Development of a convolutional neural network for diagnosing osteoarthritis, trained with knee radiographs from the ELSA-Brasil Musculoskeletal.

Objective: To develop a convolutional neural network (CNN) model, trained with the Brazilian "Estudo Longitudinal de Saúde do Adulto Musculoesquelético" (ELSA-Brasil MSK, Longitudinal Study of Adult Health, Musculoskeletal) baseline radiographic examinations, for the automated classification of knee osteoarthritis. Materials and Methods: This was a cross-sectional study carried out with 5,660 baseline posteroanterior knee radiographs from the ELSA-Brasil MSK database (5,660 baseline posteroanterior knee radiographs). The examinations were interpreted by a radiologist with specific training, and the calibration was as established previously. Results: The CNN presented an area under the receiver operating characteristic curve of 0.866 (95% CI: 0.842-0.882). The model can be optimized to achieve, not simultaneously, maximum values of 0.907 for accuracy, 0.938 for sensitivity, and 0.994 for specificity. Conclusion: The proposed CNN can be used as a screening tool, reducing the total number of examinations evaluated by the radiologists of the study, and as a double-reading tool, contributing to the reduction of possible interpretation errors.

Objetivo: Desenvolver um modelo computacional - rede neural convolucional (RNC) - treinado com radiografias da linha de base do Estudo Longitudinal de Saúde do Adulto Musculoesquelético (ELSA-Brasil Musculoesquelético), para a classificação automática de osteoartrite dos joelhos. Materiais e Métodos: Trata-se de um estudo transversal abrangendo todos os exames da linha de base do ELSA-Brasil Musculoesquelético (5.660 radiografias dos joelhos em incidência posteroanterior). Os exames foram interpretados por médico radiologista com treinamento específico e calibração previamente publicada. Resultados: A RNC desenvolvida apresentou área sob a curva característica de operação do receptor de 0,866 (IC 95%: 0,842-0,882). O modelo pode ser calibrado para alcançar, não simultaneamente, valores máximos de 0,907 para acurácia, 0,938 para sensibilidade e 0,994 para especificidade. Conclusão: A RNC desenvolvida pode ser utilizada como ferramenta de triagem, reduzindo o número total de exames avaliados pelos radiologistas do estudo, e/ou como ferramenta de segunda leitura, contribuindo com a redução de possíveis erros de interpretação.

The Effects of Prenatal Diagnosis on the Interaction of the Mother-Infant Dyad: A Longitudinal Study of Prenatal Care in the First Year of Life.

Introduction: Mother-child interactions during the first years of life have a significant impact on the emotional and cognitive development of the child. In this work, we study how a prenatal diagnosis of malformation may affect maternal representations and the quality of these early interactions. To this end, we conducted a longitudinal observational study of mother-child interactions from the gestational stage until the baby completed 12 months of age. Participants and Methods: We recruited 250 pregnant women from a local university hospital. Among them, 50 mother-infant dyads participated in all stages of the study. The study group consisted of 25 pregnant women with fetuses with some structural alteration and the control group consisted of 25 pregnant women with fetuses without structural anomalies. We collected obstetric and socio demographic data and pregnancy outcomes. Anxiety and depressive state data were collected using the COVI and Raskin Scales. We video-recorded the mother-infant interactions during several stages, including when the child was a newborn and when the child was 2, 4, 6, 9, and 12 months of age. The quality of the mother infant interactions were measured using the Coding Interactive Behavior (CIB). The interactive moments recorded on video was composed of three different activities, each one lasting appoximately 3 min, which included (1) Free Interaction, where the mother was instructed to interact "as usual" without any toy, (2) Toy Interaction, where the mother and baby played with a puppv, and (3) Song Interaction, where the mother and baby interacted while the mother sang the "Happy Birthday" song. Results: In the gestational phase, there was a significant difference between the groups with respect to anxiety and depression scores, which were significantly higher for the study group. In the postnatal phase, we found significant differences between the groups with respect to CIB scales after the child completed 6 months of age: the study group presented significantly higher values of Maternal Sensitivity at 6 months of age, of Baby Involvement at 9 and 12 months of age, and of Dyadic Reciprocity at 6, 9, and 12 months of age, while the control group presented significantly higher values of Withdrawal of the Baby at 6 months of age, and of Dyadic Negative States at 6 and 9 months of age. Conclusion: The support offered by the study favored the mother-infant bond and had a positive effect on the quality of interaction during the first year of life, despite the presence of prenatal diagnosis.

Automatic measurement of pressure ulcers using Support Vector Machines and GrabCut.

BACKGROUND AND OBJECTIVE: Pressure ulcers are regions of trauma caused by a continuous pressure applied to soft tissues between a bony prominence and a hard surface. The manual monitoring of their healing evolution can be achieved by area assessment techniques that include the use of rulers and adhesive labels in direct contact with the injury, being highly inaccurate and subjective. In this paper we present a Support Vector Machine classifier in combination with a modified version of the GrabCut method for the automatic measurement of the area affected by pressure ulcers in digital images. METHODS: Three methods of region segmentation using the superpixel strategy were evaluated from which color and texture descriptors were extracted. After the superpixel classification, the GrabCut segmentation method was applied in order to delineate the region affected by the ulcer from the rest of the image. RESULTS: Experiments on a set of 105 pressure ulcer images from a public data set resulted in an average accuracy of 96%, sensitivity of 94%, specificity of 97% and precision of 94%. CONCLUSIONS: The association of support vector machines with superpixel segmentation outperformed current methods based on deep learning and may be extended to tissue classification.

Development of a convolutional neural network for diagnosing osteoarthritis, trained with knee radiographs from the ELSA-Brasil Musculoskeletal / Desenvolvimento de rede neural convolucional para o diagnóstico radiográfico de osteoartrite dos joelhos no ELSA-Brasil Musculoesquelético

Abstract Objective: To develop a convolutional neural network (CNN) model, trained with the Brazilian "Estudo Longitudinal de Saúde do Adulto Musculoesquelético" (ELSA-Brasil MSK, Longitudinal Study of Adult Health, Musculoskeletal) baseline radiographic examinations, for the automated classification of knee osteoarthritis. Materials and Methods: This was a cross-sectional study carried out with 5,660 baseline posteroanterior knee radiographs from the ELSA-Brasil MSK database (5,660 baseline posteroanterior knee radiographs). The examinations were interpreted by a radiologist with specific training, and the calibration was as established previously. Results: The CNN presented an area under the receiver operating characteristic curve of 0.866 (95% CI: 0.842-0.882). The model can be optimized to achieve, not simultaneously, maximum values of 0.907 for accuracy, 0.938 for sensitivity, and 0.994 for specificity. Conclusion: The proposed CNN can be used as a screening tool, reducing the total number of examinations evaluated by the radiologists of the study, and as a double-reading tool, contributing to the reduction of possible interpretation errors.

Resumo Objetivo: Desenvolver um modelo computacional - rede neural convolucional (RNC) - treinado com radiografias da linha de base do Estudo Longitudinal de Saúde do Adulto Musculoesquelético (ELSA-Brasil Musculoesquelético), para a classificação automática de osteoartrite dos joelhos. Materiais e Métodos: Trata-se de um estudo transversal abrangendo todos os exames da linha de base do ELSA-Brasil Musculoesquelético (5.660 radiografias dos joelhos em incidência posteroanterior). Os exames foram interpretados por médico radiologista com treinamento específico e calibração previamente publicada. Resultados: A RNC desenvolvida apresentou área sob a curva característica de operação do receptor de 0,866 (IC 95%: 0,842-0,882). O modelo pode ser calibrado para alcançar, não simultaneamente, valores máximos de 0,907 para acurácia, 0,938 para sensibilidade e 0,994 para especificidade. Conclusão: A RNC desenvolvida pode ser utilizada como ferramenta de triagem, reduzindo o número total de exames avaliados pelos radiologistas do estudo, e/ou como ferramenta de segunda leitura, contribuindo com a redução de possíveis erros de interpretação.