INTRODUCTION: Automatic detection of blood components is an important topic in the field of hematology. Segmentation is an important step because it allows components to be grouped into common areas and processed separately. This paper proposes a method for the automatic segmentation and classification of blood components in microscopic images using a general and automatic fuzzy approach. METHODS: During pre-processing, the supports of the fuzzy sets are automatically calculated based on the histogram peaks in the green channel of the RGB image and the Euclidean distance between the leukocyte nuclei centroids and the remaining pixels. During processing, fuzzification associates the degree of pertinence of the gray level of each pixel in the regions defined in the histogram with the proximity of the leukocyte nucleus centroid closest to the pixel. The fuzzy rules are then applied, and the image is defuzzified, resulting in the classification of four regions: leukocyte nuclei, leukocyte cytoplasm, erythrocytes and blood plasma. In post-processing, false positives are reduced and the leukocytes (including the nucleus and cytoplasm), erythrocytes and blood plasma are segmented. RESULTS: A total of 530 microscopic images of blood smears were processed, and the results were compared with the results of manual segmentation by experts and the accuracy rates of other approaches. CONCLUSION: The method demonstrated average accuracy rates of 97.31% for leukocytes, 95.39% for erythrocytes and 95.06% for blood plasma, avoiding the limitations found in the literature and contributing to the practice of the segmentation of blood components.

Objetivos: Validar um sistema de teleoftalmologia móvel que tem como objetivo fornecer triagem de urgências oftalmológicas em áreas remotas e desfavorecidas do Brasil. O sistema permite que um ou mais oftalmologistas possam examinar remotamente a condição do paciente e apresentar uma decisão que descreve a gravidade do caso. Se necessário, o paciente será encaminhado ao hospital para consulta. Métodos: Através de um celular e sua câmera (modelo Nexus One com câmera de 5 megapixel), foram coletados dados de 100 pacientes aleatoriamente selecionados no pronto socorro oftalmológico da Universidade Federal de São Paulo (UNIFESP) e enviados remotamente para um sistema online, por meio do qual um oftalmologista analisou-os e enviou um diagnóstico sobre a gravidade dos casos. Os resultados enviados foram comparados com o padrão ouro que foi fornecido pelo médico no hospital. Resultados: O diagnóstico foi fornecido por dois oftalmologistas: um no hospital (padrão outro) e outro remotamente. Comparando os resultados dos diagnósticos, foi obtido 81,94% de especificidade, 92,85% de sensibilidade, acurácia de 85% e um valor preditivo negativo de 96,72%. Também foi testado o desempenho do sistema, resultando num tempo de processamento do atendimento remoto em média de 8,6 minutos por paciente. Conclusões: Esta solução de baixo custo é a primeira da literatura que utiliza apenas o telefone celular para detectar urgências oftalmológicas. Com base nos resultados, o sistema consegue oferecer um atendimento confiável, diferenciando os casos de urgências e não urgências através da plataforma de telemedicina. Como ferramenta de triagem, o mais importante é identificar os casos de urgência (ter alta sensibilidade). Assim, os resultados obtidos demonstram que a ferramenta é robusta e traz uma possibilidade real de conduzir um estudo maior para verificar sua eficiência em áreas mais distantes e desfavorecidas.

Purposes: To validate a teleophthalmology mobile system aimed at improving and providing eye urgency screenings in remote and poor area settings in Brazil. The system enables one or more ophthalmologists to remotely examine a patient's condition and submit a decision describing the gravity of the case. If necessary, the patient can be forwarded to a hospital for further consultation. Methods: A cellphone (Nexus One model, with a 5 megapixel camera) was used to collect data and pictures from 100 randomly selected patients at the Ophthalmology Emergency Room located at the General Hospital of the Federal University of São Paulo (UNIFESP). Data was then sent remotely to an online recording system to be reviewed by an ophthalmologist who provided feedback regarding the state of ocular urgency. Results were then compared to the gold standard diagnosis provided at the hospital. Results: The diagnosis of urgency was given by two ophthalmologists: one in the hospital (gold standard) and one remotely. When we compared both diagnoses we obtained results of 81.94% specificity, 92.85% sensitivity, and 85% accuracy, with a negative predictive value of 96.72%. This work also included a processing time analysis, resulting in an average time of 8.6 min per patient for remote consultations. Conclusions: This study is the first that has used only a cellphone for diagnosing the urgency of ocular cases. Based on our results, the system can provide a reliable distinction between urgent and non-urgent situations and can offer a viable alternative for the servicing of underprivileged areas. In screening techniques, the most important outcome is to identify urgent cases with a high level of sensitivity and predictive negative value. Thus, our results demonstrate that this tool is robust and we suggest that a major study aimed to verify its efficiency in resource-poor areas should be initiated.