Modelli compartimentali per l'analisi di segnali ottenuti tramite DCE MRI per studi di perfusione epatica

Il presente lavoro di tesi si inserisce all'interno di uno studio dal titolo: "Strategia di posizionamento multi-step come approccio pragmatico per ridurre il rischio di encefalopatia epatica post-TIPS (shunt trans-giugulare porto-sistemico intraepatico) in pazienti cirrotici con ascite refrattaria". Il progetto di tesi si è concentrato sull'analisi dei segnali ottenuti tramite DCE MRI, con lo scopo di implementare in ambiente MatLab due modelli differenti (Dual input - Mono compartment e Dual input - Dual compartment) che descrivono la cinetica del tracciante all'interno del sistema vascolare epatico e valutare l'efficacia dei parametri di perfusione associati nella descrizione delle variazioni in termini di microcircolazione introdotte dall'inserimento del TIPS. Inizialmente si sono voluti valutare, tramite simulazione, gli effetti in termini di amplificazione del rumore e stima dei parametri perfusionali dell'approssimazione lineare nella conversione da intensità di segnale MR a concentrazione di mezzo di contrasto. Successivamente, sempre attraverso simulazioni, per entrambi i modelli considerati è stato scelto uno schema di model-fitting e quindi testata l'affidabilità in termini di accuratezza e precisione delle stime dei parametri ottenute in funzione del livello di rumore associato alle curve di intensità di segnale. Parallelamente all'implementazione dei modelli per la stima di parametri di perfusione, sono stati realizzati dei phantom con l'obiettivo di simulare il parenchima epatico prima e dopo l'arrivo del mezzo di contrasto e poter testare la sequenza utilizzata durante l'acquisizione dei dati su paziente. Infine sono stati considerati gli esami di DCE MRI effettuati su un campione di nove pazienti pre e post-TIPS, utilizzando per l'analisi dei segnali entrambi i modelli implementati in fase di simulazione e successivamente valutando le variazioni nel valori associati ai parametri di perfusione introdotte dall'inserimento del TIPS.

Determinazione semplificata delle curve caratteristiche di un compressore centrifugo

Scopo di questa tesi è lo sviluppo di un codice nell'ambiente per il calcolo numerico Matlab atto a simulare il comportamento di un compressore centrifugo. Tale compressore centrifugo è stato studiato per l'utilizzo in un gruppo turbogas per la produzione di energia elettrica. Con l’utilizzo di alcune semplificazioni, è stata tracciata la mappa di tale compressore in intervalli di portata e salti di pressione il più possibile distanti dal funzionamento con condizioni di progetto.

Investigating interocclusal perception in tactile teeth sensibility using symmetric and asymmetric analysis

The purpose of this clinical trial was to determine the active tactile sensibility of natural teeth and to obtain a statistical analysis method fitting a psychometric function through the observed data points. On 68 complete dentulous test persons (34 males, 34 females, mean age 45.9 ± 16.1 years), one pair of healthy natural teeth each was tested: n = 24 anterior teeth and n = 44 posterior teeth. The computer-assisted, randomized measurement was done by having the subjects bite on thin copper foils of different thickness (5-200 µm) inserted between the teeth. The threshold of active tactile sensibility was defined by the 50% value of correct answers. Additionally, the gradient of the sensibility curve and the support area (90-10% value) as a description of the shape of the sensibility curve were calculated. For modeling the sensibility curve, symmetric and asymmetric functions were used. The mean sensibility threshold was 14.2 ± 12.1 µm. The older the subject, the higher the tactile threshold (r = 0.42, p = 0.0006). The support area was 41.8 ± 43.3 µm. The higher the 50% threshold, the smaller the gradient of the curve and the larger the support area. The curves showing the active tactile sensibility of natural teeth demonstrate a tendency towards asymmetry, so that the active tactile sensibility of natural teeth can mathematically best be described by using the asymmetric Weibull function.

Two-dimensional linear-combination model fitting of magnetic resonance spectra to define the macromolecule baseline using FiTAID, a Fitting Tool for Arrays of Interrelated Datasets

To propose the determination of the macromolecular baseline (MMBL) in clinical 1H MR spectra based on T(1) and T(2) differentiation using 2D fitting in FiTAID, a general Fitting Tool for Arrays of Interrelated Datasets.

Curve Decomposition for Large Deflection Analysis of Fixed-Guided Beams With Application to Statically Balanced Compliant Mechanisms

Statically balanced compliant mechanisms require no holding force throughout their range of motion while maintaining the advantages of compliant mechanisms. In this paper, a postbuckled fixed-guided beam is proposed to provide the negative stiffness to balance the positive stiffness of a compliant mechanism. To that end, a curve decomposition modeling method is presented to simplify the large deflection analysis. The modeling method facilitates parametric design insight and elucidates key points on the force-deflection curve. Experimental results validate the analysis. Furthermore, static balancing with fixed-guided beams is demonstrated for a rectilinear proof-of-concept prototype.

Diffusion-weighted MR imaging including bi-exponential fitting for the detection of recurrent or residual tumour after (chemo)radiotherapy for laryngeal and hypopharyngeal cancers

To assess whether diffusion-weighted magnetic resonance imaging (DW-MRI) including bi-exponential fitting helps to detect residual/recurrent tumours after (chemo)radiotherapy of laryngeal and hypopharyngeal carcinoma.

Classification and selection of biomarkers in genomic data using LASSO

High-throughput gene expression technologies such as microarrays have been utilized in a variety of scientific applications. Most of the work has been on assessing univariate associations between gene expression with clinical outcome (variable selection) or on developing classification procedures with gene expression data (supervised learning). We consider a hybrid variable selection/classification approach that is based on linear combinations of the gene expression profiles that maximize an accuracy measure summarized using the receiver operating characteristic curve. Under a specific probability model, this leads to consideration of linear discriminant functions. We incorporate an automated variable selection approach using LASSO. An equivalence between LASSO estimation with support vector machines allows for model fitting using standard software. We apply the proposed method to simulated data as well as data from a recently published prostate cancer study.

Fitting of Mixtures with Unspecified Number of Components Using Cross Validation Distance Estimate

Estimation of the number of mixture components (k) is an unsolved problem. Available methods for estimation of k include bootstrapping the likelihood ratio test statistics and optimizing a variety of validity functionals such as AIC, BIC/MDL, and ICOMP. We investigate the minimization of distance between fitted mixture model and the true density as a method for estimating k. The distances considered are Kullback-Leibler (KL) and “L sub 2”. We estimate these distances using cross validation. A reliable estimate of k is obtained by voting of B estimates of k corresponding to B cross validation estimates of distance. This estimation methods with KL distance is very similar to Monte Carlo cross validated likelihood methods discussed by Smyth (2000). With focus on univariate normal mixtures, we present simulation studies that compare the cross validated distance method with AIC, BIC/MDL, and ICOMP. We also apply the cross validation estimate of distance approach along with AIC, BIC/MDL and ICOMP approach, to data from an osteoporosis drug trial in order to find groups that differentially respond to treatment.