As dimensões continentais de alguns países e a distribuição heterogênea da rede hospitalar dificultam o acesso ao atendimento inicial adequado ao trauma de face principalmente aos pacientes residentes em regiões mais remotas. Um modelo de atendimento por telemedicina pode ser uma opção ao atendimento especializado presencial. Os objetivos deste trabalho são apresentar um modelo de atendimento por especialistas à distância, por meio de videoconferência utilizando-se smartphone, e analisar a concordância deste atendimento por telemedicina com o atendimento presencial, considerado padrão ouro. Cinquenta pacientes com trauma de face e suspeita de fratura de face (n=50) foram atendidos, por duas equipes independentes, uma presencialmente e outra por telemedicina. A equipe presencial, que era formada pela equipe de plantão de cirurgia plástica na unidade de emergência, prestou o atendimento à beira do leito (história, exame físico e análise de imagens por tomografia computadorizada). A equipe telemedicina, composta por um médico generalista atendendo à beira do leito, prestou atendimento em conjunto com um cirurgião plástico à distância através de videoconferência com um smartphone. Após cada atendimento as duas equipes responderam a um questionário, com informações sobre dados epidemiológicos, dados do exame físico, sobre indicações de tomografia computadorizada, achados radiológicos da tomografia e conduta. Os dados foram analisados e comparados quanto à concordância das respostas através da análise estatística kappa, cálculo de acurácia, especificidade e sensibilidade. A amostra estudada foi representativa e concordante com a literatura, com predomínio de homens jovens. Acidentes de trânsito e violência interpessoal foram os principais mecanismos causadores do trauma. A concordância das respostas para os achados de exame físico foi considerada substancial (kappa=0,720), para a indicação da tomografia computadorizada foi quase perfeita (kappa=0,957), para os achados na tomografia foi quase perfeita (kappa=0,899) e para definir a conduta também foi quase perfeita (kappa=0,891). A alta concordância dos achados radiológicos nas tomografias computadorizadas de face também foi observada ao se calcular os valores preditivo positivo (VPP=89,9%), preditivo negativo (VPN=99,3%), sensibilidade (94,2%), especificidade (98,8%) e acurácia (98,3%). O estudo concluiu que o modelo de atendimento ao trauma de face à distância por videoconferência via smartphone é factível, encontrando altos índices de concordância quando comparados ao atendimento padrão ouro presencial, sendo uma opção ao atendimento para a triagem de pacientes vítimas de trauma de face em áreas remotas que não têm à disposição o atendimento especializado presencial / The continental size of some countries and heterogeneous distributed hospital network prevent many patients who live in remote areas from getting adequate initial assessment of facial trauma. The author presents a model for trauma assessment through telemedicine, which may be an alternative to face-to-face specialized attending. The goals of this study are presenting a model for non-attending specialized assessment through video conference via smartphones, and analyzing a comparison between telemedicine and face-to-face management, the latter currently being the gold standard. Fifty patients with either a confirmed or suspected diagnosis of facial trauma (n=50) were evaluated by two teams of physicians, one face-to-face and the other one telemedicine-based. The face-to-face team, which was made up by the attending plastic surgery team in the emergency unit, attended the patients at the bedside (physical examination and CT-scan analysis). The telemedicine team was made up by an in-house general practitioner working together with an on-call plastic surgeon through videoconference via smartphones. After each evaluation, both teams answered a similar questionnaire, which contained data concerning the patient\'s epidemiology, physical examination, CT-scan indications and findings, and the treatment option to be followed. The data were analyzed and compared regarding the similarity of answers, with the use of kappa statistics and analysis of data accuracy, sensitivity and specificity. The sample studied was representative and consistent with the literature, showing a predominance of young males. Traffic accidents and personal violence were the main causes of trauma. The agreement of answers for physical examination findings was considered substantial (kappa=0.720). For CT-scan indications, it was considered almost perfect (kappa=0.957); for CT-scan findings, it was almost perfect (kappa=0.899); and for defining the treatment option, it was also almost perfect (kappa=0.891). High concurrency of face CT-scan findings was also observed after we calculated the positive predictive value (PPV=89.9%), negative predictive value (NPV=99.3%), sensitivity (94.2%), specificity (98.8%) and accuracy (98.3%). The study concluded that the model for non-attending assessment of facial trauma through video conference via smartphones is feasible, showing high concurrence rates when compared to gold-standard in-house assessment, thus being an option for first assessment of facial trauma patients who live in remote areas, where specialized medical teams are not available

藉由研究無線通信產業標準的制訂過程與主流標準的形成機制，本研究探討過去有關影響產業標準建立的文獻與個案，提出系統發展過程中，各影響因素扮演的角色與提供的功能，以瞭解在無線通信產業中，影響一標準脫穎而出成為市場主流的關鍵因素。本研究主要的研究發現如下： 一、政府的政策對標準後續發展相當重要，若政府態度是採開放競爭，則將在境內形成多種標準競爭而耗掉大部分能量，未來在標準向外推展上較吃虧。反之政府若能主導統一標準，常能使相關資源更集中焦點來發展單一標準，未來對外推展上較能成功。 二、協會能否發揮協商能力，使各業者提出之標準能相結合，是主流標準形成之關鍵之一，協會若能使各業者放棄成見，取得共識形成單一標準將有助於標準之發展。 三、對標準主導者而言，若能成為主流標準確實擁有很多潛在利益，為爭取標準的支持度，便會有很多遊說及策略聯盟產生，若能取得越多成員支持，常較能形成主流標準，因此標準擁有者若能降低權利金門檻，甚至將其免費授權並公開架構，標準成為市場主流之成長速度將會加快。但若成員過於複雜也容易阻礙標準演進的進度。 四、第三代行動電話的應用以多媒體為主，這是相當革命性的嚐試，故能否找到消費者公認的殺手級應用服務（killer application）成為成敗關鍵要素之一。 五、標準形成之發源地需有足夠的經濟規模量的支持，才能累積足夠資源向境外擴展。 六、標準形成在發酵期必須取得單一標準共識，在漸進變化期於境內快速建立成功經驗，並順利將成功複製於境外地區，這個模式較能成功成為主流標準。另在建立標準時需有明確的願景與目標、具國際化的規格設計，才可減少在對外擴展時之障礙。 / This thesis is intended to make explicit the roles played and functions provided by the various factors in the process of system development by means of an investigation of the process of the formation of a wireless communications standard and the mechanism that makes it a dominant. By studying the documents and cases on the factors affecting the establishment of a standard, we have come to recognize six fundamentals that make it stand out as a dominant. These six fundamentals are summarized as follows: 1. The policy made by the government plays an essential part in the further development of the formation of a given standard. If the government adopts an attitude of open competition, then lots of related resources will be wasted on the keen competition between vying forces in the local market, which is, in turn, detrimental to outward expansion. On the contrary, if the government integrates the related resources and helps form a unitary standard, the standard may succeed in opening other markets. 2. The Association is also an important positive contributor in the process of the formation of a unitary standard. The Association should do its best to mediate between each vying standard to reach an agreement and adopt a unitary standard, which is definitely instrumental in the further development. 3. For those who want to direct the formation of a standard and get the tremendous potential profits from it, they will do a lot of lobbying and form strategic alliances to attract more supports for the standard. When succeeding in doing so, the owners of the unitary standard can speed up the growth of its market by lowering the royalty and license limit, or even by opening its architecture with a free authorization. 4. The application of the third-generation cellular phones features multi-media, which is a considerably revolutionary attempt. Therefore, the success of it depends on whether the standard can come up with a killer application that appeals to consumers. 5. The local market from which the standard originates should be big enough to support it, so the standard can stand firm on the strong basis and reach out for other markets. 6. The formation of a standard should be based on an agreement of a unitary standard at the ferment stage. And at the transition stage, it should build up a successful experience as soon as possible in the local market and then successfully duplicate the model in the international ones. Besides, a definite goal and a clear vision as well as internationalized regulations and designs all contribute to lifting the barriers in its expanding outwards.

No description available.

As dimensões continentais de alguns países e a distribuição heterogênea da rede hospitalar dificultam o acesso ao atendimento inicial adequado ao trauma de face principalmente aos pacientes residentes em regiões mais remotas. Um modelo de atendimento por telemedicina pode ser uma opção ao atendimento especializado presencial. Os objetivos deste trabalho são apresentar um modelo de atendimento por especialistas à distância, por meio de videoconferência utilizando-se smartphone, e analisar a concordância deste atendimento por telemedicina com o atendimento presencial, considerado padrão ouro. Cinquenta pacientes com trauma de face e suspeita de fratura de face (n=50) foram atendidos, por duas equipes independentes, uma presencialmente e outra por telemedicina. A equipe presencial, que era formada pela equipe de plantão de cirurgia plástica na unidade de emergência, prestou o atendimento à beira do leito (história, exame físico e análise de imagens por tomografia computadorizada). A equipe telemedicina, composta por um médico generalista atendendo à beira do leito, prestou atendimento em conjunto com um cirurgião plástico à distância através de videoconferência com um smartphone. Após cada atendimento as duas equipes responderam a um questionário, com informações sobre dados epidemiológicos, dados do exame físico, sobre indicações de tomografia computadorizada, achados radiológicos da tomografia e conduta. Os dados foram analisados e comparados quanto à concordância das respostas através da análise estatística kappa, cálculo de acurácia, especificidade e sensibilidade. A amostra estudada foi representativa e concordante com a literatura, com predomínio de homens jovens. Acidentes de trânsito e violência interpessoal foram os principais mecanismos causadores do trauma. A concordância das respostas para os achados de exame físico foi considerada substancial (kappa=0,720), para a indicação da tomografia computadorizada foi quase perfeita (kappa=0,957), para os achados na tomografia foi quase perfeita (kappa=0,899) e para definir a conduta também foi quase perfeita (kappa=0,891). A alta concordância dos achados radiológicos nas tomografias computadorizadas de face também foi observada ao se calcular os valores preditivo positivo (VPP=89,9%), preditivo negativo (VPN=99,3%), sensibilidade (94,2%), especificidade (98,8%) e acurácia (98,3%). O estudo concluiu que o modelo de atendimento ao trauma de face à distância por videoconferência via smartphone é factível, encontrando altos índices de concordância quando comparados ao atendimento padrão ouro presencial, sendo uma opção ao atendimento para a triagem de pacientes vítimas de trauma de face em áreas remotas que não têm à disposição o atendimento especializado presencial / The continental size of some countries and heterogeneous distributed hospital network prevent many patients who live in remote areas from getting adequate initial assessment of facial trauma. The author presents a model for trauma assessment through telemedicine, which may be an alternative to face-to-face specialized attending. The goals of this study are presenting a model for non-attending specialized assessment through video conference via smartphones, and analyzing a comparison between telemedicine and face-to-face management, the latter currently being the gold standard. Fifty patients with either a confirmed or suspected diagnosis of facial trauma (n=50) were evaluated by two teams of physicians, one face-to-face and the other one telemedicine-based. The face-to-face team, which was made up by the attending plastic surgery team in the emergency unit, attended the patients at the bedside (physical examination and CT-scan analysis). The telemedicine team was made up by an in-house general practitioner working together with an on-call plastic surgeon through videoconference via smartphones. After each evaluation, both teams answered a similar questionnaire, which contained data concerning the patient\'s epidemiology, physical examination, CT-scan indications and findings, and the treatment option to be followed. The data were analyzed and compared regarding the similarity of answers, with the use of kappa statistics and analysis of data accuracy, sensitivity and specificity. The sample studied was representative and consistent with the literature, showing a predominance of young males. Traffic accidents and personal violence were the main causes of trauma. The agreement of answers for physical examination findings was considered substantial (kappa=0.720). For CT-scan indications, it was considered almost perfect (kappa=0.957); for CT-scan findings, it was almost perfect (kappa=0.899); and for defining the treatment option, it was also almost perfect (kappa=0.891). High concurrency of face CT-scan findings was also observed after we calculated the positive predictive value (PPV=89.9%), negative predictive value (NPV=99.3%), sensitivity (94.2%), specificity (98.8%) and accuracy (98.3%). The study concluded that the model for non-attending assessment of facial trauma through video conference via smartphones is feasible, showing high concurrence rates when compared to gold-standard in-house assessment, thus being an option for first assessment of facial trauma patients who live in remote areas, where specialized medical teams are not available

L'intégration de nombreuses fonctions complexes dans les systèmes embarqués, tels que les téléphones portables, conduit à optimiser la consommation d'énergie pour maintenir l'autonomie de fonctionnement. Concernant la chaine de reproduction sonore, la consommation a été réduite par l'utilisation d'amplificateurs de classe D analogique, mais la nature numérique de la source audio impose encore un convertisseur numérique analogique en amont. La forte consommation de ce CAN et la qualité de reproduction sonore de l'amplificateur sont actuellement les principales limitations de cette approche. Ce travail de recherche, réalisé en quatre phases, a donc pour objectifs d'apporter des améliorations et de proposer de nouvelles architectures pour réduire ces limitations. Concernant l'amplificateur de classe D analogique, une boucle d'asservissement analogique basée sur un modulateur auto-oscillant à hystérésis a été développée pour réduire la consommation et augmenter sa qualité de reproduction. Cette étude a été validée par la réalisation d'un circuit en technologie CMOS 130 nm. La possibilité de piloter l'amplificateur de classe D directement par un signal de commande numérique a ensuite été envisagée. Le train d'impulsions nécessaire à la commande de l'étage de puissance est obtenu par modulation numérique de la source audio. L'utilisation en boucle ouverte de l'étage de puissance ne permet néanmoins pas d'obtenir un signal audio de sortie insensible aux variations de l'alimentation. Un asservissement analogique local autour de l'étage de puissance est donc nécessaire. La réalisation en technologie CMOS 130 nm de cette architecture a permis la validation des études effectuées (fonctionnement, stabilité, bande passante, modélisations des éléments non linéaires,...). / The integration of many complex functions in embedded systems such as mobile phones, led to optimize energy consumption to maintain operational autonomy. Concerning the chain of sound reproduction, consumption was reduced by the use of analog Class D amplifiers, but the nature of the digital audio source requires a digital to analog converter further upstream. The high consumption of the ADC and the quality of sound reproduction of the amplifier are currently the main limitations of this approach. This research, conducted in four phases, thus aims to make improvements and propose new architectures to reduce these limitations. Regarding the Class D amplifier analog control loop an analog modulator based on a self-oscillating hysteresis has been developed to reduce consumption and increase its quality of reproduction. This study was validated by the realization of a circuit in CMOS 130 nm. The ability to control the class D amplifier directly from a digital control signal was then considered. The train of pulses needed to control the power stage is obtained by digital modulation of the audio source. The use of open-loop output stage, however, does not produce an audio signal output insensitive to variations in the diet. A local analog servo around the power stage is required. Realization in CMOS 130 nm of this architecture has allowed the validation studies (operation, stability, bandwidth, modeling non-linear elements ,...). The interface between the digital modulator and the analog part is very sensitive to disturbance. A digital control overall Class D amplifier with digital control has been studied to control the interface. A prototype of this control is in progress.

Keyword spotting is one of the many applications of automatic speech recognition. Its purpose is determining spots in given utterance in which some of the specified words were spoken. Keyword spotting has a great potential to enhance performance of new applications as well as the existing ones. An example could be a mobile phone voice control. Due to OS Symbian's coming to the market it is even possible for end user to implement a keyword spotting for a mobile phone on his or her own. The thesis describes theoretical prerequisites for keyword spotting and its implementation. Firstly the OS Symbian is presented with respect to the given task. Secondly each step of keyword spotting process is described. Finally the object design of keyword spotter is presented followed by implementation description. The thesis concludes with results review and notes on possible improvements.