Diss. (sammanfattning) Linköping : Univ., 2006. / Härtill 5 uppsatser.

Made available in DSpace on 2015-07-22T20:50:50Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-07-29 / Objetivos: Avaliar a concordância no diagnóstico das lesões cutâneas pigmentadas por meio do exame dermatoscópico presencial e a distância. Analisar a concordância diagnóstica da teledermatoscopia e do exame dermatoscópico presencial em relação ao resultado anatomopatológico. Verificar a acurácia, a sensibilidade e a especificidade da teledermatoscopia na detecção das lesões pigmentadas malignas. Métodos: Sessenta e quatro lesões cutâneas pigmentadas foram submetidas a exame clínico e dermatoscópico realizado presencialmente por dois examinadores com experiência em dermatoscopia, durante o ano de 2005, no ambulatório de dermatoscopia do Departamento de Dermatologia da Universidade Federal de São Paulo (UNIFESP). Após dois anos, imagens digitais clínicas e dermatoscópicas destes casos foram reexaminadas pelos mesmos especialistas por meio da transmissão de imagens e de dados clínicos via web. Resultados: Observou-se boa concordância diagnóstica entre o exame presencial e a teledermatoscopia com valor de Kappa de 0,698 (IC a 95% - [0,575; 0,821]). Verificou-se também boa concordância entre o exame presencial e a teledermatoscopia com o resultado anatomopatológico (padrão ouro) com valores de Kappa de 0,728 (IC a 95% - [0,608; 0,848]) e 0,656 (IC a 95% - [0,526; 0,785]), respectivamente. Constatou-se 84% de acurácia, 93% de sensibilidade e 78% de especificidade na detecção das lesões pigmentadas malignas no exame dermatoscópico a distância. Dois casos falsos negativos foram observados pela teledermatoscopia. Não houve diferença estatisticamente significante entre a sensibilidade, a especificidade e a acurácia das duas modalidades diagnósticas. Conclusões: O exame dermatoscópico a distância pela transmissão de imagens digitais e de dados clínicos proporcionou acurácia diagnóstica, sensibilidade e especificidade comparáveis ao exame convencional. A teledermatoscopia mostrou ser método eficaz e confiável na detecção das lesões malignas constituindo ferramenta importante na triagem das lesões pigmentadas. A boa concordância diagnóstica entre a avaliação via web e o exame presencial destaca a aplicabilidade da teledermatoscopia como instrumento de assistência especializada em regiões de difícil acesso ou carentes no setor de saúde. / Objectives: To evaluate agreement in diagnosis of pigmented skin lesions through presential or remote dermoscopic examination. To analyze diagnosis agreement of teledermoscopy and presential dermoscopic examination regarding pathological findings. To verify accuracy, sensibility and specificity of teledermoscopy in detection of malignant pigmented lesions. Methods: A total of 64 pigmented skin lesions were submitted to clinical and dermoscopic examinations performed presentially by two examiners experienced in dermoscopy, at the Dermoscopy Outpatient’s Clinic, Department of Dermatology, Universidade Federal de São Paulo (UNIFESP), in 2005. Two years later, the digital clinical and dermoscopic images of these cases were re-examined by the same specialists through transmission of images and clinical data via web. Agreement between presential diagnosis and telediagnosis was assessed by the Kappa (κ) statistic calculation and its respective 95% confidence interval (95% CI). Through this calculation, the agreement between presential presumptive diagnoses and telediagnoses with pathological findings was analyzed. Sensitivity, specificity and accuracy of two diagnostic modalities were obtained using the pathological examination as the gold standard. The respective 95% confidence interval (95% CI) was calculated for each diagnostic measurement. Results: Good diagnostic agreement was observed between presential examination and teledermoscopy, with a Kappa value of 0.698 (95% CI - [0.575; 0.821]). Good agreement was also found between conventional examination and teledermoscopy, and the pathological findings (gold standard) had Kappa values of 0.728 (95% CI - [0.608; 0.848]) and 0.656 (95% CI - [0.526; 0.785]), respectively. Accuracy of 84%, sensitivity of 93% and specificity of 78% were observed in detecting malignant pigmented lesions by remote dermoscopic examination. Two false negative cases were observed through teledermoscopy. There was no statistically significant difference in sensitivity, specificity and accuracy of both diagnostic methods. Conclusions: Remote dermoscopic examination by means of digital image transmission and of clinical data provided diagnostic accuracy, sensitivity and specificity comparable to conventional examination. Teledermoscopy proved to be an efficient and reliable method to detect malignant lesions, representing an important tool to screen pigmented lesions. The good diagnostic agreement between assessment via web and conventional examination highlights the applicability of teledermoscopy as a specialized assistance tool in regions that are difficult to reach or lack health resources. / TEDE / BV UNIFESP: Teses e dissertações

Over the years, exemplar-based methods have yielded significant improvements over their model-based counterparts in image synthesis applications. Notably, texture synthesis algorithms using an exemplar-based approach have shown success where traditional stochastic methods failed. As an illustrative example, I will present an exemplar-based approach that yields substantial benefits for user-guided terrain synthesis using Digital Elevation Models (DEMs). This success is realized through exploitation of structural properties of natural terrain. In addition to their proliferation in the image synthesis domain, as annotated image datasets become increasingly available, exemplar-based methods are also gaining in popularity for image analysis applications. This thesis addresses the intersection between exemplar-based analysis and the problem of content-based image retrieval (CBIR). A basic problem in CBIR is the process by which the search criteria are refined by the user through the manipulation of returned exemplars. Exemplar-based analysis is particularly well-suited to query refinement due to its interpretability and the ease with which it can be incorporated into an interactive system. I investigate this connection in the domain of Computer-Assisted Diagnosis (CAD) of dermatological images. I demonstrate that exemplar-based approaches in CBIR can be effective for diagnosing pigmented skin lesions (PSLs). I will present an exemplar-based algorithm for segmenting PSLs in dermatoscopic images. In addition, I will present a generalized representation of dermoscopic features for detection and matching. This representation not only leads to an exemplar-based PSL diagnosis scheme, but it also enables us to realize interactive region-of-interest retrieval, which includes a relevance feedback mechanism to facilitate more flexible query-by-example analysis. Finally, I will assess the benefit of this CBIR-CAD approach through both quantitative evaluations and user studies.