS device was developed and extensively characterized on calibrated phantoms and in-vivo, demonstrating state-of-the-art performance, excellent stability, and suitability for long-term clinical studies. In parallel, TD NIRS and DCS measurements acquired within the Trajector-Age study using an established optical system were analyzed to investigate muscle oxygenation and blood flow in aging and physically active populations. On top of offering novel insights into the physiology of aging, these measurements also revealed a strong influence of adipose tissue thickness and protocol variability on optical biomarkers. Motivated by these findings, improved TD NIRS data analysis strategies were developed based on Monte Carlo simulations and validated in-vivo, enabling correction of geometric biases in layered tissues and extending the range of reliable measurements. Additionally, a deep-learning-based denoising framework was introduced for DCS, substantially enhancing signal quality and the precision of blood flow estimation while remaining compatible with conventional analysis pipelines. Overall, this work advances the methodological robustness and interpretability of diffuse optical techniques for studying skeletal muscle aging and supports their broader translational application in clinical and physiological research.| File | Dimensione | Formato | |
|---|---|---|---|
|
PhD_Thesis_Nabacino_final.pdf
non accessibile
Descrizione: Tesi
Dimensione
23.52 MB
Formato
Adobe PDF
|
23.52 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/255617