Systematic investigation of automated plan generation for breast cancer including beam angle and isocenter selection

Il tumore al seno è il più comune tra le donne nel mondo. La radioterapia è comunemente usata dopo la chirurgia per distruggere eventuali cellule maligne rimaste nel volume del seno. Nei trattamenti di radioterapia bisogna cercare di irradiare il volume da curare limitando contemporaneamente la tossicità nei tessuti sani. In clinica i parametri che definiscono il piano di trattamento radioterapeutico sono selezionati manualmente utilizzando un software di simulazione per trattamenti. Questo processo, detto di trial and error, in cui i differenti parametri vengono modificati e il trattamento viene simulato nuovamente e valutato, può richiedere molte iterazioni rendendolo dispendioso in termini di tempo. Lo studio presentato in questa tesi si concentra sulla generazione automatica di piani di trattamento per irradiare l'intero volume del seno utilizzando due fasci approssimativamente opposti e tangenti al paziente. In particolare ci siamo concentrati sulla selezione delle direzioni dei fasci e la posizione dell'isocentro. A questo scopo, è stato investigata l'efficacia di un approccio combinatorio, nel quale sono stati generati un elevato numero di possibili piani di trattamento utilizzando differenti combinazioni delle direzioni dei due fasci. L'intensità del profilo dei fasci viene ottimizzata automaticamente da un algoritmo, chiamato iCycle, sviluppato nel ospedale Erasmus MC di Rotterdam. Inizialmente tra tutti i possibili piani di trattamento generati solo un sottogruppo viene selezionato, avente buone caratteristiche per quel che riguarda l'irraggiamento del volume del seno malato. Dopo di che i piani che mostrano caratteristiche ottimali per la salvaguardia degli organi a rischio (cuore, polmoni e seno controlaterale) vengono considerati. Questi piani di trattamento sono matematicamente equivalenti quindi per selezionare tra questi il piano migliore è stata utilizzata una somma pesata dove i pesi sono stati regolati per ottenere in media piani che abbiano caratteristiche simili ai piani di trattamento approvati in clinica. Questo metodo in confronto al processo manuale oltre a ridurre considerevol-mente il tempo di generazione di un piano di trattamento garantisce anche i piani selezionati abbiano caratteristiche ottimali nel preservare gli organi a rischio. Inizialmente è stato utilizzato l'isocentro scelto in clinica dal tecnico. Nella parte finale dello studio l'importanza dell'isocentro è stata valutata; ne è risultato che almeno per un sottogruppo di pazienti la posizione dell'isocentro può dare un importante contributo alla qualità del piano di trattamento e quindi potrebbe essere un ulteriore parametro da ottimizzare. 

Design, realization and characterization of automated millifluidic bioreactors for investigating the molecular evolution of lytic or lysogenic vector phages infecting engineered host e. Coli

In questa tesi viene presentato un bioreattore in grado di mantenere nel tempo condizioni biologiche tali che consentano di massimizzare i cicli di evoluzione molecolare di vettori di clonazione fagici: litico (T7) o lisogeno (M13). Verranno quindi introdtti concetti legati alla Teoria della Quasispecie e alla relazione tra errori di autoreplicazione e pressioni selettive naturali o artificiali su popolazioni di virus: il modello naturale del sistema evolutivo. Tuttavia, mantenere delle popolazioni di virus significa formire loro un substrato dove replicare. Per fare ciò, altri gruppi di ricerca hanno giá sviluppato complessi e costosi prototipi di macchinari per la crescita continua di popolazioni batteriche: i compartimenti dei sistemi evolutivi. Il bioreattore, oggetto di questo lavoro, fa parte del progetto europeo Evoprog: general purpose programmable machine evolution on a chip (Jaramillo’s Lab, University of Warwick) che, utilizzando tecnologie fagiche e regolazioni sintetiche esistenti, sará in grado di produrre funzionalità biocomputazionali di due ordini di grandezza più veloci rispetto alle tecniche convenzionali, riducendo allo stesso tempo i costi complessivi. Il primo prototipo consiste in uno o piú fermentatori, dove viene fatta crescere la cultura batterica in condizioni ottimizzate di coltivazione continua, e in un cellstat, un volume separato, dove avviene solo la replicazione dei virus. Entrambi i volumi sono di pochi millilitri e appropriatamente interconnessi per consentire una sorta di screening continuo delle biomolecole prodotte all’uscita. Nella parte finale verranno presentati i risultati degli esperimenti preliminari, a dimostrazione dell’affidabilità del prototipo costruito e dei protocolli seguiti per la sterilizzazione e l’assemblaggio del bioreattore. Gli esperimenti effettuati dimostrano il successo di due coltivazioni virali continue e una ricombinazione in vivo di batteriofagi litici o lisogeni ingegnerizzati. La tesi si conclude valutando i futuri sviluppi e i limiti del sistema, tenendo in considerazione, in particolare, alcune applicazioni rivolte agli studi di una terapia batteriofagica.

Frame-driven Extraction of Linked Data and Ontologies from Text

Ontology design and population -core aspects of semantic technologies- re- cently have become fields of great interest due to the increasing need of domain-specific knowledge bases that can boost the use of Semantic Web. For building such knowledge resources, the state of the art tools for ontology design require a lot of human work. Producing meaningful schemas and populating them with domain-specific data is in fact a very difficult and time-consuming task. Even more if the task consists in modelling knowledge at a web scale. The primary aim of this work is to investigate a novel and flexible method- ology for automatically learning ontology from textual data, lightening the human workload required for conceptualizing domain-specific knowledge and populating an extracted schema with real data, speeding up the whole ontology production process. Here computational linguistics plays a fundamental role, from automati- cally identifying facts from natural language and extracting frame of relations among recognized entities, to producing linked data with which extending existing knowledge bases or creating new ones. In the state of the art, automatic ontology learning systems are mainly based on plain-pipelined linguistics classifiers performing tasks such as Named Entity recognition, Entity resolution, Taxonomy and Relation extraction [11]. These approaches present some weaknesses, specially in capturing struc- tures through which the meaning of complex concepts is expressed [24]. Humans, in fact, tend to organize knowledge in well-defined patterns, which include participant entities and meaningful relations linking entities with each other. In literature, these structures have been called Semantic Frames by Fill- 6 Introduction more [20], or more recently as Knowledge Patterns [23]. Some NLP studies has recently shown the possibility of performing more accurate deep parsing with the ability of logically understanding the structure of discourse [7]. In this work, some of these technologies have been investigated and em- ployed to produce accurate ontology schemas. The long-term goal is to collect large amounts of semantically structured information from the web of crowds, through an automated process, in order to identify and investigate the cognitive patterns used by human to organize their knowledge.

Optimisation of chemical lysis protocol for point-of-care leukocyte differentiation

The full blood cell (FBC) count is the most common indicator of diseases. At present hematology analyzers are used for the blood cell characterization, but, recently, there has been interest in using techniques that take advantage of microscale devices and intrinsic properties of cells for increased automation and decreased cost. Microfluidic technologies offer solutions to handling and processing small volumes of blood (2-50 uL taken by finger prick) for point-of-care(PoC) applications. Several PoC blood analyzers are in use and may have applications in the fields of telemedicine, out patient monitoring and medical care in resource limited settings. They have the advantage to be easy to move and much cheaper than traditional analyzers, which require bulky instruments and consume large amount of reagents. The development of miniaturized point-of-care diagnostic tests may be enabled by chip-based technologies for cell separation and sorting. Many current diagnostic tests depend on fractionated blood components: plasma, red blood cells (RBCs), white blood cells (WBCs), and platelets. Specifically, white blood cell differentiation and counting provide valuable information for diagnostic purposes. For example, a low number of WBCs, called leukopenia, may be an indicator of bone marrow deficiency or failure, collagen- vascular diseases, disease of the liver or spleen. The leukocytosis, a high number of WBCs, may be due to anemia, infectious diseases, leukemia or tissue damage. In the laboratory of hybrid biodevices, at the University of Southampton,it was developed a functioning micro impedance cytometer technology for WBC differentiation and counting. It is capable to classify cells and particles on the base of their dielectric properties, in addition to their size, without the need of labeling, in a flow format similar to that of a traditional flow cytometer. It was demonstrated that the micro impedance cytometer system can detect and differentiate monocytes, neutrophils and lymphocytes, which are the three major human leukocyte populations. The simplicity and portability of the microfluidic impedance chip offer a range of potential applications in cell analysis including point-of-care diagnostic systems. The microfluidic device has been integrated into a sample preparation cartridge that semi-automatically performs erythrocyte lysis before leukocyte analysis. Generally erythrocytes are manually lysed according to a specific chemical lysis protocol, but this process has been automated in the cartridge. In this research work the chemical lysis protocol, defined in the patent US 5155044 A, was optimized in order to improve white blood cell differentiation and count performed by the integrated cartridge.

Seismic vulnerability of the ancient albergotti's manor in Arezzo, Italy

The thesis moves from the need of understanding how a historical building would behave in case of earthquake and this purpose is strongly linked to the fact that the majority of Italian structures are old ones placed in seismic sites. Primarily an architectural and chronological research is provided in order to figure out how the building has developed in time; then, after the reconstruction of the skeleton of the analyzed element (“Villa i Bossi” in Gragnone, AR), a virtual model is created such that the main walls and sections are tested according to the magnitude of expected seismic events within the reference area. This approach is basically aimed at verifying the structure’s reliability as composed by single units; the latter are treated individually in order to find out all the main critical points where rehabilitation might be needed. Finally the most harmful sections are studied in detail and proper strengthening is advised according to the current know-how.

Synthesis of novel cobalt and iron complexes for sustainable catalytic hydrogenations

Aim of the present work of thesis is to synthesize new non-noble metal based complexes to be employ in redox reactions by a metal-ligand cooperative mechanism. The need of replacing toxic and expensive precious metal complexes with more available and benign metals, has led to the development of new compounds based on cobalt and iron, which are the metals investigated in this study. A carbonyl-tetrahydroborato-bis[(2-diisopropylphosphino)ethyl]amine-cobalt complex bearing a PNP-type ligand is synthesized by a three-step route. Optimization attempt of reaction route were assessed in order to lowering reaction times and solvent waste. New cobalt complex has been tested in esters hydrogenation as well as in acceptorless dehydrogenative coupling of ethanol. Other varieties of substrates were also tested in order to evaluate any possible applications. Concerning iron complex, dicarbonyl-(η4-3,4-bis(4-methoxyphenyl)-2,5-diphenylcyclopenta-2,4-dienone)(1,3-dimethyl-ilidene)iron is synthesized by a three steps route, involving transmetallation of a silver complex, derived from an imidazolium salt, to iron complex. In order to avoid solvent waste, optimization is assessed. Studies were performed to assess activity of triscarbonyl iron precursor toward imidazolium salt and silver complexes.

Lewis acid catalysed direct glycosylation of N-acetyl-D-glucosamine

Incorporation of the relevant monosaccharide N-Acetyl-D-glucosamine (GlcNAc) into synthetic oligosaccharides by chemical glycosylation is still a very challenging object of studies, since direct reactions are low yielding. This issue is generally ascribed to its low solubility in common solvents and to the formation of a poorly reactive oxazoline intermediate, which is typically bypassed by introducing extra synthetic steps to avoid the presence of the NHAc moiety during glycosylation. Recently, a new direct Lewis acids-catalysed GlcNAc-ylation protocol has been disclosed, with acylated donors appearing to hold potential for high yielding glycosylation reactions. This master project focused indeed on a novel synthesis of promising 1-acyl GlcNAc donors, in order to test them in direct Lewis acid catalysed glycosylation without the need of N-protecting groups. Screening of various Lewis acids and reaction conditions with these acylated donors has been carried out, in presence of reactive primary alcohols as well as more challenging carbohydrate acceptor alcohols. These experiments demonstrated that the fine tuning of the leaving group combined with a suitable metal triflate could lead to a successful reaction outcome in the direct glycosylation. Successful methodology of this kind would provide rapid access to naturally occurring N-glycan motifs, such as the highly relevant human milk oligosaccharides (HMOs).

Fracture toughness characterization of aluminium-epoxy adhesive joints reinforced with electrospun Nylon 6,6 nanofibers

The use of adhesives to join two different substrates is an efficient replacement to classic joining technologies such as welding and soldering. One the one hand adhesion has different advantages over those techniques such as an improved stress distribution and the potential weight reduction of the structure; on the other hand, two of the most important drawbacks are a relatively low fracture toughness and the need of an accurate surface preparation. These two aspects will be accurately analysed in the present work: the use of Nylon nanofibers as reinforcement for the adhesive should increase fracture toughness, while a surface preparation method consisting of mechanical and chemical treatments will be developed. After the specimens are produced, they will be tested in mode I fracture using a DCB (Double Beam Cantilever) test, which allows to measure the fracture toughness during crack propagation. At the end of the test, the surfaces of the adherends will be visually observed and SEM (Scanning Electronic Microscope) analysed in order to evaluate if adhesive or cohesive fracture occurred, and thus if surface treatments has been well developed to allow a better adhesive-aluminium joining.

A Distributed Support based on Kubeflow to Ease MLOps of Federated Learning Algorithms

The main objective of my thesis work is to exploit the Google native and open-source platform Kubeflow, specifically using Kubeflow pipelines, to execute a Federated Learning scalable ML process in a 5G-like and simplified test architecture hosting a Kubernetes cluster and apply the largely adopted FedAVG algorithm and FedProx its optimization empowered by the ML platform ‘s abilities to ease the development and production cycle of this specific FL process. FL algorithms are more are and more promising and adopted both in Cloud application development and 5G communication enhancement through data coming from the monitoring of the underlying telco infrastructure and execution of training and data aggregation at edge nodes to optimize the global model of the algorithm ( that could be used for example for resource provisioning to reach an agreed QoS for the underlying network slice) and after a study and a research over the available papers and scientific articles related to FL with the help of the CTTC that suggests me to study and use Kubeflow to bear the algorithm we found out that this approach for the whole FL cycle deployment was not documented and may be interesting to investigate more in depth. This study may lead to prove the efficiency of the Kubeflow platform itself for this need of development of new FL algorithms that will support new Applications and especially test the FedAVG algorithm performances in a simulated client to cloud communication using a MNIST dataset for FL as benchmark.

Application of active and transfer learning to wound image segmentation

Wound management is a fundamental task in standard clinical practice. Automated solutions already exist for humans, but there is a lack of applications on wound management for pets. The importance of a precise and efficient wound assessment is helpful to improve diagnosis and to increase the effectiveness of treatment plans for the chronic wounds. The goal of the research was to propose an automated pipeline capable of segmenting natural light-reflected wound images of animals. Two datasets composed by light-reflected images were used in this work: Deepskin dataset, 1564 human wound images obtained during routine dermatological exams, with 145 manual annotated images; Petwound dataset, a set of 290 wound photos of dogs and cats with 0 annotated images. Two implementations of U-Net Convolutioal Neural Network model were proposed for the automated segmentation. Active Semi-Supervised Learning techniques were applied for human-wound images to perform segmentation from 10% of annotated images. Then the same models were trained, via Transfer Learning, adopting an Active Semi- upervised Learning to unlabelled animal-wound images. The combination of the two training strategies proved their effectiveness in generating large amounts of annotated samples (94% of Deepskin, 80% of PetWound) with the minimal human intervention. The correctness of automated segmentation were evaluated by clinical experts at each round of training thus we can assert that the results obtained in this thesis stands as a reliable solution to perform a correct wound image segmentation. The use of Transfer Learning and Active Semi-Supervied Learning allows to minimize labelling effort from clinicians, even requiring no starting manual annotation at all. Moreover the performances of the model with limited number of parameters suggest the implementation of smartphone-based application to this topic, helping the future standardization of light-reflected images as acknowledge medical images.