IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-21 DOI: 10.1016/j.knee.2025.10.022

Nadia Aghili , Katie Hughes , Kianoush Nazarpour , Hamish Simpson , Irene Yang

Background

The application of artificial intelligence (AI) in orthopaedic research, particularly for knee conditions, is growing rapidly. While AI offers the potential to enhance the efficiency and precision of care, its role and impact remain underexplored. This systematic scoping review evaluates studies employing AI algorithms, including machine learning and deep learning, to support clinical decision making for knee diseases.

Methods

A scoping review was conducted using the Joanna Briggs Institute methodology and PRISMA-ScR guidelines. MEDLINE, EMBASE, and ISI Web of Science were searched for English-language studies published between 2008 and 2025. Eligible studies involved adult patients and applied AI to diagnose knee conditions, predict outcomes, or support clinical processes. Extracted data included study demographics, targeted knee conditions, AI algorithms used, and their applications.

Results

From 2761 studies screened, 816 (30 %) were included after title and abstract screening. Among these, 66 % addressed diagnosis and 34 % focused on clinical prediction. Osteoarthritis was the most studied condition (71 %), followed by soft tissue damage (15 %). Deep learning was the most utilized AI method (43 %), followed by traditional machine learning (39 %). Among excluded studies, AI was indirectly applied in 21 %, primarily for identification (25 %), segmentation (42 %), and measurement (17 %).

Conclusions

Over the past 16 years, AI use in knee orthopaedic research has grown, yet only 30 % of studies directly addressed diagnostic or predictive applications. Challenges include limited reproducibility, generalizability, and clinical applicability of AI models. Future research should focus on improving reporting standards and exploring the application of AI in intra-operative, post-operative, and non-imaging use cases to enhance clinical utility.

背景：人工智能（AI）在骨科研究中的应用，特别是膝关节疾病的研究，正在迅速发展。虽然人工智能具有提高护理效率和准确性的潜力，但其作用和影响仍未得到充分探索。本系统的范围综述评估了使用人工智能算法（包括机器学习和深度学习）来支持膝关节疾病临床决策的研究。方法：采用乔安娜布里格斯研究所的方法和PRISMA-ScR指南进行范围审查。MEDLINE， EMBASE和ISI Web of Science检索了2008年至2025年间发表的英语研究。符合条件的研究涉及成年患者，并应用人工智能来诊断膝关节状况、预测预后或支持临床过程。提取的数据包括研究人口统计、目标膝关节状况、使用的人工智能算法及其应用。结果：在筛选的2761项研究中，经过标题和摘要筛选后纳入了816项（30%）研究。其中，66%致力于诊断，34%专注于临床预测。骨关节炎是研究最多的疾病（71%），其次是软组织损伤（15%）。深度学习是使用最多的人工智能方法（43%），其次是传统机器学习（39%）。在被排除的研究中，人工智能间接应用于21%的研究，主要用于识别（25%）、分割（42%）和测量（17%）。结论：在过去的16年里，人工智能在膝关节骨科研究中的应用有所增长，但只有30%的研究直接涉及诊断或预测应用。挑战包括人工智能模型有限的可重复性、普遍性和临床适用性。未来的研究应着眼于提高报告标准，探索人工智能在术中、术后和非影像学用例中的应用，以提高临床应用价值。

{"title":"The need for better quality studies: A systematic scoping review of current utility of artificial intelligence in orthopaedics and research gaps in the knee joint","authors":"Nadia Aghili , Katie Hughes , Kianoush Nazarpour , Hamish Simpson , Irene Yang","doi":"10.1016/j.knee.2025.10.022","DOIUrl":"10.1016/j.knee.2025.10.022","url":null,"abstract":"<div><h3>Background</h3><div>The application of artificial intelligence (AI) in orthopaedic research, particularly for knee conditions, is growing rapidly. While AI offers the potential to enhance the efficiency and precision of care, its role and impact remain underexplored. This systematic scoping review evaluates studies employing AI algorithms, including machine learning and deep learning, to support clinical decision making for knee diseases.</div></div><div><h3>Methods</h3><div>A scoping review was conducted using the Joanna Briggs Institute methodology and PRISMA-ScR guidelines. MEDLINE, EMBASE, and ISI Web of Science were searched for English-language studies published between 2008 and 2025. Eligible studies involved adult patients and applied AI to diagnose knee conditions, predict outcomes, or support clinical processes. Extracted data included study demographics, targeted knee conditions, AI algorithms used, and their applications.</div></div><div><h3>Results</h3><div>From 2761 studies screened, 816 (30 %) were included after title and abstract screening. Among these, 66 % addressed diagnosis and 34 % focused on clinical prediction. Osteoarthritis was the most studied condition (71 %), followed by soft tissue damage (15 %). Deep learning was the most utilized AI method (43 %), followed by traditional machine learning (39 %). Among excluded studies, AI was indirectly applied in 21 %, primarily for identification (25 %), segmentation (42 %), and measurement (17 %).</div></div><div><h3>Conclusions</h3><div>Over the past 16 years, AI use in knee orthopaedic research has grown, yet only 30 % of studies directly addressed diagnostic or predictive applications. Challenges include limited reproducibility, generalizability, and clinical applicability of AI models. Future research should focus on improving reporting standards and exploring the application of AI in intra-operative, post-operative, and non-imaging use cases to enhance clinical utility.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104268"},"PeriodicalIF":2.0,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145582767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Implant camouflage – how apparently excellent registry data can mask poor implant results: an international multi-centre case-control study 植入伪装——明显优秀的注册数据如何掩盖糟糕的植入结果：一项国际多中心病例对照研究。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-20 DOI: 10.1016/j.knee.2025.11.004

Gerard A. Sheridan , Michael E. Neufeld , Lisa C. Howard , David Keohane , Eric Masterson , Dan Cadoux-Hudson , Simon Tilley , Doug Dunlop , John Keith Tucker , Bassam A. Masri , Donald A. Garbuz

Background

The purpose of this research is to highlight a new issue facing national joint registries called ’Implant Camouflage’, whereby a poorly performing implant combination within an overall brand portfolio is hidden by other high performing implant combinations.

Methods

This was an international multi-centre case-control study assessing implant combinations within the NexGen® TKA portfolio. There were 9392 primary TKA components included with 1744 ‘high-risk’ cases and 7648 ‘standard-risk’ controls. The median time to follow-up was 7.4 (IQR 4.8–9.4) years in the cases and 10.4 (IQR 7.5–13.8) years in the control group. The primary outcome variable was revision for aseptic loosening. All-cause revision was a secondary outcome variable. Cox regression and Kaplan-Meier (KM) survival analysis was used to determine survivorship free of revision for aseptic loosening.

Results

The high-risk cases had significantly higher rates of revision for aseptic loosening (5.2 % v 0.4 %, p < 0.001) and all-cause revision (7.9 % v 2.9 %, p < 0.001). KM curves demonstrated higher rates of aseptic loosening in the high-risk cases with no intersection of 95 % confidence intervals with the controls. On univariate analysis, older age (p < 0.001) and higher BMI (p < 0.001) patients had significantly higher rates of revision for aseptic loosening. After multivariate analysis, the high-risk component cases were the most significant predictor of early failure for aseptic tibial loosening (p < 0.001).

Conclusions

The NexGen® Option-LPS Flex and Option-LPS GSF combinations have over a 10-fold higher revision rate for aseptic loosening when compared to other implant combinations in the NexGen® portfolio.

背景：本研究的目的是强调国家联合注册所面临的一个新问题，称为“种植体伪装”，即在整个品牌组合中表现不佳的种植体组合被其他高性能种植体组合所隐藏。方法：这是一项国际多中心病例对照研究，评估NexGen®TKA产品组合中的种植体组合。9392个主要TKA组成部分包括1744个“高风险”病例和7648个“标准风险”控制。中位随访时间：病例为7.4 （IQR 4.8-9.4）年，对照组为10.4 （IQR 7.5-13.8）年。主要结局变量为无菌性松动的翻修。全因修正是次要结局变量。采用Cox回归和Kaplan-Meier （KM）生存分析来确定无菌性松动不进行翻修的生存率。结果：高危病例的无菌松动翻修率明显更高（5.2% vs 0.4%, p）。结论：与NexGen®产品组合中的其他种植体组合相比，NexGen®Option-LPS Flex和Option-LPS GSF组合的无菌松动翻修率高出10倍以上。

{"title":"Implant camouflage – how apparently excellent registry data can mask poor implant results: an international multi-centre case-control study","authors":"Gerard A. Sheridan , Michael E. Neufeld , Lisa C. Howard , David Keohane , Eric Masterson , Dan Cadoux-Hudson , Simon Tilley , Doug Dunlop , John Keith Tucker , Bassam A. Masri , Donald A. Garbuz","doi":"10.1016/j.knee.2025.11.004","DOIUrl":"10.1016/j.knee.2025.11.004","url":null,"abstract":"<div><h3>Background</h3><div>The purpose of this research is to highlight a new issue facing national joint registries called ’Implant Camouflage’, whereby a poorly performing implant combination within an overall brand portfolio is hidden by other high performing implant combinations.</div></div><div><h3>Methods</h3><div>This was an international multi-centre case-control study assessing implant combinations within the NexGen® TKA portfolio. There were 9392 primary TKA components included with 1744 ‘high-risk’ cases and 7648 ‘standard-risk’ controls. The median time to follow-up was 7.4 (IQR 4.8–9.4) years in the cases and 10.4 (IQR 7.5–13.8) years in the control group. The primary outcome variable was revision for aseptic loosening. All-cause revision was a secondary outcome variable. Cox regression and Kaplan-Meier (KM) survival analysis was used to determine survivorship free of revision for aseptic loosening.</div></div><div><h3>Results</h3><div>The high-risk cases had significantly higher rates of revision for aseptic loosening (5.2 % v 0.4 %, <em>p</em> < 0.001) and all-cause revision (7.9 % v 2.9 %, <em>p</em> < 0.001). KM curves demonstrated higher rates of aseptic loosening in the high-risk cases with no intersection of 95 % confidence intervals with the controls. On univariate analysis, older age (<em>p</em> < 0.001) and higher BMI (<em>p</em> < 0.001) patients had significantly higher rates of revision for aseptic loosening. After multivariate analysis, the high-risk component cases were the most significant predictor of early failure for aseptic tibial loosening (<em>p</em> < 0.001).</div></div><div><h3>Conclusions</h3><div>The NexGen® Option-LPS Flex and Option-LPS GSF combinations have over a 10-fold higher revision rate for aseptic loosening when compared to other implant combinations in the NexGen® portfolio.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104278"},"PeriodicalIF":2.0,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145574571","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Response to letter to the editor on “Periarticular Injection Versus Placebo in Total Knee Arthroplasty with Intrathecal Morphine” 对“鞘内吗啡全膝关节置换术中关节周围注射与安慰剂”致编辑的信的回应。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-19 DOI: 10.1016/j.knee.2025.10.019

引用次数: 0

Online information on MACI knee surgery: analysis and opportunities to improve patient education and decision-making MACI膝关节手术的在线信息：分析和改善患者教育和决策的机会。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-19 DOI: 10.1016/j.knee.2025.10.027

Camila Vicioso , Ava G. Neijna , Hannah L. Terry , Luca Valdivia , Laurel Wong , Renee Ren , Sabrina M. Strickland

Background

Patients increasingly turn to online search engines to learn about orthopedic procedures. Matrix-Induced Autologous Chondrocyte Implantation (MACI) is an increasingly popular treatment for cartilage lesions. However, the visibility, range, and credibility of online information about MACI remain unclear. As MACI’s clinical use grows, understanding what patients are encountering online is essential to guiding informed discussions.

Purpose

To (1) identify frequently asked questions about MACI and their thematic categories, and (2) evaluate the types and credibility of linked websites in order to inform a patient-friendly handout with frequently asked questions (FAQs) and high-quality sources.

Methods

We conducted a cross-sectional analysis of 1620 MACI-related entries from Google’s “People Also Ask” feature. Questions were categorized using the Rothwell framework, and scored for credibility using JAMA Benchmark Criteria. Kruskal-Wallis tests compared credibility across groups; logistic regression identified predictors of high-credibility content.

Results

Most included questions (n = 1107) concerned technical details (n = 285), evaluation of surgery (n = 220), and cost (n = 138). Commercial websites were most common (40.9 %) and, while they often offered patient-friendly content, they had lower average JAMA credibility scores (1.54) compared to government (3.95) and academic (1.97) sources (p < 0.01).

Conclusion

Online information about MACI is dominated by commercial websites. While many offer helpful information, key patient concerns, such as pain, are underaddressed in online resources. Surgeon-led efforts, including recommending trusted sources and creating patient-friendly online content (e.g. using keywords, images, structured headings), may improve the visibility and credibility of MACI information, supporting patient understanding and shared decision-making.

背景：越来越多的患者通过在线搜索引擎来了解骨科手术。基质诱导的自体软骨细胞植入（MACI）是一种越来越流行的治疗软骨病变的方法。然而，关于MACI的在线信息的可见性、范围和可信度仍然不清楚。随着MACI的临床应用的增长，了解患者在网上遇到的情况对于指导知情讨论至关重要。目的：(1)确定关于MACI的常见问题及其主题类别，(2)评估链接网站的类型和可信度，以便为患者提供方便的常见问题（FAQs）和高质量来源的讲义。方法：我们对b谷歌的“People Also Ask”功能中的1620个maci相关条目进行了横断面分析。使用Rothwell框架对问题进行分类，并使用JAMA基准标准对可信度进行评分。Kruskal-Wallis测试比较各组之间的可信度；逻辑回归确定了高可信度内容的预测因子。结果：大多数被纳入的问题（n = 1107）涉及技术细节（n = 285）、手术评估（n = 220）和费用（n = 138）。商业网站最常见（40.9%），虽然他们经常提供对患者友好的内容，但与政府（3.95）和学术（1.97）来源相比，他们的JAMA可信度平均得分（1.54）较低(p结论：关于MACI的在线信息主要是商业网站。虽然许多网站提供了有用的信息，但患者关心的关键问题，如疼痛，在在线资源中没有得到充分解决。由外科医生主导的工作，包括推荐可信来源和创建对患者友好的在线内容（例如使用关键词、图像、结构化标题），可以提高MACI信息的可见性和可信度，支持患者理解和共同决策。

{"title":"Online information on MACI knee surgery: analysis and opportunities to improve patient education and decision-making","authors":"Camila Vicioso , Ava G. Neijna , Hannah L. Terry , Luca Valdivia , Laurel Wong , Renee Ren , Sabrina M. Strickland","doi":"10.1016/j.knee.2025.10.027","DOIUrl":"10.1016/j.knee.2025.10.027","url":null,"abstract":"<div><h3>Background</h3><div>Patients increasingly turn to online search engines to learn about orthopedic procedures. Matrix-Induced Autologous Chondrocyte Implantation (MACI) is an increasingly popular treatment for cartilage lesions. However, the visibility, range, and credibility of online information about MACI remain unclear. As MACI’s clinical use grows, understanding what patients are encountering online is essential to guiding informed discussions.</div></div><div><h3>Purpose</h3><div>To (1) identify frequently asked questions about MACI and their thematic categories, and (2) evaluate the types and credibility of linked websites in order to inform a patient-friendly handout with frequently asked questions (FAQs) and high-quality sources.</div></div><div><h3>Methods</h3><div>We conducted a cross-sectional analysis of 1620 MACI-related entries from Google’s “People Also Ask” feature. Questions were categorized using the Rothwell framework, and scored for credibility using JAMA Benchmark Criteria. Kruskal-Wallis tests compared credibility across groups; logistic regression identified predictors of high-credibility content.</div></div><div><h3>Results</h3><div>Most included questions (<em>n</em> = 1107) concerned technical details (<em>n</em> = 285), evaluation of surgery (<em>n</em> = 220), and cost (<em>n</em> = 138). Commercial websites were most common (40.9 %) and, while they often offered patient-friendly content, they had lower average JAMA credibility scores (1.54) compared to government (3.95) and academic (1.97) sources (<em>p</em> < 0.01).</div></div><div><h3>Conclusion</h3><div>Online information about MACI is dominated by commercial websites. While many offer helpful information, key patient concerns, such as pain, are underaddressed in online resources. Surgeon-led efforts, including recommending trusted sources and creating patient-friendly online content (e.g. using keywords, images, structured headings), may improve the visibility and credibility of MACI information, supporting patient understanding and shared decision-making.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104273"},"PeriodicalIF":2.0,"publicationDate":"2025-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145566434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

No differences in clinical outcomes and return to sport in patients with sex-mismatched meniscal allograft transplantation: Average 6-year follow up 性别不匹配的半月板同种异体移植患者的临床结果和恢复运动无差异：平均6年随访。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-17 DOI: 10.1016/j.knee.2025.11.001

Jairo Triana, Zachary I. Li, Allison Morgan, Vishal Sundaram, Guillem Gonzalez-Lomas, Michael Alaia, Eric Strauss, Laith Jazrawi, Kirk Campbell

Background

Risks associated with graft transplantation due to donor–recipient sex differences have been reported in whole organ transplantation literature but are not well-described in meniscal allograft transplantation (MAT). This study assessed the effect of donor–recipient sex mismatch on patient-reported outcomes (PROs) following MAT.

Methods

A retrospective review of patients <50 years of age that underwent unicompartmental MAT from 2010 to 2023 at a single institution with minimum 2-year follow up was conducted. Exclusion criteria were revision procedures or insufficient donor information. PROs collected included: Knee Injury and Osteoarthritis and Outcome Score (KOOS), Visual Analog scale (VAS) for pain and satisfaction, and return to sport (RTS). PROs were compared between donor–recipient graft matching groups (same-sex (SS) or different-sex (DS)) and by donor sex.

Results

Fifty-seven patients (31 males, 26 females) met inclusion criteria with mean age of 29.1 ± 8.9 years, BMI (body mass index) of 26.8 ± 5.7 kg/m², and follow up of 73.4 ± 29.2 months. Donors were mostly male (77.2 %). Thirty-six patients (63.3 %) received SS grafts and 21 (36.8 %) received DS grafts. There were no significant differences between groups in KOOS subscores, VAS pain, satisfaction scores, or RTS rates (P > 0.05). Donor sex also showed no association with PROs. Linear and logistic regression models controlling for age, BMI, and average follow up demonstrated that graft mismatch was not associated with KOOS scores, RTS, or graft failure. Decreasing patient age was associated with improved KOOS-pain (β: −0.376; 95 % CI (−1.892, −0.155); P = 0.022).

Conclusions

Patients that undergo MAT with grafts from a different sex had similar outcomes to patients receiving grafts from a donor of the same sex. Sex-mismatch was not associated with functional outcomes or RTS. The results of this paper support the use of geometrically matched donor grafts, independent of donor sex, which should expand available graft options when matching a patient for MAT.

背景：在全器官移植文献中已经报道了供体-受体性别差异导致的移植风险，但在半月板同种异体移植（MAT）中没有很好的描述。方法：回顾性分析57例患者（男31例，女26例）符合纳入标准，平均年龄29.1±8.9岁，身体质量指数（BMI） 26.8±5.7 kg/m2，随访73.4±29.2个月。献血者多为男性（77.2%）。36例（63.3%）接受SS移植，21例（36.8%）接受DS移植。两组间KOOS评分、VAS疼痛评分、满意度评分、RTS率差异均无统计学意义（P < 0.05）。供体性别也与PROs无关。控制年龄、BMI和平均随访的线性和逻辑回归模型表明，移植物不匹配与KOOS评分、RTS或移植物失败无关。患者年龄的降低与KOOS-pain的改善相关(β: -0.376; 95% CI (-1.892, -0.155)；p = 0.022)。结论：接受不同性别移植的MAT患者与接受同性供者移植的患者预后相似。性别不匹配与功能结局或RTS无关。本文的结果支持使用几何匹配的供体移植物，独立于供体性别，这应该扩大可用的移植物选择时，匹配患者的MAT。

{"title":"No differences in clinical outcomes and return to sport in patients with sex-mismatched meniscal allograft transplantation: Average 6-year follow up","authors":"Jairo Triana, Zachary I. Li, Allison Morgan, Vishal Sundaram, Guillem Gonzalez-Lomas, Michael Alaia, Eric Strauss, Laith Jazrawi, Kirk Campbell","doi":"10.1016/j.knee.2025.11.001","DOIUrl":"10.1016/j.knee.2025.11.001","url":null,"abstract":"<div><h3>Background</h3><div>Risks associated with graft transplantation due to donor–recipient sex differences have been reported in whole organ transplantation literature but are not well-described in meniscal allograft transplantation (MAT). This study assessed the effect of donor–recipient sex mismatch on patient-reported outcomes (PROs) following MAT.</div></div><div><h3>Methods</h3><div>A retrospective review of patients <50 years of age that underwent unicompartmental MAT from 2010 to 2023 at a single institution with minimum 2-year follow up was conducted. Exclusion criteria were revision procedures or insufficient donor information. PROs collected included: Knee Injury and Osteoarthritis and Outcome Score (KOOS), Visual Analog scale (VAS) for pain and satisfaction, and return to sport (RTS). PROs were compared between donor–recipient graft matching groups (same-sex (SS) or different-sex (DS)) and by donor sex.</div></div><div><h3>Results</h3><div>Fifty-seven patients (31 males, 26 females) met inclusion criteria with mean age of 29.1 ± 8.9 years, BMI (body mass index) of 26.8 ± 5.7 kg/m<sup>2</sup>, and follow up of 73.4 ± 29.2 months. Donors were mostly male (77.2 %). Thirty-six patients (63.3 %) received SS grafts and 21 (36.8 %) received DS grafts. There were no significant differences between groups in KOOS subscores, VAS pain, satisfaction scores, or RTS rates (<em>P</em> > 0.05). Donor sex also showed no association with PROs. Linear and logistic regression models controlling for age, BMI, and average follow up demonstrated that graft mismatch was not associated with KOOS scores, RTS, or graft failure. Decreasing patient age was associated with improved KOOS-pain (β: −0.376; 95 % CI (−1.892, −0.155); <em>P</em> = 0.022).</div></div><div><h3>Conclusions</h3><div>Patients that undergo MAT with grafts from a different sex had similar outcomes to patients receiving grafts from a donor of the same sex. Sex-mismatch was not associated with functional outcomes or RTS. The results of this paper support the use of geometrically matched donor grafts, independent of donor sex, which should expand available graft options when matching a patient for MAT.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104275"},"PeriodicalIF":2.0,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145552004","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

How accurately do large language models answer patient questions on anterior cruciate ligament tears? A comparative study 大型语言模型如何准确地回答前交叉韧带撕裂患者的问题？比较研究。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-17 DOI: 10.1016/j.knee.2025.11.002

Salim Youssef , Anton Brehmer , Peter Melcher , Ralf Henkelmann , Pierre Hepp , Jan Theopold , Maria Elze , Yasmin Youssef , Georg Osterhoff

Background

Large language models (LLMs) are increasingly used in the medical sector, raising questions about their reliability for patient education. With more LLMs becoming publicly available, it remains unclear whether meaningful performance differences exist between them. This is particularly relevant for anterior cruciate ligament (ACL) injuries, which mainly affect young, active individuals, those most likely to seek health advice from AI. This study aimed to evaluate and directly compare the accuracy of five leading LLMs in answering common patient questions about ACL tears.

Methods

Fourteen commonly asked patient questions were identified in a systematic online search. Each question was submitted to five LLMs: ChatGPT-4, Gemini 2.0, Llama 3.1, DeepSeek-V3, and Grok3. Responses were assessed for accuracy by orthopedic consultants using a five-point Likert scale. Word count was recorded as a proxy for readability. Statistical analysis included ANOVA by Tukey’s HSD post hoc test.

Results

All models achieved mean accuracy scores ≥3 (mostly accurate). DeepSeek (3.61) and Grok (3.59) demonstrated significantly higher mean accuracies than Llama (3.25; P < 0.05). ChatGPT and Gemini achieved mean scores of 3.48 and 3.52, respectively. Models generating longer responses, such as Grok and DeepSeek, tended to offer greater accuracy, whereas Llama produced the shortest and least accurate answers.

Conclusions

All tested LLMs show promise for patient education regarding ACL injuries, but notable performance differences exist. Model choice is therefore critical. While all responses were evaluated by clinical experts, the lack of guideline-based validation highlights the need for further studies assessing both accuracy and patient comprehension.

背景：大型语言模型（llm）越来越多地应用于医疗领域，这引起了人们对其在患者教育方面的可靠性的质疑。随着越来越多的法学硕士公开上市，它们之间是否存在有意义的性能差异仍不清楚。这与前交叉韧带（ACL）损伤尤其相关，这种损伤主要影响年轻、活跃的个体，他们最有可能向人工智能寻求健康建议。本研究旨在评估并直接比较五种领先的llm在回答有关ACL撕裂的常见患者问题时的准确性。方法：通过系统的在线搜索，确定了14个常见的患者问题。每个问题提交给五个llm: ChatGPT-4、Gemini 2.0、Llama 3.1、DeepSeek-V3和Grok3。骨科顾问使用李克特五分制评估反应的准确性。字数被记录下来作为可读性的代表。统计分析采用Tukey’s HSD事后检验的方差分析。结果：所有模型的平均准确率得分均≥3分（大部分准确）。DeepSeek（3.61）和Grok（3.59）的平均准确率明显高于Llama（3.25）。结论：所有测试的llm都有希望对患者进行ACL损伤的教育，但存在显着的性能差异。因此，模式选择至关重要。虽然所有的反应都由临床专家评估，但缺乏基于指南的验证强调了进一步研究评估准确性和患者理解的必要性。

{"title":"How accurately do large language models answer patient questions on anterior cruciate ligament tears? A comparative study","authors":"Salim Youssef , Anton Brehmer , Peter Melcher , Ralf Henkelmann , Pierre Hepp , Jan Theopold , Maria Elze , Yasmin Youssef , Georg Osterhoff","doi":"10.1016/j.knee.2025.11.002","DOIUrl":"10.1016/j.knee.2025.11.002","url":null,"abstract":"<div><h3>Background</h3><div>Large language models (LLMs) are increasingly used in the medical sector, raising questions about their reliability for patient education. With more LLMs becoming publicly available, it remains unclear whether meaningful performance differences exist between them. This is particularly relevant for anterior cruciate ligament (ACL) injuries, which mainly affect young, active individuals, those most likely to seek health advice from AI. This study aimed to evaluate and directly compare the accuracy of five leading LLMs in answering common patient questions about ACL tears.</div></div><div><h3>Methods</h3><div>Fourteen commonly asked patient questions were identified in a systematic online search. Each question was submitted to five LLMs: ChatGPT-4, Gemini 2.0, Llama 3.1, DeepSeek-V3, and Grok3. Responses were assessed for accuracy by orthopedic consultants using a five-point Likert scale. Word count was recorded as a proxy for readability. Statistical analysis included ANOVA by Tukey’s HSD post hoc test.</div></div><div><h3>Results</h3><div>All models achieved mean accuracy scores ≥3 (mostly accurate). DeepSeek (3.61) and Grok (3.59) demonstrated significantly higher mean accuracies than Llama (3.25; <em>P</em> < 0.05). ChatGPT and Gemini achieved mean scores of 3.48 and 3.52, respectively. Models generating longer responses, such as Grok and DeepSeek, tended to offer greater accuracy, whereas Llama produced the shortest and least accurate answers.</div></div><div><h3>Conclusions</h3><div>All tested LLMs show promise for patient education regarding ACL injuries, but notable performance differences exist. Model choice is therefore critical. While all responses were evaluated by clinical experts, the lack of guideline-based validation highlights the need for further studies assessing both accuracy and patient comprehension.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104276"},"PeriodicalIF":2.0,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145551934","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robotic assisted fixed bearing medial UKA: can soft tissue balancing and restoration of alignment modify PFJ kinematics in the presence of chondral wear 机器人辅助固定轴承内侧UKA：在存在软骨磨损的情况下，软组织平衡和对齐恢复能否改变PFJ的运动学。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-17 DOI: 10.1016/j.knee.2025.10.023

Ashok Kumar P.S. , Aashay Sonkusale , Shantanu Patil , Pichai Suryanarayan , Kalaivanan Kanniyan , Vijay Bose

Background

The impact of patellofemoral chondral lesions on the outcomes following Unicondylar Knee Arthroplasty (UKA) remains an enigmatic topic among arthroplasty surgeons. We aim to compare the clinical outcomes of medial UKAs, with varying grades of patellofemoral wear and ACL quality, to help in decision making for suitability of UKA.

Methods

A total of 195 robotic-assisted fixed-bearing UKAs performed between 2018 and 2024 were identified from our institutional database. Chondral lesions in the trochlea, both medial and lateral facets of the patella, and the lateral compartment were classified according to the International Cartilage Repair Society classification as low grade (0–2) and high grade (3/4). The Oxford Knee Score and Modified Kujala Score, a score specific for patellofemoral pain, were used for functional assessment. Intraclass correlation was used to compare the chondral damage and outcomes.

Results

The mean follow-up was 44 ± 10 months (11–56). There was no difference between the OKS and MKS compared across the ICRS grades for trochlea, medial/lateral patellar facet, or lateral compartment damage (p > 0.05). There was no significant difference in the varying grades of ACL damage (p > 0.05). No revisions are seen until last follow-up.

Conclusions

Intactness of 3 ‘Ls’ – ‘L’ateral compartment cartilage including lateral meniscus, adequately tracking ‘L’ateral facet of the patella and functional stability of the AC‘L’ are the most important caveats for determining the suitability of UKA. Restoration of the joint line, pre-arthritic soft tissue balance and alignment are crucial. The use of imageless robotics technology has helped produce consistent results with the optimisation of the surgical workflow.

背景：髌股软骨病变对单髁膝关节置换术（UKA）后预后的影响在关节置换术中仍然是一个谜。我们的目的是比较不同程度髌骨磨损和前交叉韧带质量的内侧UKA的临床结果，以帮助决策UKA的适用性。方法：从我们的机构数据库中识别出2018年至2024年期间进行的195例机器人辅助固定轴承uka。滑车、髌骨内侧面和外侧侧面以及外侧腔室的软骨病变按照国际软骨修复协会的分类分为低级别（0-2）和高级别（3/4）。牛津膝关节评分和改良库贾拉评分（一种专门针对髌股疼痛的评分）用于功能评估。用类内相关性比较软骨损伤和预后。结果：平均随访44±10个月（11-56）。在滑车、髌内/外侧关节面或外侧腔室损伤的ICRS分级中，OKS和MKS之间没有差异（p < 0.05）。不同程度ACL损伤比较差异无统计学意义（p < 0.05）。直到最后一次跟进才看到任何修订。结论：包括外侧半月板在内的3 'L' - ‘L‘外侧腔室软骨的完整性，髌骨’L’外侧小面的充分跟踪以及AC‘L’的功能稳定性是确定UKA适宜性的最重要因素。关节线的恢复，关节炎前的软组织平衡和对齐是至关重要的。使用无图像机器人技术有助于产生一致的结果，优化手术工作流程。

{"title":"Robotic assisted fixed bearing medial UKA: can soft tissue balancing and restoration of alignment modify PFJ kinematics in the presence of chondral wear","authors":"Ashok Kumar P.S. , Aashay Sonkusale , Shantanu Patil , Pichai Suryanarayan , Kalaivanan Kanniyan , Vijay Bose","doi":"10.1016/j.knee.2025.10.023","DOIUrl":"10.1016/j.knee.2025.10.023","url":null,"abstract":"<div><h3>Background</h3><div>The impact of patellofemoral chondral lesions on the outcomes following Unicondylar Knee Arthroplasty (UKA) remains an enigmatic topic among arthroplasty surgeons. We aim to compare the clinical outcomes of medial UKAs, with varying grades of patellofemoral wear and ACL quality, to help in decision making for suitability of UKA.</div></div><div><h3>Methods</h3><div>A total of 195 robotic-assisted fixed-bearing UKAs performed between 2018 and 2024 were identified from our institutional database. Chondral lesions in the trochlea, both medial and lateral facets of the patella, and the lateral compartment were classified according to the International Cartilage Repair Society classification as low grade (0–2) and high grade (3/4). The Oxford Knee Score and Modified Kujala Score, a score specific for patellofemoral pain, were used for functional assessment. Intraclass correlation was used to compare the chondral damage and outcomes.</div></div><div><h3>Results</h3><div>The mean follow-up was 44 ± 10 months (11–56). There was no difference between the OKS and MKS compared across the ICRS grades for trochlea, medial/lateral patellar facet, or lateral compartment damage (<em>p</em> > 0.05). There was no significant difference in the varying grades of ACL damage (<em>p</em> > 0.05). No revisions are seen until last follow-up.</div></div><div><h3>Conclusions</h3><div>Intactness of 3 ‘Ls’ – ‘L’ateral compartment cartilage including lateral meniscus, adequately tracking ‘L’ateral facet of the patella and functional stability of the AC‘L’ are the most important caveats for determining the suitability of UKA. Restoration of the joint line, pre-arthritic soft tissue balance and alignment are crucial. The use of imageless robotics technology has helped produce consistent results with the optimisation of the surgical workflow.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104269"},"PeriodicalIF":2.0,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145552010","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Can a machine learning approach contribute to monitoring post-market surveillance of total knee arthroplasty prostheses? 机器学习方法能否有助于全膝关节置换术假体上市后的监测？

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-15 DOI: 10.1016/j.knee.2025.10.020

Khashayar Ghadirinejad , Stephen Graves , Richard de Steiger , Nicole Pratt , Lucian B. Solomon , Mark Taylor , Reza Hashemi

Background

The impact of knee replacement surgery on patients can be monitored by joint registries through the surveillance of prostheses and identification of under-performing devices – outliers. The study of new statistical methods can help determine whether a device is at a higher risk of failure by considering possible confounding factors. Self-learning algorithms with the potential to involve multiple variables simultaneously are one approach to limiting the impact of confounding factors. This study aimed to assess two machine learning (ML) techniques to detect total knee outliers while controlling for patient- and device-related confounding.

Methods

The potential to identify outliers among 160 unique prostheses was evaluated for Random Survival Forest (RSF) and regularised/unregularised Cox models. The input variables included femoral/tibial components and patient characteristics provided by the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) on 265,655 primary total knee procedures. The outcome was time to first revision surgery treated as a censored case for death, and the AOANJRR gold standard was defined as the criteria to assess the effectiveness of proposed ML methods.

Results

In the study cohort, the AOANJRR standardised approach detected five conventional prosthesis combinations. Both the Cox and RSF techniques identified two of the same total knee prostheses. The regularised/unregularised Cox results were more comparable to the AOANJRR standard by detecting one additional prosthesis at a higher risk of revision.

Conclusion

Machine learning may offer a supplementary approach for the identification of prosthesis outliers. However, further analysis is required to fully comprehend the effect of confounding factors and the potential contribution of ML to the early identification of outliers.

背景：膝关节置换手术对患者的影响可以通过联合登记来监测假体和识别表现不佳的异常装置。新的统计方法的研究可以通过考虑可能的混杂因素来帮助确定一个设备是否有更高的失效风险。有可能同时涉及多个变量的自学习算法是限制混杂因素影响的一种方法。本研究旨在评估两种机器学习（ML）技术，以检测全膝关节异常值，同时控制患者和设备相关的混淆。方法：利用随机生存森林（RSF）和正则/非正则Cox模型评估160个独特假肢中识别异常值的潜力。输入变量包括由澳大利亚骨科协会国家关节置换登记处（AOANJRR）提供的265,655例初级全膝关节手术的股骨/胫骨部件和患者特征。结果是第一次翻修手术的时间，作为死亡审查病例，AOANJRR金标准被定义为评估建议的ML方法有效性的标准。结果：在研究队列中，AOANJRR标准化方法检测到5种常规假体组合。Cox和RSF技术都确定了两个相同的全膝关节假体。通过检测一个具有较高翻修风险的额外假体，规范化/非规范化Cox结果与AOANJRR标准更具可比性。结论：机器学习为义肢异常点识别提供了一种补充方法。然而，需要进一步的分析来充分理解混杂因素的影响以及ML对早期识别异常值的潜在贡献。

{"title":"Can a machine learning approach contribute to monitoring post-market surveillance of total knee arthroplasty prostheses?","authors":"Khashayar Ghadirinejad , Stephen Graves , Richard de Steiger , Nicole Pratt , Lucian B. Solomon , Mark Taylor , Reza Hashemi","doi":"10.1016/j.knee.2025.10.020","DOIUrl":"10.1016/j.knee.2025.10.020","url":null,"abstract":"<div><h3>Background</h3><div>The impact of knee replacement surgery on patients can be monitored by joint registries through the surveillance of prostheses and identification of under-performing devices – outliers. The study of new statistical methods can help determine whether a device is at a higher risk of failure by considering possible confounding factors. Self-learning algorithms with the potential to involve multiple variables simultaneously are one approach to limiting the impact of confounding factors. This study aimed to assess two machine learning (ML) techniques to detect total knee outliers while controlling for patient- and device-related confounding.</div></div><div><h3>Methods</h3><div>The potential to identify outliers among 160 unique prostheses was evaluated for Random Survival Forest (RSF) and regularised/unregularised Cox models. The input variables included femoral/tibial components and patient characteristics provided by the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) on 265,655 primary total knee procedures. The outcome was time to first revision surgery treated as a censored case for death, and the AOANJRR gold standard was defined as the criteria to assess the effectiveness of proposed ML methods.</div></div><div><h3>Results</h3><div>In the study cohort, the AOANJRR standardised approach detected five conventional prosthesis combinations. Both the Cox and RSF techniques identified two of the same total knee prostheses. The regularised/unregularised Cox results were more comparable to the AOANJRR standard by detecting one additional prosthesis at a higher risk of revision.</div></div><div><h3>Conclusion</h3><div>Machine learning may offer a supplementary approach for the identification of prosthesis outliers. However, further analysis is required to fully comprehend the effect of confounding factors and the potential contribution of ML to the early identification of outliers.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104266"},"PeriodicalIF":2.0,"publicationDate":"2025-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145535037","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Reevaluating periarticular injection for pain control in total knee arthroplasty 全膝关节置换术中关节周围注射对疼痛控制的再评价。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-15 DOI: 10.1016/j.knee.2025.10.018

Tien-Yu Fu, Chia-Hao Hsu

引用次数: 0

Deep learning-enabled segmentation of knee cartilage in conventional magnetic resonance images: Internal and external validation of different models 基于深度学习的常规磁共振图像中膝关节软骨分割：不同模型的内部和外部验证。

IF 2 4区医学 Q3 ORTHOPEDICS

Knee

Pub Date : 2025-11-14 DOI: 10.1016/j.knee.2025.10.024

Lin Yao , Xiaoling Liang , Jin Liu , Baoxin Qian , Zhongli Xiao , Dantian Zhu , Jiaxin Feng , Shouguo Zhou , Huaqian Cui , Shaolin Li , Wei Li

Background

Accurate evaluation of the cartilage anatomy of the knee is helpful for clinical evaluation of the source of knee pain and the classification and treatment of knee osteoarthritis (OA). This study proposes a deep learning model for segmentation of knee articular cartilage in conventional proton density fat-saturated MRI sequences to assess cartilage morphology for subsequent injury grading.

Methods

This retrospective study was conducted at two radiology centers, involving 254 knees from 254 patients who had previously undergone MRI scans. The training-internal validation cohort included 219 knees from Center 1. The external validation cohort comprised 35 knees from Center 2. Two musculoskeletal radiology experts manually annotated the cartilage regions. A 3D Res U-net model was employed for segmentation, and its performance was compared with 3D U-net and 3D V-net models. Segmentation results were evaluated using the Dice coefficient and Jaccard index.

Results

The 3D Res U-net model demonstrated superior segmentation performance compared to the other deep learning methods. For cartilage in the lateral femorotibial joint, medial femorotibial joint, and patellofemoral joint, the average Dice coefficients with 3D Res U-net were 0.871, 0.860, and 0.858 in internal validation and 0.846, 0.837, and 0.819 in external validation, respectively. The Jaccard index followed a similar trend.

Conclusion

The 3D Res U-net model improves knee cartilage segmentation in conventional MR imaging, contributing to the understanding of cartilage morphology and the improvement of clinically relevant decisions.

背景：准确评估膝关节软骨解剖结构有助于临床评估膝关节疼痛的来源及膝关节骨性关节炎（OA）的分类和治疗。本研究提出了一种深度学习模型，用于在常规质子密度脂肪饱和MRI序列中分割膝关节软骨，以评估软骨形态，以便随后进行损伤分级。方法：本回顾性研究在两个放射学中心进行，涉及254例患者的254个膝关节，这些患者之前接受过MRI扫描。训练-内部验证队列包括来自中心1的219个膝关节。外部验证队列包括来自中心2的35个膝关节。两名肌肉骨骼放射学专家手工注释了软骨区域。采用三维Res U-net模型进行分割，并与三维U-net模型和三维V-net模型进行了性能比较。使用Dice系数和Jaccard指数对分割结果进行评价。结果：与其他深度学习方法相比，3D Res U-net模型表现出更好的分割性能。股胫外侧关节、股胫内侧关节和髌股关节软骨的3D Res U-net平均Dice系数在内部验证中分别为0.871、0.860和0.858，在外部验证中分别为0.846、0.837和0.819。Jaccard指数也遵循了类似的趋势。结论：3D Res U-net模型改善了常规MR成像中的膝关节软骨分割，有助于对软骨形态的理解和临床相关决策的改善。

{"title":"Deep learning-enabled segmentation of knee cartilage in conventional magnetic resonance images: Internal and external validation of different models","authors":"Lin Yao , Xiaoling Liang , Jin Liu , Baoxin Qian , Zhongli Xiao , Dantian Zhu , Jiaxin Feng , Shouguo Zhou , Huaqian Cui , Shaolin Li , Wei Li","doi":"10.1016/j.knee.2025.10.024","DOIUrl":"10.1016/j.knee.2025.10.024","url":null,"abstract":"<div><h3>Background</h3><div>Accurate evaluation of the cartilage anatomy of the knee is helpful for clinical evaluation of the source of knee pain and the classification and treatment of knee osteoarthritis (OA). This study proposes a deep learning model for segmentation of knee articular cartilage in conventional proton density fat-saturated MRI sequences to assess cartilage morphology for subsequent injury grading.</div></div><div><h3>Methods</h3><div>This retrospective study was conducted at two radiology centers, involving 254 knees from 254 patients who had previously undergone MRI scans. The training-internal validation cohort included 219 knees from Center 1. The external validation cohort comprised 35 knees from Center 2. Two musculoskeletal radiology experts manually annotated the cartilage regions. A 3D Res U-net model was employed for segmentation, and its performance was compared with 3D U-net and 3D V-net models. Segmentation results were evaluated using the Dice coefficient and Jaccard index.</div></div><div><h3>Results</h3><div>The 3D Res U-net model demonstrated superior segmentation performance compared to the other deep learning methods. For cartilage in the lateral femorotibial joint, medial femorotibial joint, and patellofemoral joint, the average Dice coefficients with 3D Res U-net were 0.871, 0.860, and 0.858 in internal validation and 0.846, 0.837, and 0.819 in external validation, respectively. The Jaccard index followed a similar trend.</div></div><div><h3>Conclusion</h3><div>The 3D Res U-net model improves knee cartilage segmentation in conventional MR imaging, contributing to the understanding of cartilage morphology and the improvement of clinically relevant decisions.</div></div>","PeriodicalId":56110,"journal":{"name":"Knee","volume":"58 ","pages":"Article 104270"},"PeriodicalIF":2.0,"publicationDate":"2025-11-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145530716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Knee最新文献

Background

Methods

Results

Conclusions

Background

Methods

Results

Conclusions

Background

Purpose

Methods

Results

Conclusion

Background

Methods

Results

Conclusions

Background

Methods

Results

Conclusions

Background

Methods

Results

Conclusions

Background

Methods

Results

Conclusion

Background

Methods

Results

Conclusion