Detecting face presentation attacks in mobile devices with a patch-based CNN and a sensor-aware loss function.

With the widespread use of biometric authentication comes the exploitation of presentation attacks, possibly undermining the effectiveness of these technologies in real-world setups. One example takes place when an impostor, aiming at unlocking someone else's smartphone, deceives the built-in face recognition system by presenting a printed image of the user. In this work, we study the problem of automatically detecting presentation attacks against face authentication methods, considering the use-case of fast device unlocking and hardware constraints of mobile devices. To enrich the understanding of how a purely software-based method can be used to tackle the problem, we present a solely data-driven approach trained with multi-resolution patches and a multi-objective loss function crafted specifically to the problem. We provide a careful analysis that considers several user-disjoint and cross-factor protocols, highlighting some of the problems with current datasets and approaches. Such analysis, besides demonstrating the competitive results yielded by the proposed method, provides a better conceptual understanding of the problem. To further enhance efficacy and discriminability, we propose a method that leverages the available gallery of user data in the device and adapts the method decision-making process to the user's and the device's own characteristics. Finally, we introduce a new presentation-attack dataset tailored to the mobile-device setup, with real-world variations in lighting, including outdoors and low-light sessions, in contrast to existing public datasets.

A data-driven approach to referable diabetic retinopathy detection.

Prior art on automated screening of diabetic retinopathy and direct referral decision shows promising performance; yet most methods build upon complex hand-crafted features whose performance often fails to generalize. OBJECTIVE: We investigate data-driven approaches that extract powerful abstract representations directly from retinal images to provide a reliable referable diabetic retinopathy detector. METHODS: We gradually build the solution based on convolutional neural networks, adding data augmentation, multi-resolution training, robust feature-extraction augmentation, and a patient-basis analysis, testing the effectiveness of each improvement. RESULTS: The proposed method achieved an area under the ROC curve of 98.2% (95% CI: 97.4-98.9%) under a strict cross-dataset protocol designed to test the ability to generalize - training on the Kaggle competition dataset and testing using the Messidor-2 dataset. With a 5 × 2-fold cross-validation protocol, similar results are achieved for Messidor-2 and DR2 datasets, reducing the classification error by over 44% when compared to most published studies in existing literature. CONCLUSION: Additional boost strategies can improve performance substantially, but it is important to evaluate whether the additional (computation- and implementation-) complexity of each improvement is worth its benefits. We also corroborate that novel families of data-driven methods are the state of the art for diabetic retinopathy screening. SIGNIFICANCE: By learning powerful discriminative patterns directly from available training retinal images, it is possible to perform referral diagnostics without detecting individual lesions.

Beyond Lesion-Based Diabetic Retinopathy: A Direct Approach for Referral.

Diabetic retinopathy (DR) is the leading cause of blindness in adults, but can be managed if detected early. Automated DR screening helps by indicating which patients should be referred to the doctor. However, current techniques of automated screening still depend too much on the detection of individual lesions. In this study, we bypass lesion detection, and directly train a classifier for DR referral. Additional novelties are the use of state-of-the-art mid-level features for the retinal images: BossaNova and Fisher Vector. Those features extend the classical Bags of Visual Words and greatly improve the accuracy of complex classification tasks. The proposed technique for direct referral is promising, achieving an area under the curve of 96.4%, thus, reducing the classification error by almost 40% over the current state of the art, held by lesion-based techniques.

Automated multi-lesion detection for referable diabetic retinopathy in indigenous health care.

Diabetic Retinopathy (DR) is a complication of diabetes mellitus that affects more than one-quarter of the population with diabetes, and can lead to blindness if not discovered in time. An automated screening enables the identification of patients who need further medical attention. This study aimed to classify retinal images of Aboriginal and Torres Strait Islander peoples utilizing an automated computer-based multi-lesion eye screening program for diabetic retinopathy. The multi-lesion classifier was trained on 1,014 images from the São Paulo Eye Hospital and tested on retinal images containing no DR-related lesion, single lesions, or multiple types of lesions from the Inala Aboriginal and Torres Strait Islander health care centre. The automated multi-lesion classifier has the potential to enhance the efficiency of clinical practice delivering diabetic retinopathy screening. Our program does not necessitate image samples for training from any specific ethnic group or population being assessed and is independent of image pre- or post-processing to identify retinal lesions. In this Aboriginal and Torres Strait Islander population, the program achieved 100% sensitivity and 88.9% specificity in identifying bright lesions, while detection of red lesions achieved a sensitivity of 67% and specificity of 95%. When both bright and red lesions were present, 100% sensitivity with 88.9% specificity was obtained. All results obtained with this automated screening program meet WHO standards for diabetic retinopathy screening.

Peer-selected "best papers"-are they really that "good"?

BACKGROUND: Peer evaluation is the cornerstone of science evaluation. In this paper, we analyze whether or not a form of peer evaluation, the pre-publication selection of the best papers in Computer Science (CS) conferences, is better than random, when considering future citations received by the papers. METHODS: Considering 12 conferences (for several years), we collected the citation counts from Scopus for both the best papers and the non-best papers. For a different set of 17 conferences, we collected the data from Google Scholar. For each data set, we computed the proportion of cases whereby the best paper has more citations. We also compare this proportion for years before 2010 and after to evaluate if there is a propaganda effect. Finally, we count the proportion of best papers that are in the top 10% and 20% most cited for each conference instance. RESULTS: The probability that a best paper will receive more citations than a non best paper is 0.72 (95% CI = 0.66, 0.77) for the Scopus data, and 0.78 (95% CI = 0.74, 0.81) for the Scholar data. There are no significant changes in the probabilities for different years. Also, 51% of the best papers are among the top 10% most cited papers in each conference/year, and 64% of them are among the top 20% most cited. DISCUSSION: There is strong evidence that the selection of best papers in Computer Science conferences is better than a random selection, and that a significant number of the best papers are among the top cited papers in the conference.

Advancing bag-of-visual-words representations for lesion classification in retinal images.

Diabetic Retinopathy (DR) is a complication of diabetes that can lead to blindness if not readily discovered. Automated screening algorithms have the potential to improve identification of patients who need further medical attention. However, the identification of lesions must be accurate to be useful for clinical application. The bag-of-visual-words (BoVW) algorithm employs a maximum-margin classifier in a flexible framework that is able to detect the most common DR-related lesions such as microaneurysms, cotton-wool spots and hard exudates. BoVW allows to bypass the need for pre- and post-processing of the retinographic images, as well as the need of specific ad hoc techniques for identification of each type of lesion. An extensive evaluation of the BoVW model, using three large retinograph datasets (DR1, DR2 and Messidor) with different resolution and collected by different healthcare personnel, was performed. The results demonstrate that the BoVW classification approach can identify different lesions within an image without having to utilize different algorithms for each lesion reducing processing time and providing a more flexible diagnostic system. Our BoVW scheme is based on sparse low-level feature detection with a Speeded-Up Robust Features (SURF) local descriptor, and mid-level features based on semi-soft coding with max pooling. The best BoVW representation for retinal image classification was an area under the receiver operating characteristic curve (AUC-ROC) of 97.8% (exudates) and 93.5% (red lesions), applying a cross-dataset validation protocol. To assess the accuracy for detecting cases that require referral within one year, the sparse extraction technique associated with semi-soft coding and max pooling obtained an AUC of 94.2 ± 2.0%, outperforming current methods. Those results indicate that, for retinal image classification tasks in clinical practice, BoVW is equal and, in some instances, surpasses results obtained using dense detection (widely believed to be the best choice in many vision problems) for the low-level descriptors.

Automatic Diabetic Retinopathy detection using BossaNova representation.

The biomedical community has shown a continued interest in automated detection of Diabetic Retinopathy (DR), with new imaging techniques, evolving diagnostic criteria, and advancing computing methods. Existing state of the art for detecting DR-related lesions tends to emphasize different, specific approaches for each type of lesion. However, recent research has aimed at general frameworks adaptable for large classes of lesions. In this paper, we follow this latter trend by exploring a very flexible framework, based upon two-tiered feature extraction (low-level and mid-level) from images and Support Vector Machines. The main contribution of this work is the evaluation of BossaNova, a recent and powerful mid-level image characterization technique, which we contrast with previous art based upon classical Bag of Visual Words (BoVW). The new technique using BossaNova achieves a detection performance (measured by area under the curve - AUC) of 96.4% for hard exudates, and 93.5% for red lesions using a cross-dataset training/testing protocol.

Assessing the Need for Referral in Automatic Diabetic Retinopathy Detection.

Emerging technologies in health care aim at reducing unnecessary visits to medical specialists, minimizing overall cost of treatment and optimizing the number of patients seen by each doctor. This paper explores image recognition for the screening of diabetic retinopathy, a complication of diabetes that can lead to blindness if not discovered in its initial stages. Many previous reports on DR imaging focus on the segmentation of the retinal image, on quality assessment, and on the analysis of presence of DR-related lesions. Although this study has advanced the detection of individual DR lesions from retinal images, the simple presence of any lesion is not enough to decide on the need for referral of a patient. Deciding if a patient should be referred to a doctor is an essential requirement for the deployment of an automated screening tool for rural and remote communities. We introduce an algorithm to make that decision based on the fusion of results by metaclassification. The input of the metaclassifier is the output of several lesion detectors, creating a powerful high-level feature representation for the retinal images. We explore alternatives for the bag-of-visual-words (BoVW)-based lesion detectors, which critically depends on the choices of coding and pooling the low-level local descriptors. The final classification approach achieved an area under the curve of 93.4% using SOFT-MAX BoVW (soft-assignment coding/max pooling), without the need of normalizing the high-level feature vector of scores.

Mensuração do tempo dos registros manual e eletrônico da Sistematização da Assistência de Enfermagem em Unidade de Terapia Intensiva / Measuring the time of manual and electronic records of the Systematization of Nursing Care in theIntensive Care Unit / Midiendo el tiempo de los registros manuales y electrónicos de la Sistematización de los Cuidados deEnfermería en la Unidad de Cuidados Intensivos

Objetivos: Mensurar o tempo do registro manual e eletrônico da Sistematização da Assistência de Enfermagem. Métodos: Em estudo experimental antes e depois, o registro eletrônico foi mensurado mediante o uso de sistema informatizado e o registro manual em impressos. As tarefas foram previamente sorteadas e realizadas por oito enfermeiros. A mensuração ocorreu durante quatro horas nos períodos manhã, tarde e noite, em seis fases idênticas, com intervalos de um mês. Os dados foram analisados com Wilcoxon, p<0,05. Resultados: O registro eletrônico do "Registro do Exame Físico" e "Diagnóstico de Enfermagem" levou mais tempo de execução que o manual, p<0,05 e o registro eletrônico da "Prescrição de Enfermagem" e "Evolução de Enfermagem" levou menos tempo de realização, p<0,04 e p<0,0002. Conclusões: O aumento do tempo no registro eletrônico ocorreu devido o sistema requerer informações complexas e a diminuição do tempo ocorreu em decorrência da facilidade de acesso às informações do sistema.

Objectives:To measure the time of manual and electronic record of the Nursing Care System. Methods: In an experimental study before and after the electronic record was measured by using computerized and manual record in print. The tasks were previously drawn and performed by eight nurses. The measurement took place during four hours in the morning period, afternoon and evening, in six equal steps at intervals of a month. Data were analyzed with Wilcoxon, p <0.05. Results: The electronic record of the "Record of Physical Examination" and "Nursing Diagnosis" execution took longer than the manual, p <0.05 and the electronic registration of "Prescription of Nursing" and "Evolution of Nursing" took less time for achievement, p <0.04 and p <0.0002. Conclusions: the increased time was due in the electronic registration system requiring complex information and decreasing the time was due to the ease of access to system information.

Objetivos: medir el tiempo de registro manual y electrónico del Sistema de Cuidados de Enfermería. Método: En un estudio experimental antes y después del registro electrónico se midió mediante el uso de registros automatizados y manuales en formato impreso. Las tareas fueron elaborados previamente y realizado por ocho enfermeras. La medición se llevó a cabo durante cuatro horas en el período de la mañana, tarde y noche, en seis etapas iguales a intervalos de un mes. Los datos fueron analizados con Wilcoxon, p <0,05. Resultados: El registro electrónico de la "Acta de Examen Físico" y la ejecución de "Diagnóstico de Enfermería" tomó más tiempo que el manual, p <0,05 y el registro electrónico de la "Prescripción de la Enfermería" y "Evolución de la Enfermería" tomó menos tiempo para el logro, p <0,04 yp <0,0002. Conclusiones: el tiempo de aumento se debió en el sistema de registro electrónico que requieren información compleja y disminuyendo el tiempo se debe a la facilidad de acceso a la información del sistema.

Machine learning and pattern classification in identification of indigenous retinal pathology.

Diabetic retinopathy (DR) is a complication of diabetes, which if untreated leads to blindness. DR early diagnosis and treatment improve outcomes. Automated assessment of single lesions associated with DR has been investigated for sometime. To improve on classification, especially across different ethnic groups, we present an approach using points-of-interest and visual dictionary that contains important features required to identify retinal pathology. Variation in images of the human retina with respect to differences in pigmentation and presence of diverse lesions can be analyzed without the necessity of preprocessing and utilizing different training sets to account for ethnic differences for instance.

Telemedicina: contexto e desdobramento a partir do estudo de três casos no período de 2005 e 2006 / Telemedicine: context and deployment from the study of three cases in period 2005 to 2006

Este artigo apresenta o resultado da tese de doutorado do autor, cujo objetivo foi conhece o contexto de três projetos de Telemedicina, os respectivos desdobramentos para a sociedade. Para a realização da pesquisa, foi utilizado o modelo de estudo de caso, com uso de ferramentas de entrevista de campo e observações etnográficas. Os projetos estudados foram,nas cidades de Belo Horizonte, Recife e Porto Alegre. Os resultados alcançados são parciais, ou seja, eles reforçam a tese de que é preciso se estabelecer mecanismos de avaliação pré, durante e pós-implantação dos projetos de Telemedicina. constatamos que:Impacto no Acesso ? não há evidências de melhoria e ampliação do acesso aos serviços de saúde. Ao contrário, este acesso foi limitado há programação da realização de videoconferências; Impacto Econômico ? para os casos em que ocorreu o processo de segunda opinião, de fato, há ganhos para o paciente.; Impacto na Aceitação ? por parte dos usuários se deve ao fato do ineditismo, participação em videoconferência.

This article shows an summary of research realized to determine where they were implanted with the telemedicine projects and their consequences for society. To conduct the study, we used the model case study, with the use of interview tools and description of it, from ethnographic observations. A total of three projects in telemedicine, the cities of Belo Horizonte, Recife and Porto Alegre. The results are partial, that is, they reinforce the view that it is necessary to establish mechanisms for the pre, during and after implementation of telemedicine projects. We also found that: Impact on Access - did not find evidence of improving and expanding access to health services; Economic Impact - for cases that occurred in the process of second opinion, in fact, there are gains for the patient. ; impact on the acceptance - by among users is because of the novelty, participation in videoconferencing.

How do experts recognize schizophrenia: the role of the disorganization symptom.

OBJECTIVE: Research on clinical reasoning has been useful in developing expert systems. These tools are based on Artificial Intelligence techniques which assist the physician in the diagnosis of complex diseases. The development of these systems is based on a cognitive model extracted through the identification of the clinical reasoning patterns applied by experts within the clinical decision-making context. This study describes the method of knowledge acquisition for the identification of the triggering symptoms used in the reasoning of three experts for the diagnosis of schizophrenia. METHOD: Three experts on schizophrenia, from two University centers in Sao Paulo, were interviewed and asked to identify and to represent the triggering symptoms for the diagnosis of schizophrenia according to the graph methodology. RESULTS: Graph methodology showed a remarkable disagreement on how the three experts established their diagnosis of schizophrenia. They differed in their choice of triggering-symptoms for the diagnosis of schizophrenia: disorganization, blunted affect and thought disturbances. CONCLUSIONS: The results indicate substantial differences between the experts as to their diagnostic reasoning patterns, probably under the influence of different theoretical tendencies. The disorganization symptom was considered to be the more appropriate to represent the heterogeneity of schizophrenia and also, to further develop an expert system for the diagnosis of schizophrenia.

How do experts recognize schizophrenia: the role of the disorganization symptom / Como os especialistas reconhecem a esquizofrenia: o papel do sintoma desorganização

OBJETIVE: Research on clinical reasoning has been useful in developing expert systems. These tools are based on Artificial Intelligence techniques which assist the physician in the diagnosis of complex diseases. The development of these systems is based on a cognitive model extracted through the identification of the clinical reasoning patterns applied by experts within the clinical decision-making context. This study describes the method of knowledge acquisition for the identification of the triggering symptoms used in the reasoning of three experts for the diagnosis of schizophrenia. METHOD: Three experts on schizophrenia, from two University centers in Sao Paulo, were interviewed and asked to identify and to represent the triggering symptoms for the diagnosis of schizophrenia according to the graph methodology. RESULTS: Graph methodology showed a remarkable disagreement on how the three experts established their diagnosis of schizophrenia. They differed in their choice of triggering-symptoms for the diagnosis of schizophrenia: disorganization, blunted affect and thought disturbances. CONCLUSIONS: The results indicate substantial differences between the experts as to their diagnostic reasoning patterns, probably under the influence of different theoretical tendencies. The disorganization symptom was considered to be the more appropriate to represent the heterogeneity of schizophrenia and also, to further develop an expert system for the diagnosis of schizophrenia.

OBJETIVO: As pesquisas sobre o raciocínio clínico foram importantes para o surgimento de sistemas de apoio à decisão diagnóstica. Essas ferramentas são desenvolvidas por meio de técnicas de inteligência artificial e têm com objetivo principal auxiliar o médico no diagnóstico de doenças complexas. A abordagem utilizada para a construção desses sistemas constitui na formulação de um modelo baseado na identificação de padrões no raciocínio dos expertos quando de uma tomada de decisão diagnóstica. Este estudo descreve a metodologia empregada para identificar os elementos-chave utilizados no raciocínio de três expertos no processo de diagnóstico do transtorno da esquizofrenia. MÉTODO: Para explorar o raciocínio clínico foram selecionados três expertos em esquizofrenia de dois centros universitários de São Paulo. Foi utilizado o método dos grafos, por meio do qual o experto podia esquematizar a combinação de sintomas-chave que ele utilizava para identificar um diagnóstico de esquizofrenia. RESULTADOS: A partir da análise qualitativa dos grafos foi possível notar uma diferença marcante nos padrões de raciocínio diagnóstico. Essa diferença ocorreu, sobretudo, nos sintomas-chave do processo de decisão diagnóstica: desorganização, afeto embotado e distúrbio do pensamento. CONCLUSÕES: Os resultados apontam para uma diferença substancial entre os expertos quanto a um padrão de raciocínio diagnóstico provavelmente influenciado por diferentes correntes teóricas. Essas diferenças constituem um impedimento para a construção de um modelo único. O sintoma desorganização foi considerado o elemento-chave mais apropriado para representar a heterogeneidade da esquizofrenia e ser modelado para a construção de sistema de apoio à decisão diagnóstica.