An experimental and theoretical approach to correct for the scattered radiation in an X-ray computer tomography system for industrial applications

The main problem connected to cone beam computed tomography (CT) systems for industrial applications employing 450 kV X-ray tubes is the high amount of scattered radiation which is added to the primary radiation (signal). This stray radiation leads to a significant degradation of the image quality. A better understanding of the scattering and methods to reduce its effects are therefore necessary to improve the image quality. Several studies have been carried out in the medical field at lower energies, whereas studies in industrial CT, especially for energies up to 450 kV, are lacking. Moreover, the studies reported in literature do not consider the scattered radiation generated by the CT system structure and the walls of the X-ray room (environmental scatter). In order to investigate the scattering on CT projections a GEANT4-based Monte Carlo (MC) model was developed. The model, which has been validated against experimental data, has enabled the calculation of the scattering including the environmental scatter, the optimization of an anti-scatter grid suitable for the CT system, and the optimization of the hardware components of the CT system. The investigation of multiple scattering in the CT projections showed that its contribution is 2.3 times the one of primary radiation for certain objects. The results of the environmental scatter showed that it is the major component of the scattering for aluminum box objects of front size 70 x 70 mm2 and that it strongly depends on the thickness of the object and therefore on the projection. For that reason, its correction is one of the key factors for achieving high quality images. The anti-scatter grid optimized by means of the developed MC model was found to reduce the scatter-toprimary ratio in the reconstructed images by 20 %. The object and environmental scatter calculated by means of the simulation were used to improve the scatter correction algorithm which could be patented by Empa. The results showed that the cupping effect in the corrected image is strongly reduced. The developed CT simulation is a powerful tool to optimize the design of the CT system and to evaluate the contribution of the scattered radiation to the image. Besides, it has offered a basis for a new scatter correction approach by which it has been possible to achieve images with the same spatial resolution as state-of-the-art well collimated fan-beam CT with a gain in the reconstruction time of a factor 10. This result has a high economic impact in non-destructive testing and evaluation, and reverse engineering.

Adaptive multiscale biological signal processing

Biological processes are very complex mechanisms, most of them being accompanied by or manifested as signals that reflect their essential characteristics and qualities. The development of diagnostic techniques based on signal and image acquisition from the human body is commonly retained as one of the propelling factors in the advancements in medicine and biosciences recorded in the recent past. It is a fact that the instruments used for biological signal and image recording, like any other acquisition system, are affected by non-idealities which, by different degrees, negatively impact on the accuracy of the recording. This work discusses how it is possible to attenuate, and ideally to remove, these effects, with a particular attention toward ultrasound imaging and extracellular recordings. Original algorithms developed during the Ph.D. research activity will be examined and compared to ones in literature tackling the same problems; results will be drawn on the base of comparative tests on both synthetic and in-vivo acquisitions, evaluating standard metrics in the respective field of application. All the developed algorithms share an adaptive approach to signal analysis, meaning that their behavior is not dependent only on designer choices, but driven by input signal characteristics too. Performance comparisons following the state of the art concerning image quality assessment, contrast gain estimation and resolution gain quantification as well as visual inspection highlighted very good results featured by the proposed ultrasound image deconvolution and restoring algorithms: axial resolution up to 5 times better than algorithms in literature are possible. Concerning extracellular recordings, the results of the proposed denoising technique compared to other signal processing algorithms pointed out an improvement of the state of the art of almost 4 dB.

Performance evaluation of detectors for digital radiography

To date the hospital radiological workflow is completing a transition from analog to digital technology. Since the X-rays digital detection technologies have become mature, hospitals are trading on the natural devices turnover to replace the conventional screen film devices with digital ones. The transition process is complex and involves not just the equipment replacement but also new arrangements for image transmission, display (and reporting) and storage. This work is focused on 2D digital detector’s characterization with a concern to specific clinical application; the systems features linked to the image quality are analyzed to assess the clinical performances, the conversion efficiency, and the minimum dose necessary to get an acceptable image. The first section overviews the digital detector technologies focusing on the recent and promising technological developments. The second section contains a description of the characterization methods considered in this thesis categorized in physical, psychophysical and clinical; theory, models and procedures are described as well. The third section contains a set of characterizations performed on new equipments that appears to be some of the most advanced technologies available to date. The fourth section deals with some procedures and schemes employed for quality assurance programs.

Enhancing the efficiency of solar concentrators by controlled optical aberrations

The diameters of traditional dish concentrators can reach several tens of meters, the construction of monolithic mirrors being difficult at these scales: cheap flat reflecting facets mounted on a common frame generally reproduce a paraboloidal surface. When a standard imaging mirror is coupled with a PV dense array, problems arise since the solar image focused is intrinsically circular. Moreover, the corresponding irradiance distribution is bell-shaped in contrast with the requirement of having all the cells under the same illumination. Mismatch losses occur when interconnected cells experience different conditions, in particular in series connections. In this PhD Thesis, we aim at solving these issues by a multidisciplinary approach, exploiting optical concepts and applications developed specifically for astronomical use, where the improvement of the image quality is a very important issue. The strategy we propose is to boost the spot uniformity acting uniquely on the primary reflector and avoiding the big mirrors segmentation into numerous smaller elements that need to be accurately mounted and aligned. In the proposed method, the shape of the mirrors is analytically described by the Zernike polynomials and its optimization is numerically obtained to give a non-imaging optics able to produce a quasi-square spot, spatially uniform and with prescribed concentration level. The freeform primary optics leads to a substantial gain in efficiency without secondary optics. Simple electrical schemes for the receiver are also required. The concept has been investigated theoretically modeling an example of CPV dense array application, including the development of non-optical aspects as the design of the detector and of the supporting mechanics. For the method proposed and the specific CPV system described, a patent application has been filed in Italy with the number TO2014A000016. The patent has been developed thanks to the collaboration between the University of Bologna and INAF (National Institute for Astrophysics).

CO2-EVAR: An innovative approach to Automated Carbon Dioxide Angiography during Endovascular Abdominal Aortic Aneurysm repair

Objectives CO2-EVAR was proposed for treatment of AAA especially in patients with CKD. Issues regarding standardization, such as visualization of lowest renal artery (LoRA) and quality image in angiographies performed from pigtail or introducer-sheath, are still unsolved. Aim of the study was to analyze different steps of CO2-EVAR to create an operative protocol to standardize the procedure. Methods Patients undergoing CO2-EVAR were prospectively enrolled in 5 European centers (2018-2021). CO2-EVAR was performed using an automated injector. LoRA visualization and image quality (1-4) were analyzed and compared at different procedure steps: preoperative CO2-angiography from Pigtail/Introducer-sheath (1st Step), angiographies from Pigtail at 0%,50%,100% main body (MB) deployment (2nd Step), contralateral hypogastric artery (CHA) visualization with CO2 injection from femoral Introducer-sheath (3rd Step) and completion angiogram from Pigtail/Introducer-sheath (4th Step). Intra-/postoperative adverse events were evaluated. Results Sixty-five patients undergoing CO2-EVAR were enrolled, 55/65(84.5%) male, median age 75(11.5) years. Median ICM was 20(54)cc; 19/65(29.2%) procedures were performed with 0-iodine. 1st Step: median image quality was significantly higher with CO2 injected from femoral introducer [Pigtail2(3)vs.3(3)Introducer,p=.008]. 2nd Step: LoRA was more frequently detected at 50% (93%vs.73.2%, p=.002) and 100% (94.1%vs.78.4%, p=.01) of MB deployment compared with first angiography from Pigtail; image quality was significantly higher at 50% [3(3)vs.2(3),p=<.001] and 100% [4(3) vs.2(3),p=.001] of MB deployment. CHA was detected in 93% cases (3rd Step). Mean image quality was significantly higher when final angiogram (4th Step) was performed from introducer (Pigtail2.6±1.1vs.3.1±0.9Introducer,p=<.001). Rates of intra-/postoperative adverse events (pain,vomit,diarrhea) were 7.7% and 12.5%. Conclusions Preimplant CO2-angiography should be performed from Introducer-sheath. MB steric bulk during its deployment should be used to improve image quality and LoRA visualization with CO2. CHA can be satisfactorily visualized with CO2. Completion CO2-angiogram should be performed from femoral Introducer-sheath. This operative protocol allows to perform CO2-EVAR with minimal ICM and low rate of mild complications.

Solutions to common issues in widefield microscopy: vignetting, mosaicing and depth of focus

In this thesis we have developed solutions to common issues regarding widefield microscopes, facing the problem of the intensity inhomogeneity of an image and dealing with two strong limitations: the impossibility of acquiring either high detailed images representative of whole samples or deep 3D objects. First, we cope with the problem of the non-uniform distribution of the light signal inside a single image, named vignetting. In particular we proposed, for both light and fluorescent microscopy, non-parametric multi-image based methods, where the vignetting function is estimated directly from the sample without requiring any prior information. After getting flat-field corrected images, we studied how to fix the problem related to the limitation of the field of view of the camera, so to be able to acquire large areas at high magnification. To this purpose, we developed mosaicing techniques capable to work on-line. Starting from a set of overlapping images manually acquired, we validated a fast registration approach to accurately stitch together the images. Finally, we worked to virtually extend the field of view of the camera in the third dimension, with the purpose of reconstructing a single image completely in focus, stemming from objects having a relevant depth or being displaced in different focus planes. After studying the existing approaches for extending the depth of focus of the microscope, we proposed a general method that does not require any prior information. In order to compare the outcome of existing methods, different standard metrics are commonly used in literature. However, no metric is available to compare different methods in real cases. First, we validated a metric able to rank the methods as the Universal Quality Index does, but without needing any reference ground truth. Second, we proved that the approach we developed performs better in both synthetic and real cases.

Fish quality assessment through the application of chemico-physical, sensory and microbiological analyses

The quality of fish products is indispensably linked to the freshness of the raw material modulated by appropriate manipulation and storage conditions, specially the storage temperature after catch. The purpose of the research presented in this thesis, which was largely conducted in the context of a research project funded by Italian Ministry of Agricultural, Food and Forestry Policies (MIPAAF), concerned the evaluation of the freshness of farmed and wild fish species, in relation to different storage conditions, under ice (0°C) or at refrigeration temperature (4°C). Several specimens of different species, bogue (Boops boops), red mullet (Mullus barbatus), sea bream (Sparus aurata) and sea bass (Dicentrarchus labrax), during storage, under the different temperature conditions adopted, have been examined. The assessed control parameters were physical (texture, through the use of a dynamometer; visual quality using a computer vision system (CVS)), chemical (through footprint metabolomics 1H-NMR) and sensory (Quality Index Method (QIM). Microbiological determinations were also carried out on the species of hake (Merluccius merluccius). In general obtained results confirmed that the temperature of manipulation/conservation is a key factor in maintaining fish freshness. NMR spectroscopy showed to be able to quantify and evaluate the kinetics for unselected compounds during fish degradation, even a posteriori. This can be suitable for the development of new parameters related to quality and freshness. The development of physical methods, particularly the image analysis performed by computer vision system (CVS), for the evaluation of fish degradation, is very promising. Among CVS parameters, skin colour, presence and distribution of gill mucus, and eye shape modification evidenced a high sensibility for the estimation of fish quality loss, as a function of the adopted storage conditions. Particularly the eye concavity index detected on fish eye showed a high positive correlation with total QIM score.

Quality assessment of a landslide inventory map and its application to land‐use planning. A case study in the Northern Apennines (Emilia‐Romagna region, Italy)

Landslide hazard and risk are growing as a consequence of climate change and demographic pressure. Land‐use planning represents a powerful tool to manage this socio‐economic problem and build sustainable and landslide resilient communities. Landslide inventory maps are a cornerstone of land‐use planning and, consequently, their quality assessment represents a burning issue. This work aimed to define the quality parameters of a landslide inventory and assess its spatial and temporal accuracy with regard to its possible applications to land‐use planning. In this sense, I proceeded according to a two‐steps approach. An overall assessment of the accuracy of data geographic positioning was performed on four case study sites located in the Italian Northern Apennines. The quantification of the overall spatial and temporal accuracy, instead, focused on the Dorgola Valley (Province of Reggio Emilia). The assessment of spatial accuracy involved a comparison between remotely sensed and field survey data, as well as an innovative fuzzylike analysis of a multi‐temporal landslide inventory map. Conversely, long‐ and short‐term landslide temporal persistence was appraised over a period of 60 years with the aid of 18 remotely sensed image sets. These results were eventually compared with the current Territorial Plan for Provincial Coordination (PTCP) of the Province of Reggio Emilia. The outcome of this work suggested that geomorphologically detected and mapped landslides are a significant approximation of a more complex reality. In order to convey to the end‐users this intrinsic uncertainty, a new form of cartographic representation is needed. In this sense, a fuzzy raster landslide map may be an option. With regard to land‐use planning, landslide inventory maps, if appropriately updated, confirmed to be essential decision‐support tools. This research, however, proved that their spatial and temporal uncertainty discourages any direct use as zoning maps, especially when zoning itself is associated to statutory or advisory regulations.

Variational and learning models for image and time series inverse problems

Inverse problems are at the core of many challenging applications. Variational and learning models provide estimated solutions of inverse problems as the outcome of specific reconstruction maps. In the variational approach, the result of the reconstruction map is the solution of a regularized minimization problem encoding information on the acquisition process and prior knowledge on the solution. In the learning approach, the reconstruction map is a parametric function whose parameters are identified by solving a minimization problem depending on a large set of data. In this thesis, we go beyond this apparent dichotomy between variational and learning models and we show they can be harmoniously merged in unified hybrid frameworks preserving their main advantages. We develop several highly efficient methods based on both these model-driven and data-driven strategies, for which we provide a detailed convergence analysis. The arising algorithms are applied to solve inverse problems involving images and time series. For each task, we show the proposed schemes improve the performances of many other existing methods in terms of both computational burden and quality of the solution. In the first part, we focus on gradient-based regularized variational models which are shown to be effective for segmentation purposes and thermal and medical image enhancement. We consider gradient sparsity-promoting regularized models for which we develop different strategies to estimate the regularization strength. Furthermore, we introduce a novel gradient-based Plug-and-Play convergent scheme considering a deep learning based denoiser trained on the gradient domain. In the second part, we address the tasks of natural image deblurring, image and video super resolution microscopy and positioning time series prediction, through deep learning based methods. We boost the performances of supervised, such as trained convolutional and recurrent networks, and unsupervised deep learning strategies, such as Deep Image Prior, by penalizing the losses with handcrafted regularization terms.

Simulation of optical observables for space missions to small bodies

The navigation of deep space spacecraft requires accurate measurement of the probe’s state and attitude with respect to a body whose ephemerides may not be known with good accuracy. The heliocentric state of the spacecraft is estimated through radiometric techniques (ranging, Doppler, and Delta-DOR), while optical observables can be introduced to improve the uncertainty in the relative position and attitude with respect to the target body. In this study, we analyze how simulated optical observables affect the estimation of parameters in an orbit determination problem, considering the case of the ESA’s Hera mission towards the binary asteroid system composed of Didymos and Dimorphos. To this extent, a shape model and a photometric function are used to create synthetic onboard camera images. Then, using a stereophotoclinometry technique on some of the simulated images, we create a database of maplets that describe the 3D geometry of the surface around a set of landmarks. The matching of maplets with the simulated images provides the optical observables, expressed as pixel coordinates in the camera frame, which are fed to an orbit determination filter to estimate a certain number of solve-for parameters. The noise introduced in the output optical observables by the image processing can be quantified using as a metric the quality of the residuals, which is used to fine-tune the maplet-matching parameters. In particular, the best results are obtained when using small maplets, with high correlation coefficients and occupation factors.

Content-aware approach for improving biomedical image analysis: an interdisciplinary study series

Biomedicine is a highly interdisciplinary research area at the interface of sciences, anatomy, physiology, and medicine. In the last decade, biomedical studies have been greatly enhanced by the introduction of new technologies and techniques for automated quantitative imaging, thus considerably advancing the possibility to investigate biological phenomena through image analysis. However, the effectiveness of this interdisciplinary approach is bounded by the limited knowledge that a biologist and a computer scientist, by professional training, have of each other’s fields. The possible solution to make up for both these lacks lies in training biologists to make them interdisciplinary researchers able to develop dedicated image processing and analysis tools by exploiting a content-aware approach. The aim of this Thesis is to show the effectiveness of a content-aware approach to automated quantitative imaging, by its application to different biomedical studies, with the secondary desirable purpose of motivating researchers to invest in interdisciplinarity. Such content-aware approach has been applied firstly to the phenomization of tumour cell response to stress by confocal fluorescent imaging, and secondly, to the texture analysis of trabecular bone microarchitecture in micro-CT scans. Third, this approach served the characterization of new 3-D multicellular spheroids of human stem cells, and the investigation of the role of the Nogo-A protein in tooth innervation. Finally, the content-aware approach also prompted to the development of two novel methods for local image analysis and colocalization quantification. In conclusion, the content-aware approach has proved its benefit through building new approaches that have improved the quality of image analysis, strengthening the statistical significance to allow unveiling biological phenomena. Hopefully, this Thesis will contribute to inspire researchers to striving hard for pursuing interdisciplinarity.

Regularization meets GreenAI: a new framework for image reconstruction in life sciences applications

Ill-conditioned inverse problems frequently arise in life sciences, particularly in the context of image deblurring and medical image reconstruction. These problems have been addressed through iterative variational algorithms, which regularize the reconstruction by adding prior knowledge about the problem's solution. Despite the theoretical reliability of these methods, their practical utility is constrained by the time required to converge. Recently, the advent of neural networks allowed the development of reconstruction algorithms that can compute highly accurate solutions with minimal time demands. Regrettably, it is well-known that neural networks are sensitive to unexpected noise, and the quality of their reconstructions quickly deteriorates when the input is slightly perturbed. Modern efforts to address this challenge have led to the creation of massive neural network architectures, but this approach is unsustainable from both ecological and economic standpoints. The recently introduced GreenAI paradigm argues that developing sustainable neural network models is essential for practical applications. In this thesis, we aim to bridge the gap between theory and practice by introducing a novel framework that combines the reliability of model-based iterative algorithms with the speed and accuracy of end-to-end neural networks. Additionally, we demonstrate that our framework yields results comparable to state-of-the-art methods while using relatively small, sustainable models. In the first part of this thesis, we discuss the proposed framework from a theoretical perspective. We provide an extension of classical regularization theory, applicable in scenarios where neural networks are employed to solve inverse problems, and we show there exists a trade-off between accuracy and stability. Furthermore, we demonstrate the effectiveness of our methods in common life science-related scenarios. In the second part of the thesis, we initiate an exploration extending the proposed method into the probabilistic domain. We analyze some properties of deep generative models, revealing their potential applicability in addressing ill-posed inverse problems.