Acurácia diagnóstica e persistência de saos em crianças por meio do uso da inteligência artificial
Carregando...
Arquivos
Data
2023-08-16
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Introdução: A síndrome da apneia obstrutiva do sono (SAOS) caracterizase por episódios de obstrução das vias aéreas superiores, com repercussões sistêmicas, como alterações craniofaciais, cardiovasculares, metabólicas e neurocognitivas. O diagnóstico é confirmado pela polissonografia (PSG), disponível em poucos centros. A inteligência artificial (IA), tem sido utilizada na medicina, facilitando classificação e prognóstico de doenças, possibilitando avaliar associações não encontradas com a metodologia científica tradicional. Hipotetizamos que a IA possa auxiliar no reconhecimento de SAOS moderada-grave e/ou persistência de SAOS após a adenotonsilectomia, a partir da idade, peso, OSA-18 e sonoendoscopia. Objetivos: Avaliar a acurácia diagnóstica do conjunto de dados do OSA-18, idade, z-score peso/idade e de exame de sonoendoscopia para o diagnóstico de SAOS moderada-grave, e de persistência de SAOS em crianças, por meio de IA. Metodologia: Foram incluídos os dados do OSA-18, idade, escore-Z peso/idade, PSG e sonoendoscopia de 95 crianças de 4 a 9 anos, que realizaram PSG para diagnóstico e tratamento de SAOS no HCFMB no período de 2015 a 2021. Crianças com dados incompletos, neuropatas ou sindrômicas foram excluídas. Utilizando IA, com prendizado de máquina, foi avaliada a acurácia diagnóstica para SAOS normal-leve e moderadagrave em duas etapas, sendo: 1. a partir de dados do OSA-18, idade, peso e de sonoendoscopia pré-operatórios; 2. pelo conjunto dos mesmos dados, porém após adenotonsilectomia, avaliando a acurácia para a persistência de SAOS. Resultados: Os resultados mostraram elevada acurácia (93%) para o diagnóstico pré-operatório de Saos, quando utilizados dados do OSA-18, idade e z-score peso/idade, refletindo elevada sensibilidade, 100% em SAOS normal-leve e 88% em Saos moderado-grave, além de elevada especificidade, 88% e 100%, respectivamente. Já os dados de sono-endoscopia para diagnóstico mostraram redução de acurácia, inviabilizando sua execução. No entanto, para avaliar a persistência de Saos em crianças, a sono-endoscopia mostrou-se melhor que o questionário OSA-18, levando a uma acurácia de 79%, ante 75% combinando-se OSA-18, idade e peso. Conclusão: a utilização de IA com aprendizado de máquina mostrou-se eficaz no diagnóstico e seguimento de Saos, podendo contribuir para a substituição ou melhor indicação de PSG.
Introduction: Obstructive sleep apnea syndrome (OSA) is characterized by episodes of upper airway obstruction, with systemic repercussions, such as craniofacial, cardiovascular, metabolic and neurocognitive alterations. The diagnosis is confirmed by polysomnography (PSG), available in few centers. Artificial intelligence (AI) has been used in medicine, facilitating the classification and prognosis of diseases, making it possible to assess associations not found with traditional scientific methodology. We hypothesize that AI can help in the recognition of moderate-severe OSA and/or persistence of OSA after adenotonsillectomy, based on age, weight, OSA-18 and sleep endoscopy. Objectives: To evaluate the diagnostic accuracy of the OSA-18 dataset, age, weight/age z-score and sonoendoscopic examination for the diagnosis of moderate-severe OSA, and persistence of OSA in children, through AI. Methodology: Data from the OSA-18, age, weight/age Z-score, PSG and sleep endoscopy of 95 children aged 4 to 9 years who underwent PSG for diagnosis and treatment of OSA at the HCFMB from 2015 to 2021 were included. Children with incomplete data, neuropathic or syndromic were excluded. Using AI, with machine learning, the diagnostic accuracy for normal-mild and moderatesevere OSAS was evaluated in two stages, as follows: 1. from OSA-18 data, age, weight and preoperative sleep endoscopy ; 2. by the set of the same data, but after adenotonsillectomy, evaluating the accuracy for the persistence of OSAS. Results: The results showed high accuracy (93%) for the preoperative diagnosis of OSA, when data from the OSA-18, age and weight/age z-score were used, reflecting high sensitivity, 100% in normalmild OSA and 88% in moderate-severe OSA, in addition to high specificity, 88% and 100%, respectively. On the other hand, sleep-endoscopy data for diagnosis showed a reduction in accuracy, making its execution unfeasible. However, to assess the persistence of OSA in children, sleep endoscopy proved to be better than the OSA-18 questionnaire, leading to an accuracy of 79%, against 75% combining OSA-18, age and weight . Conclusion: the use of AI with machine learning proved to be effective in the diagnosis and follow-up of Saos, and may contribute to the replacement or better indication of PSG.
Introduction: Obstructive sleep apnea syndrome (OSA) is characterized by episodes of upper airway obstruction, with systemic repercussions, such as craniofacial, cardiovascular, metabolic and neurocognitive alterations. The diagnosis is confirmed by polysomnography (PSG), available in few centers. Artificial intelligence (AI) has been used in medicine, facilitating the classification and prognosis of diseases, making it possible to assess associations not found with traditional scientific methodology. We hypothesize that AI can help in the recognition of moderate-severe OSA and/or persistence of OSA after adenotonsillectomy, based on age, weight, OSA-18 and sleep endoscopy. Objectives: To evaluate the diagnostic accuracy of the OSA-18 dataset, age, weight/age z-score and sonoendoscopic examination for the diagnosis of moderate-severe OSA, and persistence of OSA in children, through AI. Methodology: Data from the OSA-18, age, weight/age Z-score, PSG and sleep endoscopy of 95 children aged 4 to 9 years who underwent PSG for diagnosis and treatment of OSA at the HCFMB from 2015 to 2021 were included. Children with incomplete data, neuropathic or syndromic were excluded. Using AI, with machine learning, the diagnostic accuracy for normal-mild and moderatesevere OSAS was evaluated in two stages, as follows: 1. from OSA-18 data, age, weight and preoperative sleep endoscopy ; 2. by the set of the same data, but after adenotonsillectomy, evaluating the accuracy for the persistence of OSAS. Results: The results showed high accuracy (93%) for the preoperative diagnosis of OSA, when data from the OSA-18, age and weight/age z-score were used, reflecting high sensitivity, 100% in normalmild OSA and 88% in moderate-severe OSA, in addition to high specificity, 88% and 100%, respectively. On the other hand, sleep-endoscopy data for diagnosis showed a reduction in accuracy, making its execution unfeasible. However, to assess the persistence of OSA in children, sleep endoscopy proved to be better than the OSA-18 questionnaire, leading to an accuracy of 79%, against 75% combining OSA-18, age and weight . Conclusion: the use of AI with machine learning proved to be effective in the diagnosis and follow-up of Saos, and may contribute to the replacement or better indication of PSG.
Descrição
Palavras-chave
SAOS, Sonoendoscopia, Inteligência artificial, Adenotonsilectomia, Crianças, OSA-18, Sleep endoscopy, Artificial intelligence, Adenotonsillectomy, Children