Flow encoding magnetic resonance imaging: development of advanced computational methods for the assessment of in vivo fluid dynamics

Recent developments in flow encoding magnetic resonance imaging consisted in the definition of novel 3D time-resolved phase-contrast cardiac magnetic resonance pulse sequences with three-directional velocity-encoding. These imaging technique, namely 4D Flow, allowed for capturing complex 3D time-resolved velocity patterns for retrospective investigations of any location within a volume of interest, without any restriction to predefined 2D imaging planes. Several studies proved the potential of 4D Flow in comprehensively describing the in vivo fluid dynamics on a patient-specific basis, so to overcome the inherent limitations of in silico modeling approaches and in vitro testing set-ups. Nevertheless, from a technical standpoint, the reliability of 4D Flow sequences is still under investigation, considering their current limitations in terms of spatial resolution, ranging from 1.5-3 mm; temporal discretization of the cardiac cycle, resulting in 20-30 phases (i.e., about 30-50 ms); velocity to noise ratio; and lack of standardized post-processing approaches. The aforementioned limitations hamper the reliability of estimated hemodynamic markers and consequently the application of 4D Flow sequences in the clinical scenario, especially considering those clinical scenarios where the biomechanical implications of altered fluid dynamics on the onset and progression of cardiovascular diseases are yet unknown. In this specific context, three major challenges were tackled: i) to improve and assess the reliability of 4D Flow derived markers, e.g., wall shear stresses, through the definition of realistic in silico set of 4D Flow datasets, considered as benchmark for innovative algorithms; ii) to test the relevance of 4D Flow in the analysis, risk stratification and monitoring of diseases affecting the thoracic aorta; iii) to investigate the feasibility of 4D Flow in quantifying the complex 3D fluid dynamics of blood processing devices, in sight of being a valid tool for devices optimization. Considering the first topic, two new methods were developed for the quantification of wall shear stress (WSS) acting on the aortic, aiming at improving the estimation of peak values and the identification of high-stress regions. The extensive benchmarking highlighted the need for data spatial resolution at least comparable to current clinical guidelines, the low sensitivity of the methods to data noise, and their capability, when used jointly, to compute more realistic peak WSS values as compared to state-of-the-art methods. In light of the improved estimation of shear forces acting on the endothelium of the aortic wall, a specific clinical scenario was taking in consideration to assess the role of WSSs in the progression of aortopathy related to the presence of bicuspid aortic valve and to the associated altered fluid dynamics. To test this hypothesis, 4D Flow was used to analyze the in vivo fluid dynamics in the thoracic aorta of normo-functional BAV patients with no aortic dilation. Despite minor bulk flow disturbances at peak systole, evident alterations of WSS distribution and peak values, and WSS-related indexes were found. Despite these fluid-dynamic alterations, no clinically relevant anatomical remodeling was observed in a three-year follow-up, suggesting that WSS alterations may precede the onset of aortopathy and may contribute to its triggering. As regards the third topic, 4D flow sequences with sub-millimetric spatial resolution and region-dependent velocity encodings were tested on a real device that integrates an oxygenator and a heat exchanger. The effects of fine geometrical features of the device on the local fluid-dynamics were highlighted, these phenomena being unlikely measurable by current in vitro approaches. Also, the effects of non-idealities on the flow field distribution were captured, thanks to the absence of the simplifying assumptions that typically characterize in silico approaches. This evidences proved the ability of 4D Flow mapping in extracting sound information on the fluid-dynamic performances of blood processing devices.

Gli sviluppi recenti della tecnica di risonanza magnetica a contrasto di fase hanno permesso lo sviluppo di una nuova sequenza 3D discretizzata nel tempo con codifica delle fasi di velocità su tre direzioni mutuamente ortogonali. Tale tecnica, denominata 4D Flow, permette di misurare e cogliere pattern 3D complessi del campo fluidodinamico all’interno di un volume di interesse, con la possibilità di navigare tale dato in modo retrospettivo, senza alcuna restrizione di piani 2D precedentemente definiti. Numerosi studi in letteratura hanno dimostrato il potenziale della tecnica di 4D Flow nel descrivere la fluido dinamica paziente-specifica con misure in vivo, così da superare le limitazioni che caratterizzano modelli numerici in silico e set-up in vitro. Ciononostante, l’affidabilità delle sequenze di 4D Flow, dal punto di vista tecnica, è ancora in fase di analisi, considerando le principali limitazioni in termini di: risoluzione spaziale (voxel pari a 1.5-3 mm) e temporale (fasi di circa 30-50 ms), rumore dipendente dalla codifica di velocità, mancanza di approcci standard per l’analisi dei dati. Le limitazioni sopradescritte inficiano la accuratezza delle stime fluidodinamiche derivanti dall’analisi dei dati 4D Flow, così da limitarne l’utilizzo in clinica, in particolare nelle situazioni per le quali vi è ancora incertezza riguardo all’impatto delle implicazioni biomeccaniche dovute ad alterazioni fluido dinamiche sullo sviluppo e sul progredire di numerose malattie cardiovascolari. In questo contesto specifico, tre punti critici sono stati affrontati, al fine di: i) valutare e migliorare la affidabilità della stima di marker fluidodinamici, quali ad esempio wall shear stress, mediante l’utilizzo di innovativi dati 4D Flow artificiali creati per replicare condizioni in vivo, ma conoscendo il dato di riferimento con il quale confrontare le stime ottenute; ii) valutare la rilevanza delle analisi 4D Flow nei processi di diagnosi, stratificazione del rischio e monitoraggio di malattie cardiovascolari del distretto aortico; iii) valutare la fattibilità di utilizzare la tecnica di 4D Flow nel quantificare pattern fluidodinamici 3D all’interno di dispositivi artificiali per il processamento di sangue, con la prospettiva di renderla una tecnica valida nel processo di ottimizzazione di tali dispositivi.