Search

The problem of identifying and therefore modelling a complex system makes use of various techniques and strategies whose computational efforts change drastically. It is not straightforward to analyse the complexity of a system as a whole because of myriads of factors, such as the way of arranging its constituent items and how they interact mutually. Intuitively, the bigger the set of sub-parts is, the more numerous the degrees of freedom are. Additionally there is not a specific and global criterion for optimally determining an always-working method that makes the identification procedure easier, especially in those contexts where the number of unknown variables can make the difference. In this sense, plasma physics is not an exception, being a field where complex phenomena, such as plasma instabilities, easily arise. From a systemic, high-level perspective, the possibility of employing a model that can describe these behaviours is particularly appealing, since it can be exploited for control applications that have not to neglect the underlying physical nature. So far, most of the work published in literature has focused on more physically-grounded models, which could describe how plasma physics works in detail, but very little has been done as mentioned before, with the aim of providing a computational, yet system-oriented, insight of these physical systems. Starting from real flux measurements recorded thanks to suitable sensors installed inside Tokamak machines, the paper attempts to provide a solution based on already known tools available in literature to solve the aforementioned problem, by combining both machine learning-based strategies for dimensionality reduction and control theory. More in detail, the whole architecture presented in this work is founded on the use of auto-encoders, which are intrinsically capable of compressing input features thanks to their structure, and Hammerstein–Wiener models, which are structurally endowed with both linear and non-linear sub-modellers for better capturing the whole dynamics to identify. By merging these functional blocks, it is possible to address both the issue of establishing the most relevant sub-set of variables for identification and the identification problem itself, resulting in a fully customisable approach to data-driven modelling.

In this contribution, a new aspect in robotic basic applications is approached. The problem regards the human-machine modeling for remote ultrasound scan equipments. Even if robotic systems for ultrasound scan applications with remote operations have been widely studied, in this research the remote force feedback control has been investigated. This means that the human operator does receive as input the correct force perception that is transmitted by the remote ultrasound scan equipment in analyzing the body of the patient. Two principal aspects have been investigated. The first one is to consider an artificial body model that will receive the control signals from the remote equipment. The second one is to study the suitable feedback control law that try to compensate both the uncertainty between the artificial body and the patient body also taking into account the transmission delay. Therefore the task is to give to the operator the real perception also taking into account of the force effect, thus improving the experience of the remote operation in performing ultrasound scan procedures.

Plasma physics is an example of research field where many measurements carried out at very specific working conditions need to be collected and processed. By looking at the properties of these data, it can be possible to explore their hidden features in order to solve challenging problems that usually require high computational efforts, such as the tomographic reconstruction. In this paper, preliminary but nontrivial analyses of flux measurements produced in a Tokamak machine are shown and discussed, with the aim of introducing an application of some algorithms for feature selection to detect hidden, relevant relationships within given sets of channels. All the statistical details, and therefore the feature selection procedure itself, are introduced in view of further deepenings, such as the aforementioned problem of tomographically reconstructing plasma profiles from flux measurements or modelling the system in terms of its input-output relationship.

This paper aims to present a framework for supervised binary classification of n -Boolean functions through Echo State Networks endowed with Laplacian Eigenmaps for dimensionality reduction. The proposed method is applied both to improve the classification performance when the learnt weights are quantised in view of a digital implementation and as a computational demonstration of the neural reuse theory when parallel outputs are allowed. Our analysis focuses on the effect of various forms of noise (i.e., normal noise, uniform noise and quantisation noise) when all the possible Boolean functions of n input bits are learnt. External disturbances are applied both over the learnt weights and the input features so that we can analyse how resilient the whole architecture is when various forms of parametric noise is injected into the system. Results presented here show that dimensionality reduction allowed by the Laplacian Eigenmaps-based approach improves robustness to these different sources of noise, leading to reduced memory storage requirements while maintaining high classification performance. Our results are compared to those derived from other more common classification techniques in terms of learning performance and computational complexity, also considering a realistic dataset describing a decision making task in a wall-following navigation session with mobile robots.

This manuscript aims at showing the effects of feature selection and manifold reduction methods in dealing with the wall-following problem in mobile robotics, a well-known nonlinearly separable classification problem in which sensor recordings are associated to controlled motor responses. The capabilities of state manifold reduction in Echo State Networks (ESNs) through Laplacian Eigenmaps (LEs) are described in terms of noise rejection over the trained weights. Furthermore, various machine learning-based and data mining-based methodologies are applied to show the advantages of using the most informative contents drawn from the original sensor readings.

In this paper, an insect brain-inspired computational structure was developed. The peculiarity of the core processing layer is the local connectivity among the spiking neurons, which allows for a representation under the cellular nonlinear network paradigm. Moreover, the processing layer works as a liquid state network with fixed internal connections and trainable output weights. Learning was accomplished by adopting a simple supervised, batch approach based on the calculation of the Moore–Penrose matrix. The architecture, taking inspiration from a specific neuropile of the insect brain, the mushroom bodies, is evaluated and compared with other standard and bio-inspired solutions present in the literature, referring to three different scenarios.

The manuscript introduces a data-driven technique founded on Laplacian Eigenmaps for manifold reduction in bio-inspired Liquid State classifiers. Starting from a preliminary introduction about the algorithm and the need of using manifold reduction methods for data representation, a statistical analysis of hyperparameters involved in the Laplacian Eigenmaps technique is presented and the effects of quantisation on trained weights is discussed with a view to efficiently implement multiple parallel mappings in the digital domain.

In this paper an insect brain-inspired neural processing architecture was developed to be applied on board of a bio-robot for solving feature-to-action tasks. The system, accounting on visual features, is able to solve a classification problems using a spatial temporal approach that is typical of bio-inspired neural architectures. The proposed neural structure, taking inspiration from a specific neuropile of the insect brain, called mushroom bodies, is applied to solve tasks shown in insect experiments where non-elemental learning strategies are taken into account. An important peculiarity of the hidden processing layer of the proposed multi-layer architecture is the local, CNN-like connectivity among the spiking neurons, opening the way for an hardware implementation on neuromorphic chips.

Serrated lesions are recognized as important contributors to colorectal cancer incidence. We aimed to prospectively assess the prevalence of serrated lesions and identify potential predictors of these lesions during colonoscopy in an Italian population.Prospective cross sectional study involving 8 endoscopy units from February 1st to July 31st 2012.Out of 2468 colonoscopies, 886 precancerous lesions were detected in 567 patients. Of these, 173 SELs were diagnosed in 148 patients (140 serrated/hyperplastic polyps and 33 serrated adenomas). Prevalence was 7% (173/2468). Serrated lesions accounted for 19.5% of all precancerous lesions. Serrated polyps were prevalent in the left colon (42.1%) and serrated adenomas in the proximal colon (54.5%). Independent clinical predictors of serrated lesions were patient age (OR 0.98 [0.97-1.00]) and post-polypectomy surveillance (OR 1.87 [1.24-2.82]). Endoscopic predictors were right colon location (OR 2.65 [1.63-4.30] vs. rectum; and 1.53 [1.03-2.26] vs. left colon), polypoid shape (OR 0.41 [027-0.64]) and size6 mm (OR 0.49 [0.33-0.72] vs. 6-10 mm; and 0.14 [0.07-0.28] vs.10 mm). There was no independent predictor of serrated adenoma.In our Italian study population, the prevalence of colorectal serrated lesions was 7%. Their diagnosis is associated with younger age and surveillance colonoscopy, right-sided colorectal location, non-polypoid shape and size6 mm.

Strategic innovations in the fight against cancer have resulted in improved diagnosis of multiple primary malignancies (MPM). In fact, early detection has translated into an increased reported incidence of MPM, due to the spread of screening programs and to increasingly sophisticated instrumental diagnostic surveys. The high sensitivity and specificity of modern diagnostic technologies are such that second primary malignant neoplasms can be identified in their early stages, during follow-up for the first primary malignant tumor and even if the symptomatology is vague or the findings incidental. Also, the combined approach (with or without a neoadjuvant) to cancer therapy has yielded better long-term survival. In turn, compared to 30 years ago, more patients are being diagnosed with a second primary tumor [1-4].

Background A short right renal vein complicates transplantation causing traction and difficulties during anastomosis. When we perform a kidney transplantation from a cadaveric donor, this problem may be resolved by using the vena cava to create a venous duct. This elongation technique is proposed to be performed during bench surgery. We propose a small change in the technique: execution of an “elongation patch” during harvesting and under cold perfusion. Materials and Methods From January 2004 to June 2006, we performed 12 in situ (during the harvesting procedure) vena cava elongation patches. In these cases, the right renal vein was too short. The elongation patch was used in only 8 transplantations. In the other 4 cases we sacrificed the patch to perform a direct venous anastomosis due to favorable recipient anatomical characteristics. Results The 8 transplantations performed with the elongation patch did not show vascular complications and the venous anastomosis was easy to perform. Conclusions The “extension patch” was not associated with a greater incidence of vascular complications. Using the elongation patch during the harvest showed some advantages: performed during cold perfusion with a reduction in bench ischemia; the anatomical relationships are preserved so we can perform a calibrated suture; the perfusion of the organ allows us to observe the integrity of the anastomosis. This technique did not significantly increase the harvesting time.

Given the recent advances in anesthesia, surgery, and monitoring technology, day surgery (DS) offers potential advantages for elderly patients undergoing elective surgery. Epidemiological data show a continuous expansion of elderly population, associated with an increased demand for surgical treatments by older patients. This review summarizes current selection criteria, anaesthesiology techniques and methods for perioperative management in geriatric outpatient surgery.