Background Generalized knee tissue segmentation,such as cartilage and meniscus in magnetic resonance imaging(MRI),plays a vital role in the clinical assessment of knee osteoarthritis(OA).However,domain variability bet...Background Generalized knee tissue segmentation,such as cartilage and meniscus in magnetic resonance imaging(MRI),plays a vital role in the clinical assessment of knee osteoarthritis(OA).However,domain variability between MRI datasets poses a significant challenge for the application of robust segmentation methods in real-world clinical settings.Existing unsupervised domain adaptation(UDA)approaches,which rely on one-to-one assumptions between the source and target domains,often fail to preserve knee tissues such as cartilage and meniscus,which are critical for OA diagnosis in diverse clinical settings.Methods We propose a source-independent segmentation approach tailored for multi-domain knee MRI datasets.Our method emphasizes knee tissue regions to reduce domain gaps and label inconsistencies.By introducing a stepwise adaptation strategy,segmentation performance was refined progressively from intermediate domains to the final target domain.Pseudo-label attention mechanisms were integrated into the adaptation pipeline,enabling iterative fine-tuning of domain-specific segmentations while leveraging unidirectional generative adversarial networks to enhance tissue-specific adaptation.This iterative training process ensures the generation of reliable pseudo-labels,thereby improving segmentation accuracy in diverse clinical MRI datasets.Results We demonstrated the effectiveness of our approach on the OA initiative dataset as the source domain and self-collected,T1-weighted fast field echo(T1FFE)as the intermediate domain and three-dimensional fast spin echo(3D FSE)as the final target domain.Our method achieved an average dice scores of 0.8701 and 0.7990 for source and target domains,respectively,surpassing the typical UDA methods explored in our experiments.Conclusion The experiments conducted on clinical MRI data,spanning OA severity from healthy knees to KL Grades 1–4,validated the effectiveness of the proposed domain adaptation method in precise segmentation of the cartilage and meniscus.展开更多
Background The association of long-term bisphosphonate treatment for osteoporosis and related safety problems such as atypical fractures were not clearly defined. This study was to evaluate the structural, densitometr...Background The association of long-term bisphosphonate treatment for osteoporosis and related safety problems such as atypical fractures were not clearly defined. This study was to evaluate the structural, densitometric and biomechanical properties of the prolonged bisphosphonate-loaded bones. Methods Bone mineral density （BMD） at hip and femoral midshaft, bone cross-sectional area, moment of inertia of both femurs, bone formation and resorption biochemical markers were compared between 28 elderly with at least 4 years of bisphosphonate treatment from 2002 through 2006 and age-matched group of 37 elderly. Results The total hip BMD and t-score were found not different between two groups. However, bisphosphonate treated patients were found to have significantly lower bone mineral content in the femoral shaft （P 〈0.05）; morphological study showed lower cross-sectional area in subtrochanteric and mid-diaphyseal region and thus significantly lower moment of inertia （P 〈0.01）. High resolution-peripheral quantitative computed tomography showed significantly decreased trabeculardensity, bone volume ratio, trabecular number but increased trabecular spacing in tibia and distal radius. Finite element analysis further confirmed significantly lower stiffness and failure load in tibia. Biochemical studies also showed lower bone resorption and severely suppressed bone formation activity （P 〈0.001）. Conclusions The unchanged total hip BMD between two groups confirmed the beneficial effects of bisphosphonate on trabecular bone, thus preventing osteoporotic fractures at large in previous studies. However, the inferior structural, densitometric and biomechanical properties at cortical bones, especially femur midshaft, need a special attention to look into the association between long-term bisphosphonate intake and the occurrence of stress fractures. When patients taking bisphosphonate complain of proximal thigh pain or discomfort, plain X-ray film can be the first line screening. All patients prescribed with bisphosphonate should be informed of such a complication though we must stress its rarity.展开更多
基金supported by a grant from the Innovation and Technology Commission of the Hong Kong SAR(Project MRP/001/18X)a grant from the Research Grants Council of the Hong Kong SAR(UGC/FDS24/E18/22).
文摘Background Generalized knee tissue segmentation,such as cartilage and meniscus in magnetic resonance imaging(MRI),plays a vital role in the clinical assessment of knee osteoarthritis(OA).However,domain variability between MRI datasets poses a significant challenge for the application of robust segmentation methods in real-world clinical settings.Existing unsupervised domain adaptation(UDA)approaches,which rely on one-to-one assumptions between the source and target domains,often fail to preserve knee tissues such as cartilage and meniscus,which are critical for OA diagnosis in diverse clinical settings.Methods We propose a source-independent segmentation approach tailored for multi-domain knee MRI datasets.Our method emphasizes knee tissue regions to reduce domain gaps and label inconsistencies.By introducing a stepwise adaptation strategy,segmentation performance was refined progressively from intermediate domains to the final target domain.Pseudo-label attention mechanisms were integrated into the adaptation pipeline,enabling iterative fine-tuning of domain-specific segmentations while leveraging unidirectional generative adversarial networks to enhance tissue-specific adaptation.This iterative training process ensures the generation of reliable pseudo-labels,thereby improving segmentation accuracy in diverse clinical MRI datasets.Results We demonstrated the effectiveness of our approach on the OA initiative dataset as the source domain and self-collected,T1-weighted fast field echo(T1FFE)as the intermediate domain and three-dimensional fast spin echo(3D FSE)as the final target domain.Our method achieved an average dice scores of 0.8701 and 0.7990 for source and target domains,respectively,surpassing the typical UDA methods explored in our experiments.Conclusion The experiments conducted on clinical MRI data,spanning OA severity from healthy knees to KL Grades 1–4,validated the effectiveness of the proposed domain adaptation method in precise segmentation of the cartilage and meniscus.
文摘Background The association of long-term bisphosphonate treatment for osteoporosis and related safety problems such as atypical fractures were not clearly defined. This study was to evaluate the structural, densitometric and biomechanical properties of the prolonged bisphosphonate-loaded bones. Methods Bone mineral density （BMD） at hip and femoral midshaft, bone cross-sectional area, moment of inertia of both femurs, bone formation and resorption biochemical markers were compared between 28 elderly with at least 4 years of bisphosphonate treatment from 2002 through 2006 and age-matched group of 37 elderly. Results The total hip BMD and t-score were found not different between two groups. However, bisphosphonate treated patients were found to have significantly lower bone mineral content in the femoral shaft （P 〈0.05）; morphological study showed lower cross-sectional area in subtrochanteric and mid-diaphyseal region and thus significantly lower moment of inertia （P 〈0.01）. High resolution-peripheral quantitative computed tomography showed significantly decreased trabeculardensity, bone volume ratio, trabecular number but increased trabecular spacing in tibia and distal radius. Finite element analysis further confirmed significantly lower stiffness and failure load in tibia. Biochemical studies also showed lower bone resorption and severely suppressed bone formation activity （P 〈0.001）. Conclusions The unchanged total hip BMD between two groups confirmed the beneficial effects of bisphosphonate on trabecular bone, thus preventing osteoporotic fractures at large in previous studies. However, the inferior structural, densitometric and biomechanical properties at cortical bones, especially femur midshaft, need a special attention to look into the association between long-term bisphosphonate intake and the occurrence of stress fractures. When patients taking bisphosphonate complain of proximal thigh pain or discomfort, plain X-ray film can be the first line screening. All patients prescribed with bisphosphonate should be informed of such a complication though we must stress its rarity.
