5603 research outputs found
Sort by
Learning Variational Models via Bilevel Optimization and Unfolded Algorithms
No full textI modelli variazionali rappresentano uno degli strumenti classici per risolvere problemi inversi lineari in vari contesti. In contrasto con altri metodologie più moderne, essi poggiano basi teoriche che sono state studiate per decenni. La maggior interpretabili che offrono, combinata con una minor richiesta in termini di risorse computazionali fa si che lo studi di questi metodi sia ancora rilevante al giorno d'oggi. In questo lavoro, per migliorare la performance di questi modelli, ci soffermiamo sui metodi di selezione dei loro parametri. È ben noto che i funzionali che costituiscono i modelli variazionali dipendono da uno o più parametri di cui però non si hanno regole precise su come sceglierli in modo efficace. Prendendo spunto da tecniche nell'ambito dell'apprendimento automatico, andiamo ad imparare i modelli variazioni con lo scopo di trovare in modo automatico una configurazione soddisfacente di parametri. In questo lavoro proponiamo un algoritmo di ottimizzazione di tipo forward-backward, di cui dimostriamo la convergenza ad un punto stazionario di un problema di ottimizzazione bilivello per l'apprendimento di un modello variazionale. Gli esperimenti numerici mostrano che l'algoritmo riesce a raggiungere una buona soluzione in poco tempo. Per aggirare alcuni limiti di questo approccio, modifichiamo il problema inferiore inserendo, invece che una minimizzazione completa, l'unfolding di un metodo di ottimizzazione iterativo. Il numero di iterazioni che compie tale metodo è fissato a priori. L'apprendimento di questo algoritmo "srotolato" mira a trovare la migliore configurazione di parametri (anche quelli dell'algoritmo stesso!) per sfruttare al meglio le iterazioni concesse. Abbiamo testato queste idee su diversi problemi di ricostruzione di immagini, con buoni risultati, soprattutto in confronto con altri metodi di tipo deep.Variational models are a classical tool to solve inverse problems in a multitude of contexts. In contrast to other modern methodologies, they lay their foundations on an established theoretical background. The advantage offered by their interpretability, combined with a less intense requirement in terms of computational resources makes the study and use of them still relevant to this day. To improve the performance of these models, we address one of their main weaknesses: parameters selection. These models are defined through an energy function which itself is characterized by one or more variables. The absence of broad rules to set these variables often leads to a tedious, and perhaps time consuming, empirical search for a good configuration. Borrowing ideas from the machine learning realm to automate this process, we effectively train the energy functional with the goal of finding a more than satisfying parameter setup. We study a general formulation of a bilevel optimization problem to carry out this task in a variety of imaging applications. In detail, we develop an iterative general purpose inexact forward-backward algorithm which is able to converge to a stationary point of the bilevel problem. The experiments show that the algorithm is able to find more than valid solutions in a small amount of time. Although these results are promising, there are still limitations to the approach. To circumvent some of these, we also relax the lower level minimization problem by replacing it with the unfolding of an iterative optimization algorithm. The number of iterations that the chosen scheme performs is fixed a priori, so that the whole lower level resembles a very basic neural network. The training of this algorithm unfolding then aims at finding the best parameters (including some of the optimization algorithm itself!) to make the most out of the given iterations. We tested these ideas on different imaging problems, with surprising results, especially when compared to deep learning methods
Interferenza di memoria e previsione delle prestazioni nei sistemi eterogenei accelerati da GPU.
No full textNowadays, a variety of applications, including automated factories, autonomous vehicles, and Cyber Physical Systems (CPS), are experiencing significant growth. Given the diverse range of challenges that must be addressed, such as real-time management and visualization of a factory's current state through a 3D digital twin, trajectory calculation within autonomous vehicles, visualizing Human Machine Interfaces (HMI), traffic management in smart cities equipped with cameras, IoT devices, and their associated features, a broad array of \emph{heterogeneous devices} with various \emph{hardware accelerators} are being utilized to solve these different problems. In such applications, \emph{power consumption} and \emph{task execution latency} are key aspects to consider, hence investigating approaches that mitigate power consumption while still fully utilizing the computational power provided by the devices becomes necessary.
Modern devices use hardware processors that enable the acceleration of highly parallel and data hungry computational workloads; a widely known example of such parallel processor is the \emph{Graphic Process Unit} (GPU), a hardware peripheral traditionally used for graphics rendering but nowadays it is also used as a general purpose compute accelerator.
This thesis addresses an analysis of the state of the art of techniques that can be employed to \emph{optimize power consumption} and task execution latency, as well as two types of latencies/interference that tasks can potentially experience:
latencies arising from tasks that are concurrently scheduled on the same acceleration unit, i.e., on a partitioned GPU, and the second type under consideration is the latencies experienced by tasks running on embedded boards, specifically on GPU-embedded systems, with a high computational load on the CPU side. Methods are proposed to understand and derive predictive models for latencies in both of the two types of interference.
Furthermore, this thesis concludes with a comparative study of two GPU memory management methodologies: explicit copies versus unified virtual memory.Oggigiorno, una varietà di applicazioni, tra cui fabbriche automatizzate, veicoli autonomi e Sistemi Cyber Fisici (CPS), stanno vivendo una crescita significativa. Date le diverse sfide che devono essere affrontate, come la gestione in tempo reale e la visualizzazione dello stato attuale di una fabbrica attraverso un gemello digitale 3D, il calcolo della traiettoria all'interno dei veicoli autonomi, la visualizzazione delle Interfacce Uomo-Macchina (HMI), la gestione del traffico nelle città intelligenti dotate di telecamere, dispositivi IoT e relative funzionalità, viene utilizzata una vasta gamma di dispositivi eterogenei con vari acceleratori hardware per risolvere questi diversi problemi. In tali applicazioni, il consumo energetico e la latenza nell'esecuzione delle attività sono aspetti chiave da considerare, pertanto diventa necessario investigare approcci che mitigano il consumo energetico pur utilizzando appieno la potenza computazionale fornita dai dispositivi.
I dispositivi moderni utilizzano processori hardware che consentono l'accelerazione di carichi di lavoro computazionali altamente paralleli e avidi di dati; un esempio ampiamente conosciuto di tale processore parallelo è l'Unità di Elaborazione Grafica (GPU), un periferico hardware tradizionalmente utilizzato per il rendering grafico ma oggi utilizzato anche come acceleratore di calcolo generale.
Questa tesi affronta un'analisi dello stato dell'arte delle tecniche che possono essere impiegate per ottimizzare il consumo energetico e la latenza nell'esecuzione delle attività, nonché due tipi di latenze/interferenze che le attività possono potenzialmente sperimentare: latenze generate da attività pianificate contemporaneamente sulla stessa unità di accelerazione, cioè su una GPU partizionata, e il secondo tipo preso in considerazione è rappresentato dalle latenze sperimentate dalle attività in esecuzione su schede embedded, in particolare su sistemi embedded con GPU, con un elevato carico computazionale sul lato della CPU. Vengono proposti metodi per comprendere e derivare modelli predittivi per le latenze in entrambi i tipi di interferenze.
Inoltre, questa tesi si conclude con uno studio comparativo di due metodologie di gestione della memoria GPU: copie esplicite versus memoria virtuale unificata
Respirable nano/micro systems for prevention and treatment of pulmonary infectious diseases
No full textBACKGROUND - Pulmonary drug delivery is a promising substitute for traditional drug delivery routes, which proved efficient in treating local and systemic diseases, including tuberculosis. The lungs may also represent the target site for administering active ingredients intended to prevent infectious diseases, for example, stimulating the immune response. Among immunostimulants imiquimod, now approved for topical administration and employed extensively in off-label therapeutic trials, is appealing. However, it has been successfully formulated to date only as semi-solid lipophilic formulations for topical application, is poorly soluble in most solvents, and solubility literature data are discordant.
AIM - This work aims to investigate imiquimod solubility in solvents suitable to develop innovative formulations and to clarify literature discrepancies to collect reliable data. Another aim is to design an imiquimod powder suitable for inhalation and helpful in preventing infectious diseases due to its ability to stimulate innate immunity in the lungs, but also potentially valuable for fighting them, as an adjuvant in allergen immunotherapy or intended to be combined with an antigen for the formulation of vaccines for inhalation or nasal administration. Finally, this work aims to exploit the benefits of inhalation administration by designing a spray-dried powder of microparticles embedding innovative pyrazinoic acid nanoconjugates targeting alveolar macrophages with the final aim of improving the actual, disadvantageous treatment of tuberculosis.
RESULTS - Imiquimod solubility in water, ethanol, methanol, acetonitrile, and dimethyl sulfoxide was accurately determined at different temperatures. Experimental conditions like temperature and stirring time were found to significantly affect the time required to achieve complete dissolution. The Scatchard-Hildebrand equation does not apply to imiquimod solutions studied because of association phenomena due to intermolecular hydrogen bonds and/or π-stacking, as supported by the hyperchromic effect very pronounced in highly polar solvents. Micronization and spray drying were found to be suitable methods for producing highly crystalline and respirable imiquimod inhalation powders, with the possibility to modify the administered dose by making adhesive mixtures with coarse lactose, which proved to maintain a favorable aerodynamic behavior. In vitro studies showed powders’ non-toxicity and ability to induce a controlled production of IL-6 and IL-8. Conjugates between pyrazinoic acid and poly (malic acid) proved to self-assemble, after simple nanoprecipitation in water, into nanometric, negatively surface-charged, spherical, and stable over time nanoconjugates, which gave a sustained release of active ingredients. Nanoconjugates proved to be degraded if exposed to aqueous enzyme solutions, which would reasonably happen also inside alveolar macrophages. A potentially respirable spray-dried microparticle powder was produced starting from a solution of these nanoconjugates, leucine, and sodium hyaluronate, and proved to be able to yield original nanoconjugates upon redispersion in aqueous media, maintaining drug concentration over time and avoiding burst release that usually affects nanosystems, and reducing doses and frequency of administration, reducing side effects, and improving tuberculosis therapy.
CONCLUSION - This thesis highlights the peculiar imiquimod behavior in solution for the first time, revealing the association between imiquimod molecules in very polar solvents and providing reliable solubility data. Nanoparticle-based imiquimod crystalline and respirable powders were produced, proposing for the first time an inhalation administration of this active ingredient with the primary aim of designing an inhalation platform that stimulates the immune response to prevent infectious diseases. Finally, a significant step towards improving tuberculosis therapy has been made by developing an inhalation powder of microparticles embedding innovative nanoconjugates of pyrazinoic acid that target alveolar macrophages, providing a sustained release of active ingredients
L'esperienza spirituale dei professionisti della salute in contesti di emergenza dopo un corso di spiritualità. Uno studio longitudinale qualitativo
No full textIn emergency clinical settings, health professionals can be overwhelmed, and the practical care provision might be hindered. Literature underlines the need for specific skills to develop education in emergency and critical care settings focused on health professionals’ spiritual competence.
Prospective longitudinal qualitative research was carried out to evaluate a training format on spirituality. 25 semi-structured interviews to health professionals working in emergency and critical care settings, were performed before and after a delivered webinar on spirituality. Data were analyzed through thematic analysis.
Results confirmed a shift in meaning in four central themes defined as: (1) definition of spirituality in personal experience; (2) how to recognize one’s own spirituality and that of others; (3) the personal and professional development of one’s own spirituality; (4) expectations on the training course of spirituality.
Findings underline changing point of view on spirituality experience through an increased self-awareness. A compassionate connection with oneself, others, nature, and spiritual practices led to a spiritual grow. The course enabled to identify specific training needs in emergency settings. Health professionals require tools to expand the personal spiritual skill during clinical healthcare approach. Research should be integrated with multiple time sessions reflecting the integrated training format of multiple theoretical and experiential elements, discussion groups and longer-term training
Le formiche e li pest agricoli nel Mediterraneo: esplorando i paradigmi di antagonismo e mutualismo per il controllo biologico
No full textAnts are insects with an almost unparalleled ecological impact on terrestrial ecosystems, providing many important ecosystem services through their complex interactions with countless other organisms. Ant role in agroecosystems has drawn significant attention by humans for centuries but is still better investigated for the tropics as compared to temperate regions. Most ants are generalist predators, and their predatory abilities have also been appreciated for their antagonistic effects on many agricultural pest insects. At the same time, ants establish mutualistic relationships with several honeydew-producing hemipterans, forming keystone ecological associations that can also develop into a problem for agricultural settings in which some ant-mutualist hemipterans are important pests. We explore ecological and behavioral interactions between ants and agricultural pest insects through antagonism and mutualism paradigms using laboratory and field experiments, while also gathering baseline data on the diversity and distribution of ant species in agroecosystems across Italy. We present our results through eleven chapters grouped in three sections: i) Ants as biological control agents of insect pests; ii) Ants as mutualist partners of insect pests; iii) First data on ant diversity and distribution in Italian agroecosystems. In the first section, we describe for the first time the role of ants as natural enemies of key agricultural pests accounting both direct and direct interactions and analyzing potential conflicts with other natural enemies, while also presenting encouraging results on behavioral manipulation of ants through artificial nectaries. In the second section, we describe ecological and behavioral aspects of the interactions between ants, honeydew-producing mutualist hemipterans, and their natural enemies, emphasizing the variety of outcomes that the species-specific characteristics of the involved actors may produce. Furthermore, we present a comparative review of the methodologies developed to manage ant-hemipteran mutualism when this is responsible for significant pest outbreaks. In the third section, we present first data on ant diversity in Italian agroecosystems as well as the inclusion of the ant species studied in this thesis in the DNA barcoding library of European ants
FOODTURE: Analysis and development of the social fridge "Frigo di Quartiere" project by Im.patto Nova Coop
No full textIl presente lavoro di tesi analizza l'iniziativa “Frigo di Quartiere”, nata con Nova Coop per ridurre lo spreco alimentare e sostenere le comunità vulnerabili, attraverso la redistinazione delle eccedenze dai punti vendita ai Frigo. L’obiettivo è realizzare un toolkit per Nova Coop facilitando la creazione di un manuale destinato alle cooperative, con raccomandazioni su formazione, linee guida, exhibit design e riconoscibilità, per favorire la replicabilità del modello in contesti già esistenti e nuovi.This thesis work analyzes the “Frigo di Quartiere” initiative established with Nova Coop to reduce food waste and support vulnerable communities through the redistination of surplus from stores to Fridges. The goal is to create a toolkit for Nova Coop by facilitating the creation of a manual for use by cooperatives, with recommendations on training, guidelines, exhibit design and recognizability, to encourage replicability of the model in existing and new contexts
Metals in medicine: bioinorganic chemistry to fight infectious diseases and cancer.
No full textMetals have played a vital role in the field of medicine, underscoring the enduring connection between inorganic chemistry and healthcare. This thesis embarks on an exciting journey, harnessing the tools of bioinorganic chemistry, to address pressing biomedical healthcare challenges. With a primary aim to develop innovative metal-based compounds for diverse applications, the research explores antimicrobial resistance, antibacterial drug discovery, cancer therapeutics, and antiparasitic agents. Along this journey, novel bismuth(III), gallium(III), antimony(III), manganese(I), and gold(III) complexes have been synthesized and rigorously studied, providing valuable insights for future advancements. This thesis represents a significant stride towards simplifying the underlying chemistry, enhancing the assessment of biological activity, and fostering hope for pioneering therapeutic solutions in the ever-evolving healthcare landscape
I depositi di lobi torbiditici in bacini di avanfossa tettonicamente confinati (Formazione Marnoso-arenacea, Appennini Settentrionali)
No full textll lavoro presentato in questa tesi di è incentrato sullo studio dei depositi torbiditici di età serravalliana della Formazione Marnoso-arenacea che affiorano nell’Appennino Settentrionale. Tali depositi, conosciuti in letteratura come sistema di Firenzuola o Unità V, registrano la progressiva chiusura del bacino interno della FMA ed il conseguente spostamento del depocentro verso il bacino esterno a causa dell’innalzamento simultaneo del thrust di Monte Castellaccio e dell’alto di Verghereto. Geometrie e facies dei depositi torbiditici dell’Unità V, inoltre, risultano anche influenzati dalla messa in posto del Caotico di Casaglia, che nella sua area di origine raggiunge uno spessore massimo di 500 metri.
L’analisi dei depositi dell’Unità V è stata perseguita mediante un lavoro di terreno che ha riguardato la misurazione e l’analisi di facies di dettaglio di 28 sezioni stratigrafiche, situate nelle Valli del Santerno e del Senio, e che misurano 3000 metri di spessore complessivo. L’attività di terreno è stata inoltre corroborata da attività di laboratorio che ha riguardato l’analisi granulometrica di 260 campioni utilizzando un granulometro a diffrazione laser.
Le sezioni misurate sono state correlate a formare un pannello stratigrafico generale dell’area di studio; tale pannello, che comprende l’intero spessore stratigrafico dell’Unità V, è stato suddiviso in due sotto-Unità (Va e Vb) grazie alla presenza del caotico di Bedetta, osservato per la prima volta in Val Santerno durante lo svolgimento di questo lavoro. Le due sotto-Unità individuate differiscono profondamente sia in termini stratigrafici che di facies. Le torbiditi sottostanti al suddetto caotico hanno caratteristiche similari a depositi di basin plain, nell'Unità Vb si registra un drastico incremento del rapporto arenaria/pelite, di strutture quali mud draped scour, di strati arenacei da massivi a pseudolaminati mal classati e con granulometria medio-grossolana, spesso caratterizzati dall'abbondate presenza di clasti pelitici, strutture a fiamma e superfici di bypass.
Combinando i dati relativi alle facies osservate con le correlazioni effettuate, è stato possibile suddividere gli strati osservati in 7 categorie, a loro volta organizzati in due distinti facies tract, uno per ciascuna delle sotto-Unità individuate. Scopo dei facies tract è stato quello di descrivere l’evoluzione sottocorrente dei flussi che hanno portato alla formazione di questi depositi.
La costruzione di 5 pannelli stratigrafici di dettaglio all’interno dell’Unità Vb ha consentito la suddivisione di quest’ultima in 17 depositi di lobo, separati dai corrispettivi intervalli fini di interlobo. Dove possibile inoltre, i lobi sono stati ulteriormente suddivisi nei relativi elementi di lobo che li compongono. Da un'analisi circa la distribuzione latero-verticale delle facies osservate nei lobi è stato infine possibile ricostruire le fasi evolutive dei lobi che compongono il sistema di Firenzuola.The work presented in this thesis is focused on the study of the serravallian turbidite deposits of the Marnoso-arenacea Formation, which exensively crop out in the Northern Apennines (Italy). These deposits, known in literature as Firenzuola system of Unit V, record the progressive infill of the inner basin of the Marnoso-arenacea Formation and the consequent depocenter shift towards the external basin due to the simultaneous uplift of the Mt. Castellaccio thrust and the Verghereto high. Furthermore, the geometries and facies of the turbidite deposits of Unit V are also influenced by the emplacement of the Casaglia Chaotic Complex, which reaches a maximum thickness of 500 meters in its origin area.
The analysis of Unit V deposits was pursued through fieldwork involving the measurement and detailed facies analysis of 28 stratigraphic logs located in Santerno and Senio Valleys and with a total thickness of about 3000 meters. Field activities were further supported by a laboratory work, involving grain size analysis of 260 samples using a laser diffraction particle size analyser.
The measured logs were correlated using a hierarchical approach, in order to reconstruct a general stratigraphic section of the study area. This section includes the entire stratigraphic thickness of the Unit V and has been split into two sub-Units (Va and Vb) thanks to the presence of Bedetta mass transport deposit (MTD), observed for the first time in Santerno Valley during this study. The two sub-Units differ significantly in both stratigraphic and facies terms. The turbidites underlying Bedetta MTD exhibit characteristics similar to basin plain deposits, while in Unit Vb there is a drastic increase in net to gross, mud draped scours, poorly sorted and crudely laminated medium to coarse-grained beds, often characterized by abundant mudstone clasts, flame structures and bypass surfaces.
By combining facies with the stratigraphic cross sections, it was possible to classify the observed beds into 7 categories, organized into two distinct facies tracts, one for each sub-Unit. The purpose of the facies tract is that to describe the downcurrent evolution of the flows that led to the formation of these deposits.
The construction of 5 detailed stratigraphic cross sections within Unit Vb allowed the subdivision of the latter into 17 turbidite lobe deposits, separated by interlobe deposits. Where possible, depositional lobes were further subdivided into their respective lobe elements. Finally, an analysis of the lateral and vertical facies distribution observed in the lobes made it possible to reconstruct the evolutionary phases of the depositional lobes characterizing the Firenzuola system
Information seeking in bipolar disorder : the search for information on the Internet during different phases of bipolar disorder
No full textLo studio vuole indagare se esista una correlazione tra i comportamenti di information seeking e il disturbo bipolare. In particolare, sono state analizzate le cronologie delle pagine web visitate da pazienti affetti da suddetta patologia e successivamente confrontate con i dati estratti dalle relative cartelle cliniche. Dalle analisi emergono delle differenze statisticamente significative rispetto a: numero di pagine visitate, la loro coerenza tematica e la loro valenza emotiva durante le fasi attive di malattia rispetto alle fasi eutimiche. Sono state trovate delle correlazioni in 2 soggetti, su 30 reclutati, rispetto all'andamento del disturbo e la valenza delle pagine cercate. In un caso si è manifestata una correlazione diretta, ovvero la valenza era coerente con la fase della malattia, mentre nell'altro la correlazione era inversa.The study aims to investigate whether there is a correlation between information-seeking behaviors and bipolar disorder. Specifically, the browsing histories of web pages visited by patients affected by this condition were analyzed and then compared with data extracted from their medical records. The analyses revealed statistically significant differences in terms of the number of pages visited, their thematic consistency, and their emotional valence during active phases of the disorder compared to euthymic phases. Correlations were found in 2 subjects out of the 30 recruited, with respect to the course of the disorder and the valence of the pages searched. In one case, a direct correlation was observed, meaning that the valence was consistent with the phase of the disorder, while in the other case, the correlation was inverse
Qualitative description of electroencephalographic (EEG) tracings in healthy, awake donkeys
No full textL'elettroencefalografia (EEG) rappresenta uno strumento diagnostico di fondamentale importanza in neurologia, anche in medicina veterinaria. Lo scopo di questo lavoro preliminare è effettuare un’analisi dei tracciati EEG standard in nove asini, dei quali otto sani e uno in cui sono stati osservati episodi di epilessia parziale. Questo studio si propone inoltre di indagare l'efficacia di tecniche avanzate di analisi, come l'Analisi delle Componenti Indipendenti (ICA) e l'analisi della Densità Spettrale di Potenza (PSD), nel caratterizzare in modo quantitativo le registrazioni EEG in asini non sedati e tra questi uno con sospetta epilessia focale, in cui l'EEG standard non ha fornito risultati diagnostici chiari. I tracciati EEG standard non hanno rivelato anomalie significative né tra gli individui sani, né in quello con sospetta epilessia parziale. Tale risultato è frequente nella pratica clinica veterinaria a causa di artefatti muscolari, le condizioni di registrazione sub ottimali e la variabilità anatomica di specie. L'uso combinato di ICA e PSD ha permesso però di identificare differenze funzionali significative, non rilevabili nell'analisi standard, tra lo stato fisiologico e quello patologico. L'ICA ha migliorato la qualità del segnale EEG separando gli artefatti dai segnali cerebrali, evidenziando potenziali segni di epilessia focale. L'analisi PSD ha mostrato un aumento dell'attività nelle bande delta e theta e un'asimmetria nell'attività ad alta frequenza (22-30 Hz) nel soggetto patologico, coerenti con un quadro di epilessia focale. Questi risultati suggeriscono che tecniche avanzate di analisi EEG possono migliorare la diagnosi neurologica negli equidi.Qualitative description of electroencephalographic (eeg) tracings in healthy, awake donkeys (EEG) is an important diagnostic tool in neurology, in both human and veterinary medicine. The objective of this work is to conduct an analysis of standard EEG tracings in nine donkeys, with eight of them being healthy and one exhibiting episodes of partial epilepsy. This study also aims to investigate the effectiveness of advanced analysis techniques, such as Independent Component Analysis (ICA) and Power Spectral Density (PSD) analysis, in quantitatively characterizing EEG recordings in non-sedated donkeys including one with suspected focal epilepsy, in which standard EEG did not provide clear diagnostic results. Standard EEG tracings revealed no significant abnormalities between the two samples, as it frequently occurs in veterinary clinical practice due to muscle artifacts, suboptimal recording conditions and anatomical species variability. However, the combined use of ICA and PSD allowed to identify significant functional differences, not detectable in standard analysis, between the physiological and pathological conditions. ICA improved EEG signal quality by separating artifacts from brain signals, revealing potential signs of focal epilepsy. PSD analysis showed increased activity in delta and theta bands and asymmetry in high-frequency activity (22-30 Hz) in the pathological subject, consistent with a focal epilepsy condition. These results suggest that advanced EEG analysis techniques can improve neurological diagnosis in equids