Sviluppo di tecniche di Speckle tracking per la correzione del movimento in ecografia epatica perfusionale

L'ecografia con mezzo di contrasto è una tecnica non invasiva che consente di visualizzare la micro e la macrocircolazione grazie all'utilizzo di microbolle gassose che si distribuiscono in tutto il sistema cardiovascolare. Le informazioni emodinamiche e perfusionali ricavabili dalle immagini eco con contrasto possono essere utilizzate per costruire un modello a grafo dell'albero vascolare epatico. L'analisi della connettività del grafo rappresenta una strategia molto promettente che potrebbe consentire di sostituire la misurazione del gradiente pressorio venoso del fegato, che richiede cateterismo, determinando un notevole miglioramento nella gestione dei pazienti cirrotici. La presente tesi si occupa della correzione dei movimenti del fegato, che deve essere realizzata prima di costruire il grafo per garantire un'accuratezza adeguata. Per correggere i movimenti è proposta una tecnica di Speckle tracking, testata sia in vitro su sequenze eco sintetiche, sia in vivo su sequenze reali fornite dal Policlinico Sant'Orsola.

Development of a health parameters tracking wearable device for detection of stress episodes and principles of panic attacks

A proof of concept for a wearable device is presented to help patients who suffer from panic attacks due to panic disorder. The aim of this device is to enable such patients manage these stressful episodes by guiding them to regulate their breathing and by informing the care taker. Panic attack prediction is deployed that can enable the healthcare providers to not only monitor and manage the panic attacks of a patient but also carry out an early intervention to reduce the symptom severity of the approaching panic attack. The patient can acquire the help they need, ultimately regaining control. The concept of panic attack prediction can lead to a personalized treatment of the patient. The study is conducted using a small real-world dataset, and only two primary symptoms of panic attack are used. These symptoms include pacing heart rate and hyperventilation or abnormal breathing rate. This thesis project is developed in collaboration with ALTEN italia and all the required hardware is provided by them.

The behaviour of osteocyte dendritic processes in bone under cyclic load (il comportamento dei processi dendritici degli osteociti nel tessuto osseo sottoposti a carico ciclico)

Bone is continually being removed and replaced through the actions of basic multicellular units (BMU). This constant upkeep is necessary to remove microdamage formed naturally due to fatigue and thus maintain the integrity of the bone. The repair process in bone is targeted, meaning that a BMU travels directly to the site of damage and repairs it. It is still unclear how targeted remodelling is stimulated and directed but it is highly likely that osteocytes play a role. A number of theories have been advanced to explain the microcrack osteocyte interaction but no complete mechanism has been demonstrated. Osteocytes are connected to each other by dendritic processes. The “scissors model" proposed that the rupture of these processes where they cross microcracks signals the degree of damage and the urgency of the necessary repair. In its original form it was proposed that under applied compressive loading, microcrack faces will be pressed together and undergo relative shear movement. If this movement is greater than the width of an osteocyte process, then the process will be cut in a “scissors like" motion, releasing RANKL, a cytokine known to be essential in the formation of osteoclasts from pre-osteoclasts. The main aim of this thesis was to investigate this theoretical model with a specific focus on microscopy and finite element modelling. Previous studies had proved that cyclic stress was necessary for osteocyte process rupture to occur. This was a divergence from the original “scissors model" which had proposed that the cutting of cell material occurred in one single action. The present thesis is the first study to show fatigue failure in cellular processes spanning naturally occurring cracks and it's the first study to estimate the cyclic strain range and relate it to the number of cycles to failure, for any type of cell. Rupture due to shear movement was ruled out as microcrack closing never occurred, as a result of plastic deformation of the bone. Fatigue failure was found to occur due to cyclic tensile stress in the locality of the damage. The strain range necessary for osteocyte process rupture was quantified. It was found that the lower the process strain range the greater the number of cycles to cell process failure. FEM modelling allowed to predict stress in the vicinity of an osteocyte process and to analyse its interaction with the bone surrounding it: simulations revealed evident creep effects in bone during cyclic loading. This thesis confirms and dismisses aspects of the “scissors model". The observations support the model as a viable mechanism of microcrack detection by the osteocyte network, albeit in a slightly modified form where cyclic loading is necessary and the method of rupture is fatigue failure due to cyclic tensile motion. An in depth study was performed focusing on microscopy analysis of naturally occurring cracks in bone and FEM simulation analysis of an osteocyte process spanning a microcrack in bone under cyclic load.

UWB radar sensor networks: Detection algorithms design and experimental analysis

In the last years radar sensor networks for localization and tracking in indoor environment have generated more and more interest, especially for anti-intrusion security systems. These networks often use Ultra Wide Band (UWB) technology, which consists in sending very short (few nanoseconds) impulse signals. This approach guarantees high resolution and accuracy and also other advantages such as low price, low power consumption and narrow-band interference (jamming) robustness. In this thesis the overall data processing (done in MATLAB environment) is discussed, starting from experimental measures from sensor devices, ending with the 2D visualization of targets movements over time and focusing mainly on detection and localization algorithms. Moreover, two different scenarios and both single and multiple target tracking are analyzed.

Analisi sperimentale sull'utilizzo del Microsoft Kinect One come sistema di Body Tracking per la realta virtuale in riabilitazione

In questo progetto di tesi sarà innanzitutto presentato il Kinect One e sarà fatta una panoramica sull’uso della realtà virtuale in ambito riabilitativo. In seguito sarà analizzato l’algoritmo di Body tracking, valutandone il comportamento in diverse situazioni pratiche e poi stimandone la precisione in statica. Sarà presentato un filtraggio per limitare il rumore in tempo reale e valutarne i pro ed i contro in funzione delle caratteristiche impostabili. Saranno presentate inoltre le metodologie con cui gli algoritmi integrati del Kinect permettono di ricavare una stima dell’orientamento delle parti anatomiche nello spazio ed alcune considerazioni circa le implicazioni pratiche di tali metodologie, anche in base alle osservazioni sul campo ottenute durante i mesi di realizzazione di questo progetto. Lo scopo è determinare se e come sia possibile utilizzare il Microsoft Kinect One come unico sistema di motion tracking del paziente in applicazioni cliniche di riabilitazione, quali limiti ci sono nel suo utilizzo e quali categorie di scenari e prove potrebbe supportare.

A deep learning-based approach for 3D people tracking

Questa tesi si occupa dell’estensione di un framework software finalizzato all'individuazione e al tracciamento di persone in una scena ripresa da telecamera stereoscopica. In primo luogo è rimossa la necessità di una calibrazione manuale offline del sistema sfruttando algoritmi che consentono di individuare, a partire da un fotogramma acquisito dalla camera, il piano su cui i soggetti tracciati si muovono. Inoltre, è introdotto un modulo software basato su deep learning con lo scopo di migliorare la precisione del tracciamento. Questo componente, che è in grado di individuare le teste presenti in un fotogramma, consente ridurre i dati analizzati al solo intorno della posizione effettiva di una persona, escludendo oggetti che l’algoritmo di tracciamento sarebbe portato a individuare come persone.

Sviluppo di un algoritmo di Object Detection per la privacy in ambito industriale

In questa tesi è descritto il lavoro svolto presso un'azienda informatica locale, allo scopo di ricerca ed implementazione di un algoritmo per individuare ed offuscare i volti presenti all'interno di video di e-learning in ambito industriale, al fine di garantire la privacy degli operai presenti. Tale algoritmo sarebbe stato poi da includere in un modulo software da inserire all'interno di un applicazione web già esistente per la gestione di questi video. Si è ricercata una soluzione ad hoc considerando le caratteristiche particolare del problema in questione, studiando le principali tecniche della Computer Vision per comprendere meglio quale strada percorrere. Si è deciso quindi di implementare un algoritmo di Blob Tracking basato sul colore.

Motion Control of an Advanced Electric Implement for Agricultural Autonomous Vehicles

Nowadays, the development of intelligent and autonomous vehicles used to perform agricultural activities is essential to improve quantity and quality of agricultural productions. Moreover, with automation techniques it is possible to reduce the usage of agrochemicals and minimize the pollution. The University of Bologna is developing an innovative system for orchard management called ORTO (Orchard Rapid Transportation System). This system involves an autonomous electric vehicle capable to perform agricultural activities inside an orchard structure. The vehicle is equipped with an implement capable to perform different tasks. The purpose of this thesis project is to control the vehicle and the implement to perform an inter-row grass mowing. This kind of task requires a synchronized motion between the traction motors and the implement motors. A motion control system has been developed to generate trajectories and manage their synchronization. Two main trajectories type have been used: a five order polynomial trajectory and a trapezoidal trajectory. These two kinds of trajectories have been chosen in order to perform a uniform grass mowing, paying a particular attention to the constrains of the system. To synchronize the motions, the electronic cams approach has been adopted. A master profile has been generated and all the trajectories have been linked to the master motion. Moreover, a safety system has been developed. The aim of this system is firstly to improve the safety during the motion, furthermore it allows to manage obstacle detection and avoidance. Using some particular techniques obstacles can be detected and recovery action can be performed to overcome the problem. Once the measured force reaches the predefined force threshold, then the vehicle stops immediately its motion. The whole project has been developed by employing Matlab and Simulink. Eventually, the software has been translated into C code and executed on the TI Lauchpad XL board.

Design and implementation of an optimal motion cueing algorithm for the Stuttgart Driving Simulator

Driving simulators emulate a real vehicle drive in a virtual environment. One of the most challenging problems in this field is to create a simulated drive as real as possible to deceive the driver's senses and cause the believing to be in a real vehicle. This thesis first provides an overview of the Stuttgart driving simulator with a description of the overall system, followed by a theoretical presentation of the commonly used motion cueing algorithms. The second and predominant part of the work presents the implementation of the classical and optimal washout algorithms in a Simulink environment. The project aims to create a new optimal washout algorithm and compare the obtained results with the results of the classical washout. The classical washout algorithm, already implemented in the Stuttgart driving simulator, is the most used in the motion control of the simulator. This classical algorithm is based on a sequence of filters in which each parameter has a clear physical meaning and a unique assignment to a single degree of freedom. However, the effects on human perception are not exploited, and each parameter must be tuned online by an engineer in the control room, depending on the driver's feeling. To overcome this problem and also consider the driver's sensations, the optimal washout motion cueing algorithm was implemented. This optimal control-base algorithm treats motion cueing as a tracking problem, forcing the accelerations perceived in the simulator to track the accelerations that would have been perceived in a real vehicle, by minimizing the perception error within the constraints of the motion platform. The last chapter presents a comparison between the two algorithms, based on the driver's feelings after the test drive. Firstly it was implemented an off-line test with a step signal as an input acceleration to verify the behaviour of the simulator. Secondly, the algorithms were executed in the simulator during a test drive on several tracks.

Development of a state-variable filter for a motion control application.

In collaboration with G.D. SpA I attended an internship with the purpose of developing a filter for the position control of industrial machines during testing and maintenance operations. The filter elaborates a signal in position provided by an electonic handwheel, in order to enable the application to be controlled with a signal in velocity with arbitrarily dynamics chosen during the design phase. Limiting the dynamics of the filter provide a more stable and less demanding reference trajectory which reduce the vibrations and tracking errors of the motor controlled by it. It also prevents misusages of the handwheel from the technician which could end up in harmful interferences between the mechanical parts moved by the handwheel.

Image Processing Algorithms applied to Radio SLAM using MIMO Radars

Radio Simultaneous Location and Mapping (SLAM) consists of the simultaneous tracking of the target and estimation of the surrounding environment, to build a map and estimate the target movements within it. It is an increasingly exploited technique for automotive applications, in order to improve the localization of obstacles and the target relative movement with respect to them, for emergency situations, for example when it is necessary to explore (with a drone or a robot) environments with a limited visibility, or for personal radar applications, thanks to its versatility and cheapness. Until today, these systems were based on light detection and ranging (lidar) or visual cameras, high-accuracy and expensive approaches that are limited to specific environments and weather conditions. Instead, in case of smoke, fog or simply darkness, radar-based systems can operate exactly in the same way. In this thesis activity, the Fourier-Mellin algorithm is analyzed and implemented, to verify the applicability to Radio SLAM, in which the radar frames can be treated as images and the radar motion between consecutive frames can be covered with registration. Furthermore, a simplified version of that algorithm is proposed, in order to solve the problems of the Fourier-Mellin algorithm when working with real radar images and improve the performance. The INRAS RBK2, a MIMO 2x16 mmWave radar, is used for experimental acquisitions, consisting of multiple tests performed in Lab-E of the Cesena Campus, University of Bologna. The different performances of Fourier-Mellin and its simplified version are compared also with the MatchScan algorithm, a classic algorithm for SLAM systems.