Desenvolvimento de aplicativo para mensuração da atividade de desgaste dentário e monitoramento de bruxismo do sono
Carregando...
Data
2024-03-21
Autores
Tersi, Patricia Danuze Patricio [UNIFESP]
Orientadores
Nakano, Luis Carlos Uta [UNIFESP]
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Objetivo: Apresentar o desenvolvimento baseado em evidências dos resultados preliminares de um aplicativo para smartphone (e-Brux) que possibilita a abordagem multifatorial do bruxismo do sono e a mensuração (em pixels) da atividade de desgaste dentário. Métodos: A primeira fase do projeto consistiu no desenvolvimento do aplicativo e dos algoritmos. Para a validação interna foram analisadas 2.070 mensurações do aplicativo e-Brux. As análises de acurácia, compararam-nas quanto ao método de mensuração (n=45 e-Brux e n=45 ImageJ) e quanto ao método de aquisição das imagens (n=90 método manual e n=90 hands-free). As análises reprodutibilidade analisaram-nas em dois níveis fatoriais de estudo: momento de aquisição das imagens (n=900 T 1 e n=900 T2) e resolução da câmera (n=600 para câmera de 13 MP, n=600 para 24 MP, n=600 para 48 MP). As amostras foram analisadas por meio dos testes ANOVA, Coeficiente de Correlação Intraclasse (CCI) e Bland-Altman (IC95%, p>0,05). Resultados: o tempo de processamento da mensuração das áreas de atividade de desgaste registrado sobre as placas foi 375 vezes mais rápido que o método convencional. As análises de Bland-Altman revelaram concordância entre os métodos de mensuração (e-Brux e ImageJ) e entre os métodos de aquisição de imagem (manual e hands free) bem como a estabilidade das medidas entre os dois momentos de coleta. Os CCIs indicaram excelentes níveis de concordância entre os momentos de coleta T1 e T2 (CCI=0,8559) e entre os modos de aquisição manual e hands-free (CCI >0,99, IC = 95%). As ANOVAs sugeriram a ausência de diferenças estatisticamente significantes entre os momentos de coleta, as resoluções dos smartphones, entre as aferições feitas pelo e-Brux e pelo padrão de referência ImageJ e entre os modos de aquisição hands free e manual. Conclusão: O aplicativo foi concebido com sucesso. Consideradas as limitações técnicas, as aferições do aplicativo e-Brux demonstraram adequadas responsividade, reprodutibilidade, estabilidade e acurácia segundo os dados preliminares. O inovador aplicativo realizou as mensurações em um tempo notavelmente menor, podendo ser uma promissora ferramenta diagnóstica com aplicabilidade clínica e acadêmica para monitoramento do bruxismo do sono e mensuração de padrões de desgaste dentário.
Objective: To present the evidence-based development and preliminary results of a smartphone application (e-Brux) that enables sleep bruxism multifactorial approach and the measurement (in pixels) of tooth wear activity. Methods: the first phase of the project was the application and algorithms development. For internal validation, 2,070 measurements from the e-Brux application were analyzed. The accuracy analyses compared the measurement method (n=45 e-Brux and n=45 ImageJ) and the image acquisition method (n=90 manual method and n=90 hands-free). The reproducibility analyses analyzed them at two factorial levels of study: moment of image acquisition (n=900 T 1 and n=900 T2) and camera resolution (n=600 for 13 MP camera, n=600 for 24 MP, n=600 for 48 MP). The samples were analyzed using the ANOVA, Intraclass Correlation Coefficient (ICC), and Bland-Altman (95% CI, p>0.05) tests. Results: The processing time for measuring areas of wear activity recorded on the plates was 375 times faster than the conventional method. Bland-Altman analysis revealed agreement between the methods e-Brux and ImageJ, manual and hands-free, and between the two data collection moments. The CCIs indicated excellent levels of correlation between collection moments T1 and T2 (CCI=0.8559) and between manual and hands-free methods (CCI >0. 99, CI = 95%). The ANOVAs suggested the absence of statistically significant differences between the collection moments, the smartphone resolutions, the measurements made by e-Brux and the ImageJ, and in hands-free and manual acquisition modes. Conclusion: The application was developed successfully. According to preliminary data and considering the limitations, the e-Brux application measurements demonstrated adequate responsiveness, reproducibility, stability, and accuracy compared to the study factors analyzed. The new mobile app concluded the measurements in a notable short. It can be a promising diagnostic tool for clinical and academic purposes, such as monitoring sleep bruxism and measuring tooth wear patterns.
Objective: To present the evidence-based development and preliminary results of a smartphone application (e-Brux) that enables sleep bruxism multifactorial approach and the measurement (in pixels) of tooth wear activity. Methods: the first phase of the project was the application and algorithms development. For internal validation, 2,070 measurements from the e-Brux application were analyzed. The accuracy analyses compared the measurement method (n=45 e-Brux and n=45 ImageJ) and the image acquisition method (n=90 manual method and n=90 hands-free). The reproducibility analyses analyzed them at two factorial levels of study: moment of image acquisition (n=900 T 1 and n=900 T2) and camera resolution (n=600 for 13 MP camera, n=600 for 24 MP, n=600 for 48 MP). The samples were analyzed using the ANOVA, Intraclass Correlation Coefficient (ICC), and Bland-Altman (95% CI, p>0.05) tests. Results: The processing time for measuring areas of wear activity recorded on the plates was 375 times faster than the conventional method. Bland-Altman analysis revealed agreement between the methods e-Brux and ImageJ, manual and hands-free, and between the two data collection moments. The CCIs indicated excellent levels of correlation between collection moments T1 and T2 (CCI=0.8559) and between manual and hands-free methods (CCI >0. 99, CI = 95%). The ANOVAs suggested the absence of statistically significant differences between the collection moments, the smartphone resolutions, the measurements made by e-Brux and the ImageJ, and in hands-free and manual acquisition modes. Conclusion: The application was developed successfully. According to preliminary data and considering the limitations, the e-Brux application measurements demonstrated adequate responsiveness, reproducibility, stability, and accuracy compared to the study factors analyzed. The new mobile app concluded the measurements in a notable short. It can be a promising diagnostic tool for clinical and academic purposes, such as monitoring sleep bruxism and measuring tooth wear patterns.
Descrição
Citação
Tese de doutorado