Relação entre as medidas radiográficas e da ressonância magnética na avaliação do bunionette
Carregando...
Data
2023-12-27
Autores
Garcia, Guilherme Thomé [UNIFESP]
Orientadores
Yamada, André Fukunishi [UNIFESP]
Tipo
Dissertação de mestrado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Introdução: O bunionette é uma deformidade caracterizada pela proeminência óssea na face lateral do antepé, na altura da cabeça do quinto metatarso, que pode ou não ser sintomática. A avaliação por imagem consiste na realização de radiografia com carga do pé (RC), e diversos ângulos foram descritos para caracterização das deformidades. No entanto, não existem relatos sobre a aplicabilidade dessas medidas nos exames de ressonância magnética do antepé (RM). As hipóteses são que as medidas obtidas por RC e RM se relacionam e são reprodutíveis. Objetivos: Avaliar a reprodutibilidade interobservador das medidas dos ângulos metatarsofalângico do 5° dedo (AMF), intermetatársico do 4° e 5° dedos (AIM) e ângulo de desvio lateral (ADL), obtidas em RC e RM; determinar a acurácia das medidas obtidas através da RM na caracterização de exames normais e alterados, usando a RC como método de referência (padrão-ouro). Métodos: Foram selecionados retrospectivamente 50 conjuntos de exames (RC e sua respectiva RM) de 33 pacientes. Os exames foram realizadas conforme a rotina e protocolo institucionais. Dois avaliadores com 2 anos de experiência em radiologia musculoesquelética obtiveram as medidas dos ângulos AMF, AIM e ADL nas RC e RM. As leituras foram realizadas nas incidências dorsoplantar das RC e sequências axiais T1 puro das RMs. Os exames foram divididos em dois grupos (normais e alterados/bunionette) conforme a RC (padrão-ouro). O coeficiente de correlação e concordância de Lin foi usado para avaliação da reprodutibilidade interobservador e para análise da correlação entre os métodos de imagem. Modelos de regressão linear foram usados para estimar a diferença média das medidas obtidas entre os métodos. Curvas da característica operacional do receptor foram geradas para definição dos melhores pontos de corte na diferenciação entre os grupos pela RM e cálculo da acurácia das medidas obtidas na RM. Resultados: A concordância interobservador para o ângulo AMF foi substancial nas RCs e na RMs (CCC = 0,98 e 0,99, respectivamente), moderada nas RCs e RMs para o ângulo AIM (CCC = 0,96 e 0,90, respectivamente) e pobre para ADL (CCC = 0,83 e 0,69, respectivamente). Não se observou forte correlação entre as medidas obtidas nas RMs com as RCs nos três ângulos avaliados. Na regressão linear observou-se diferença média entre as RCs e RMs de 4,3° para o ângulo AMF (IC 95% 2,10 - 6,60; p<0,001), 0,6° para AIM (IC 95% 0,04 - 1,10; p<0,036) e -1,53° para ADL (IC 95% -2,7; -0,38; p<0,010). As curvas ROC produziram valores de corte de 15,4° para o ângulo MFA (AUC = 94,9%; IC 95% 88,9-100%; p<0,001), 8,9° para IMA (AUC = 86.5%; IC 95% 75,3-97,9%; p<0,001) e 1,25° para ADL (AUC = 43,1%; IC 95% 6,2%-80,0%; p=0,698). Quando comparados com os exames de RC, os pontos de corte definidos nas RMs para AMF e AIM resultaram em valores de acurácia de 89,8% e 80%, respectivamente. Conclusão: As medidas AMF e AIM são aplicáveis nos exames de RM do antepé usando as mesmas técnicas descritas para as RCs do pé. Os dois ângulos mostraram-se reprodutíveis e observou-se acurácia satisfatória dos pontos de corte definidos para diferenciar exames normais e alterados. Tais medidas podem ser consideradas ferramentas adicionais na rotina do radiologista para investigação do bunionette nos exames de RM do antepé.
Introduction: Bunionette is a deformity characterized by a bony prominence on the lateral aspect of the forefoot, at the level of the fifth metatarsal head, which may or may not be symptomatic. Imaging assessment involves performing weight-bearing foot radiography (WBR), and various angles have been described to characterize the deformities. However, there are no reports on the applicability of these measurements in forefoot magnetic resonance imaging (MRI) examinations. The hypotheses are that the measurements obtained by RC and MRI are related and reproducible. Objectives: To evaluate the interobserver reproducibility of metatarsophalangeal angle of the 5th toe (MFA), intermetatarsal angle between the 4th and 5th toes (IMA), and lateral deviation angle (LDA) obtained in RC and MRI; to determine the accuracy of measurements obtained through MRI in characterizing normal and abnormal examinations, using RC as the reference method (gold standard). Methods: 50 sets of examinations (RC and their respective MRI) from 33 patients were retrospectively selected. The examinations were conducted following institutional routine and protocols. Two evaluators with 2 years of experience in musculoskeletal radiology obtained MFA, IMA, and LDA measurements in both RC and MRI. Readings were performed in dorsoplantar views of RC and pure T1 axial sequences of MRI. Examinations were divided into two groups (normal and abnormal/bunionette) based on RC (gold standard). The Lin concordance and correlation coefficient was used to assess interobserver reproducibility and to analyze the correlation between imaging methods. Linear regression models were used to estimate the mean difference in measurements obtained between the methods. Receiver operating characteristic (ROC) curves were generated to define the best cutoff points for differentiation between groups by MRI and to calculate the accuracy of measurements obtained in MRI. Results: Interobserver agreement for the MFA angle was substantial in both RC and MRI (CCC = 0.98 and 0.99, respectively), moderate in RC and MRI for the IMA angle (CCC = 0.96 and 0.90, respectively), and poor for LDA (CCC = 0.83 and 0.69, respectively). Strong correlation was not observed between measurements obtained in MRI and RC for the three evaluated angles. Linear regression revealed a mean difference between RC and MRI of 4.3° for MFA angle (95% CI 2.10 - 6.60; p<0.001), 0.6° for IMA (95% CI 0.04 - 1.10; p<0.036), and -1.53° for LDA (95% CI -2.7; -0.38; p<0.010). ROC curves produced cutoff values of 15.4° for the MFA angle (AUC = 94.9%; 95% CI 88.9-100%; p<0.001), 8.9° for IMA (AUC = 86.5%; 95% CI 75.3-97.9%; p<0.001), and 1.25° for LDA (AUC = 43.1%; 95% CI 6.2%-80.0%; p=0.698). When compared to RC examinations, the cutoff points defined in MRI for MFA and IMA resulted in accuracy values of 89.8% and 80%, respectively. Conclusion: MFA and IMA measurements are applicable in forefoot MRI examinations using the same techniques described for RC of the foot. Both angles are reproducible, and there was satisfactory accuracy of cutoff points defined to differentiate normal and abnormal examinations. These measurements can be considered additional tools in the radiologist's routine for investigating bunionette in forefoot MRI examinations.
Introduction: Bunionette is a deformity characterized by a bony prominence on the lateral aspect of the forefoot, at the level of the fifth metatarsal head, which may or may not be symptomatic. Imaging assessment involves performing weight-bearing foot radiography (WBR), and various angles have been described to characterize the deformities. However, there are no reports on the applicability of these measurements in forefoot magnetic resonance imaging (MRI) examinations. The hypotheses are that the measurements obtained by RC and MRI are related and reproducible. Objectives: To evaluate the interobserver reproducibility of metatarsophalangeal angle of the 5th toe (MFA), intermetatarsal angle between the 4th and 5th toes (IMA), and lateral deviation angle (LDA) obtained in RC and MRI; to determine the accuracy of measurements obtained through MRI in characterizing normal and abnormal examinations, using RC as the reference method (gold standard). Methods: 50 sets of examinations (RC and their respective MRI) from 33 patients were retrospectively selected. The examinations were conducted following institutional routine and protocols. Two evaluators with 2 years of experience in musculoskeletal radiology obtained MFA, IMA, and LDA measurements in both RC and MRI. Readings were performed in dorsoplantar views of RC and pure T1 axial sequences of MRI. Examinations were divided into two groups (normal and abnormal/bunionette) based on RC (gold standard). The Lin concordance and correlation coefficient was used to assess interobserver reproducibility and to analyze the correlation between imaging methods. Linear regression models were used to estimate the mean difference in measurements obtained between the methods. Receiver operating characteristic (ROC) curves were generated to define the best cutoff points for differentiation between groups by MRI and to calculate the accuracy of measurements obtained in MRI. Results: Interobserver agreement for the MFA angle was substantial in both RC and MRI (CCC = 0.98 and 0.99, respectively), moderate in RC and MRI for the IMA angle (CCC = 0.96 and 0.90, respectively), and poor for LDA (CCC = 0.83 and 0.69, respectively). Strong correlation was not observed between measurements obtained in MRI and RC for the three evaluated angles. Linear regression revealed a mean difference between RC and MRI of 4.3° for MFA angle (95% CI 2.10 - 6.60; p<0.001), 0.6° for IMA (95% CI 0.04 - 1.10; p<0.036), and -1.53° for LDA (95% CI -2.7; -0.38; p<0.010). ROC curves produced cutoff values of 15.4° for the MFA angle (AUC = 94.9%; 95% CI 88.9-100%; p<0.001), 8.9° for IMA (AUC = 86.5%; 95% CI 75.3-97.9%; p<0.001), and 1.25° for LDA (AUC = 43.1%; 95% CI 6.2%-80.0%; p=0.698). When compared to RC examinations, the cutoff points defined in MRI for MFA and IMA resulted in accuracy values of 89.8% and 80%, respectively. Conclusion: MFA and IMA measurements are applicable in forefoot MRI examinations using the same techniques described for RC of the foot. Both angles are reproducible, and there was satisfactory accuracy of cutoff points defined to differentiate normal and abnormal examinations. These measurements can be considered additional tools in the radiologist's routine for investigating bunionette in forefoot MRI examinations.
Descrição
Citação
GARCIA, Guilherme Thomé. Relação entre as medidas radiográficas e da ressonância magnética na avaliação do bunionette. 2023. 58 f. Dissertação (Mestrado em Radiologia Clínica) - Escola Paulista de Medicina, Universidade Federal de São Paulo (UNIFESP). São Paulo, 2023.