| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 3.27 MB | Adobe PDF |
Advisor(s)
Abstract(s)
Objetivo: Explorar o desempenho, no que respeita a artefactos, qualidade de imagem, avaliação de patologia/estruturas anatómicas e da relação sinal-ruído das sequências 3D aceleradas com CAIPIRINHA e com CS, comparativamente ao protocolo standard 2D, para o estudo e diagnóstico de patologias do joelho, estudadas por ressonância magnética, tendo em conta o fator de aceleração e, consequentemente o tempo de aquisição. Métodos: Após aprovação das entidades competentes e preenchimento do consentimento informado, foi efetuada a aquisição de um protocolo standard 2D, e das duas sequências 3D com técnicas de aceleração a 36 joelhos num equipamento de 3T. Após anonimização dos dados, dois Médicos Radiologistas da área procederam à avaliação das imagens através de uma escala de Likert. Foram desenhadas ROI’s em certas regiões anatómicas que permitiram o estudo da relação sinal-ruído entre as diferentes sequências de imagem para cada região. Resultados: A sequência 2D apresentou uma avaliação médica superior, seguida da sequência 3D com técnica CAIPIRINHA, sendo a sequência 3D com técnica CS a que apresentou uma classificação menor, verificando-se que as duas primeiras sequências são as que se assemelham na maioria dos parâmetros avaliados na imagem. Estatisticamente, os níveis de concordância de kappa foram maioritariamente baixos, no entanto verificaram-se as percentagens de concordância de 44.4 a 86.2% para a sequência 2D, 33.4 a 55.6% para a sequência 3D com técnica CAIPIRINHA e 30.6 a 47.3% para a sequência 3D com técnica CS. Para todas as ROI’s, à exceção da ROI 2 (osso), o valor médio da RSR é mais baixo na sequência 2D, e mais elevado na sequência 3D com técnica CS, existindo diferenças estatisticamente significativas entre as combinações de todos os pares de sequências à exceção das ROI’s 1 (cartilagem) e 2 (osso) das sequências 3D com técnica CAIPIRINHA e 3D com técnica CS. Conclusão: A implementação e inclusão na prática clínica diária de uma sequência 3D associada a uma técnica de aceleração de imagem, em substituição de um protocolo efetuado apenas com sequências 2D, apresenta qualidade de imagem suficiente para o diagnóstico de patologias do joelho. Verificou-se que a sequência 3D mais semelhante à sequência 2D, e com uma menor degradação da imagem de acordo com o fator de aceleração previamente estabelecido, em termos de artefactos e estudo de patologias e estruturas anatómicas foi a sequência 3D com técnica CAIPIRINHA, com uma redução do tempo total de exame de 48.1%.
ABSTRACT - Objective: Explore the magnetic resonance performance, artifacts, image quality, evaluation of pathology/anatomical structures, and signal to noise ratio of the 3D CAIPIRINHA technique and 3D CS technique, compared to the standard 2D protocol, for the study and diagnosis of knee pathologies, taking into account the acceleration factor and, consequently, the acquisition time. Methods: After approval from the competent authorities and informed consent, a standard 2D protocol was acquired, as well as the 3D CAIPIRINHA technique and 3D CS technique, on 36 knee of participants with a medical prescription to perform an MRI of the knee, in a 3T equipment. Following data anonymization, two Musculoskeletal Radiologists evaluated the images using a Likert scale. ROIs were drawn in certain anatomical regions to study the signal-to-noise ratio between the different image sequences for each region. Results: The 2D sequence was evaluated with a superior scale, followed by the 3D CAIPIRINHA technique, and the 3D CS technique being the one with the lowest classification, verifying that the first two sequences are the ones that are similar in most of the parameters evaluated in the image. Statistically, kappa’s of the agreement were mostly low, however, there were agreement percentages from 44.4 to 86.2% for the 2D sequence, 33.4 to 55.6% for the 3D CAIPIRINHA technique, and 30.6 to 47.3% for the 3D CS technique. For all ROIs, except for ROI 2 (bone), the mean of RSR is lower in the 2D sequence, and higher in the 3D CS technique, with statistically significant differences between the combinations of all pairs of sequences except for ROI 1 (cartilage) and 2 (bone) of the 3D CAIPIRINHA technique and 3D CS technique. Conclusion: The inclusion in daily clinical practice of a 3D sequence with an acceleration factor, replacing a protocol performed only with 2D sequences, presents sufficient image quality for the diagnosis of knee pathologies. It was found that the 3D sequence was most similar to the 2D sequence, and with less image degradation according to the previously established acceleration factor, in terms of artifacts and study of pathologies and anatomical structures was the 3D CAIPIRINHA technique, with a reduction of the total examination time of 48.1%.
ABSTRACT - Objective: Explore the magnetic resonance performance, artifacts, image quality, evaluation of pathology/anatomical structures, and signal to noise ratio of the 3D CAIPIRINHA technique and 3D CS technique, compared to the standard 2D protocol, for the study and diagnosis of knee pathologies, taking into account the acceleration factor and, consequently, the acquisition time. Methods: After approval from the competent authorities and informed consent, a standard 2D protocol was acquired, as well as the 3D CAIPIRINHA technique and 3D CS technique, on 36 knee of participants with a medical prescription to perform an MRI of the knee, in a 3T equipment. Following data anonymization, two Musculoskeletal Radiologists evaluated the images using a Likert scale. ROIs were drawn in certain anatomical regions to study the signal-to-noise ratio between the different image sequences for each region. Results: The 2D sequence was evaluated with a superior scale, followed by the 3D CAIPIRINHA technique, and the 3D CS technique being the one with the lowest classification, verifying that the first two sequences are the ones that are similar in most of the parameters evaluated in the image. Statistically, kappa’s of the agreement were mostly low, however, there were agreement percentages from 44.4 to 86.2% for the 2D sequence, 33.4 to 55.6% for the 3D CAIPIRINHA technique, and 30.6 to 47.3% for the 3D CS technique. For all ROIs, except for ROI 2 (bone), the mean of RSR is lower in the 2D sequence, and higher in the 3D CS technique, with statistically significant differences between the combinations of all pairs of sequences except for ROI 1 (cartilage) and 2 (bone) of the 3D CAIPIRINHA technique and 3D CS technique. Conclusion: The inclusion in daily clinical practice of a 3D sequence with an acceleration factor, replacing a protocol performed only with 2D sequences, presents sufficient image quality for the diagnosis of knee pathologies. It was found that the 3D sequence was most similar to the 2D sequence, and with less image degradation according to the previously established acceleration factor, in terms of artifacts and study of pathologies and anatomical structures was the 3D CAIPIRINHA technique, with a reduction of the total examination time of 48.1%.
Description
Mestrado em Radiações Aplicadas às Tecnologias da Saúde. - Área de especialização: Imagem por Ressonância Magnética
Keywords
Ressonância magnética Qualidade de imagem Joelho CAIPIRINHA Compressed sensing 3D Magnetic resonance Image quality Knee
Citation
Guerreiro TC. Comparação dos protocolos standard vs. 3D CAIPIRINHA TSE vs. 3D compressed sensing no estudo do joelho por RM [dissertation]. Lisboa: Escola Superior de Tecnologia da Saúde de Lisboa/Instituto Politécnico de Lisboa; 2021.
Publisher
Instituto Politécnico de Lisboa, Escola Superior de Tecnologia da Saúde de Lisboa