Diagnóstico agroambiental rápido dos setores de produção e processamento da cana-de-açúcar nos estados do Maranhão e Piauí.

2008

Smart Textile Sensors for Healthcare Monitoring

Wearable electronic textiles are an emerging research field playing a pivotal role among several different technological areas such as sensing, communication, clothing, health monitoring, information technology, and microsystems. The possibility to realise a fully-textile platform, endowed with various sensors directly realised with textile fibres and fabric, represents a new challenge for the entire research community. Among several high-performing materials, the intrinsically conductive poly(3,4-ethylenedioxythiophene) (PEDOT), doped with poly(styrenesulfonic acid) (PSS), or PEDOT:PSS, is one of the most representative and utilised, having an excellent chemical and thermal stability, as well as reversible doping state and high conductivity. This work relies on PEDOT:PSS combined with sensible materials to design, realise, and develop textile chemical and physical sensors. In particular, chloride concentration and pH level sensors in human sweat for continuous monitoring of the wearer's hydration status and stress level are reported. Additionally, a prototype smart bandage detecting the moisture level and pH value of a bed wound to allow the remote monitoring of the healing process of severe and chronic wounds is described. Physical sensors used to monitor the pressure distribution for rehabilitation, workplace safety, or sport tracking are also presented together with a novel fully-textile device able to measure the incident X-ray dose for medical or security applications where thin, comfortable, and flexible features are essential. Finally, a proof-of-concept for an organic-inorganic textile thermoelectric generator that harvests energy directly from body heat has been proposed. Though further efforts must be dedicated to overcome issues such as durability, washability, power consumption, and large-scale production, the novel, versatile, and widely encompassing area of electronic textiles is a promising protagonist in the upcoming technological revolution.

The birth of a tragic hero(ine) On tragic excess in contemporary theatre

The idea of the tragic is unthinkable. It is precisely within the moment in which an ordinary human being, a heroine or a hero – incapable of scrutinizing fully their own position within the whole – is invited to respond, to accept or refuse it all, that the tragic unfolds, changing their life irremediably. What are the causes and the consequences of "god’s arrival", as in case of Dionysus who visits Pentheus’ home in Euripides’ "The Bacchae"? Through episodes in the stories of characters from Ancient Greek dramas – such as Oedipus, Antigone, Ajax, Io, through Dostoevsky’s or Kafka’s imagery, in Prince Myshkin’s, the Ridiculous Man’s or Gregor Samsa’s experiences, this doctoral research proposes to examine the aspects which compete in the creation of a tragic hero. Theatrical performances – such as Jan Fabre’s "Mount Olympus: To Glorify the Cult of Tragedy, a 24-Hour Performance", immersed in a cycle of life, death and re-birth; Oliver Frljić’s "Trilogija o hrvatskom fašizmu", in its careful analysis of the wounds of a heritage of war; and Cristian Ceresoli’s and Silvia Gallerano’s tragic testimony of an estranged, almost soulless body in "La Merda" – open up the dialogue on our contemporary idea of the tragic. This doctoral work chooses excess as its privileged channel through which to approach the concept of the tragic – by its nature elusive, hostile to any definition, strictly personal and, thus, visible only through one’s own lens. In an excess of pain, devotion, desire, rage, arrogance or beauty, opposites collide, time concentrates into a moment and the hero is invited to choose, to live or die, to transform.

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.