Application of MRI radiomics in classification diagnosis of anterior cruciate ligament injury
-
摘要:
目的 分析MRI影像组学对前交叉韧带损伤分级的诊断价值。 方法 回顾性选择2017~2022年滨州医学院附属医院放射科质子加权成像提示前交叉韧带异常信号患者212例，以关节镜检查结果为金标准，将患者分为重度损伤组（n=141）和轻度损伤组（n=71）。提取前交叉韧带图像的组学特征包括形状特征、一阶特征、纹理特征和小波特征。通过SMOTE法进行过采样，解决数据不均衡问题。通过组内及组间相关系数进行一致性检验。在两组中通过7:3的比例随机分割为训练集和测试集。并用LASSO算法筛选出最佳影像组学特征，采用Logistic回归建立组学模型。基于影像组学特征建立影像组学模型，基于两组的临床参数建立临床模型，以及联合两者建立Nomogram模型。分别在训练集及测试集绘制ROC曲线，计算敏感度、特异性、准确度评估模型的诊断效能，通过绘制校准曲线来评估模型预测值和实际观测值之间的差异，通过绘制临床决策曲线分析评价其临床有效性。 结果 通过特征提取在质子加权成像横断位、冠状位、矢状位共获得2553个特征，通过特征筛选及降维最终保留12个特征参数。影像组学模型训练集的曲线下面积（AUC）为0.9105，测试组为0.8561。性别、关节不稳、关节交锁以及组学积分是诊断前交叉韧带重度损伤的最佳特征集。临床模型训练集的AUC值为0.6989，测试组为0.6415。Nomogram模型训练集的AUC值为0.9449，测试组为0.8661。Nomogram模型与临床模型差异有统计学意义（P < 0.05），Nomogram的AUC高于组学模型, 但在测试集差异无统计学意义（P>0.05）。 结论 基于MRI的影像组学方法可为诊断前交叉韧带损伤分级提供一种新型的检测手段，使前交叉韧带损伤的临床诊断准确率得到很大提升，Nomogram模型比临床模型以及组学模型的诊断效能更好。 Abstract:Objective To analyze the diagnostic value of MRI radiomics in the classification of anterior cruciate ligament (ACL) injury. Methods A total of 212 patients with abnormal signals of the ACL suggested by proton density weighted imaging in the Department of Radiology of the Affiliated Hospital of Binzhou Medical University from 2017 to 2022 were reviewed. The parents were divided into severe injury group (n=141) and mild injury group (n=71) based on arthroscopic findings, which was the gold standard in the classification of ACL injury. The radiomics features of the ACL images were extracted including shape features, first order features, texture features and wavelet features. The imbalance problem of data was solved by performing oversampling using SMOTE method. Consistency test was performed by intra-group and inter-group correlation coefficients. Patients were randomly divided into a training set and a test set at 7:3. The best radiomics features were selected by LASSO algorithm, and the radiomics model was established by Logistic regression. The radiomics model was established based on the radiomics features and the clinical model was established based on the clinical parameters of the two groups. The Nomogram model was established by the radiomics features and the clinical parameters as mentioned above. The diagnostic efficacy of the model was evaluated by ROC curves, sensitivity, specificity and accuracy in the training set and the test set. The difference between the model's predicted values and the actual observed values was evaluated by the calibration curve, and the clinical efficacy was evaluated by the clinical decision curve analysis. Results A total of 2553 features were obtained from proton density weighted imaging transverse, coronal and sagittal positions through feature extraction, and 12 feature parameters were retained through feature filtering and dimensionality reduction. The area under the curve (AUC) was 0.9105 for the training set of the radiomics model and 0.8561 for the test set. The best feature sets of diagnosing severe ACL injury were gender, joint instability, joint interlock and radiomics. In the clinical model, the AUC values of the training set was 0.6989 and the test set was 0.6415. In the Nonogram model, the AUC values of the training set was 0.9449 and the test set was 0.8661. The difference between the Nomogram model and the clinical model was statistically significant (P < 0.05). The AUC of the Nomogram was higher than that of the radiomics model, but the difference was not statistically significant in the test set(P> 0.05). Conclusion The radiomics method based on MRI images can provide a new diagnostic method in classification of ACL injury which can greatly improve the clinical diagnosis accuracy of ACL injury. Furthermore, the diagnostic efficiency of the Nomogram model is the best compared with the clinical model and the radiomics model. -
Key words:
- MRI /
- radiomics /
- anterior cruciate ligament injury /
- classification
-
图 1 LASSO回归模型分析选择诊断ACL损伤的特征
Figure 1. LASSO regression model analysis of the selected characteristics for the diagnosis of ACL injury. A: Plot of changes in characteristics chosen by LASSO through adjusting the parameterλ; B: The change plots of tuning parameterλchosen by 10-fold cross validation.
图 2 影像组学特征及其对应的系数所占比重
Figure 2. Proportion of imaging omics characteristics and their corresponding coefficients.
图 3 影像组学模型诊断ACL损伤的ROC曲线
Figure 3. ROC curves of the ACL injury diagnosed the radiomics model.
图 4 Nomogram模型诊断ACL损伤的ROC曲线
Figure 4. ROC curves of the Nomogram model for diagnostic ACL injury.
图 5 融合影像组学特征及临床特征的Nomogram
Figure 5. Nomogram fusion of radiomics features and clinical features.
图 6 临床特征诊断ACL损伤的ROC曲线
Figure 6. ROC curves of clinical features for diagnosed ACL injury.
图 7 Nomogram在训练集和测试集的校准曲线
Figure 7. The calibration curves of Nomogram in the training and test sets. A: Training set; B: Test set.
图 8 Nomogram、影像组学模型及临床模型的决策曲线
Figure 8. Decision curves of the Nomogram, radiomics model and clinical model.
表 1 临床模型与Nomogram诊断效能对比（De-long检验）
Table 1. Comparison of diagnostic efficacy between clinical model and Nomogram (De-long test).
Groups AUCClinical AUCNomogram Z P Training group 0.6969 0.9449 8.7696 < 0.01 Test group 0.6415 0.8661 5.0182 < 0.01 
下载: 导出CSV
表 2 影像组学模型与Nomogram诊断效能对比（De-long检验）
Table 2. Comparison of diagnostic efficacy between radiomics model and Nomogram (De-long test).
Groups AUCRadiomics AUCNomogram Z P Training group 0.9105 0.9449 2.7476 0.006 Test group 0.8561 0.8661 0.57432 0.5658
下载: 导出CSV
-
[1] 李长树, 李政, 李雳, 等. 膝前交叉韧带断裂合并内侧副韧带Ⅲ度急性损伤一期联合修复重建的短期临床疗效研究[J]. 中国中医骨伤科杂志, 2023, 31(2): 18-23. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGZG202302004.htm [2] Kokkotis C, Moustakidis S, Tsatalas T, et al. Leveraging explainable machine learning to identify gait biomechanical parameters associated with anterior cruciate ligament injury[J]. Sci Rep, 2022, 12(1): 6647. doi: 10.1038/s41598-022-10666-2 [3] 金伟林, 程亚栋, 曾冠楠, 等. 前交叉韧带断裂损伤后并发关节软骨损伤的相关因素研究[J]. 创伤外科杂志, 2020, 22(8): 588-91. https://www.cnki.com.cn/Article/CJFDTOTAL-CXWK202008026.htm [4] 向珊珊, 傅明, 张志奇, 等. 关节镜下自体或同种异体肌腱移植修复急性和陈旧性前交叉韧带损伤: 重建效果评价[J]. 中国组织工程研究, 2018, 22(26): 4174-8. doi: 10.3969/j.issn.2095-4344.0939 [5] Chen DS, Wang TF, Zhu JW, et al. A novel application of unsupervised machine learning and supervised machine learningderived radiomics in anterior cruciate ligament rupture[J]. Risk Manag Healthc Policy, 2021, 14: 2657-64. doi: 10.2147/RMHP.S312330 [6] Awan MJ, Mohd Rahim MS, Salim N, et al. Machine learningbased performance comparison to diagnose anterior cruciate ligament tears[J]. J Healthc Eng, 2022, 2022: 2550120. [7] 王晓晖, 吕晓烁, 刘展, 等. 基于影像组学构建识别临床高风险颈动脉斑块的机器学习模型[J]. 中国胸心血管外科临床杂志, 2024, 31(1): 24-34. https://www.cnki.com.cn/Article/CJFDTOTAL-ZXYX202401004.htm [8] 刘广銮, 郭宗磊, 戈进, 等. 前交叉韧带损伤伴发内侧半月板后根撕裂: 解剖学的危险因素分析[J]. 中国组织工程研究, 2023, 27 (5): 663-8. https://www.cnki.com.cn/Article/CJFDTOTAL-XDKF202424021.htm [9] Kim HJ, Lee JH, Lee DH. Proprioception in patients with anterior cruciate ligament tears: a meta-analysis comparing injured and uninjured limbs[J]. Am J Sports Med, 2017, 45(12): 2916-22. doi: 10.1177/0363546516682231 [10] Samuelsen BT, Webster KE, Johnson NR, et al. Hamstring autograft versus patellar tendon autograft for ACL reconstruction: is there a difference in graft failure rate? A meta-analysis of 47, 613 patients[J]. Clin Orthop Relat Res, 2017, 475(10): 2459-68. doi: 10.1007/s11999-017-5278-9 [11] Nasseri A, Khataee H, Bryant AL, et al. Modelling the loading mechanics of anterior cruciate ligament[J]. Comput Methods Programs Biomed, 2020, 184: 105098. doi: 10.1016/j.cmpb.2019.105098 [12] O'Sullivan O, Ladlow P, Steiner K, et al. Knee MRI biomarkers associated with structural, functional and symptomatic changes at least a year from ACL injury - A systematic review[J]. Osteoarthr Cartil Open, 2023, 5(3): 100385. doi: 10.1016/j.ocarto.2023.100385 [13] 段寿生, 陈用国. MRI对膝关节交叉韧带损伤的诊断价值研究[J]. 浙江创伤外科, 2023, 28(6): 1154-6. doi: 10.3969/j.issn.1009-7147.2023.06.047 [14] 梁继华, 梁小瑜. 磁共振成像技术诊断膝关节前交叉韧带断裂的准确率分析[J]. 影像研究与医学应用, 2022, 6(24): 173-5. https://www.cnki.com.cn/Article/CJFDTOTAL-YXYY202224057.htm [15] 王海杰, 高明勇, 涂滨. 超声与MRI检查对前交叉韧带损伤诊断的对比研究[J]. 中国超声医学杂志, 2020, 36(7): 651-3. doi: 10.3969/j.issn.1002-0101.2020.07.025 [16] 孙小煜, 宋璐, 张雪峰, 等. 应用MRI、MSCT检查在外伤性膝关节交叉韧带损伤诊治中临床价值研究[J]. 中国CT和MRI杂志, 2021, 19(9): 180-2. https://www.cnki.com.cn/Article/CJFDTOTAL-CTMR202109059.htm [17] 王忠, 文明, 刘成渝, 等. 高分辨率CT诊断膝关节半月板及交叉韧带损伤的价值观察[J]. 中国CT和MRI杂志, 2022, 20(12): 153-5. https://www.cnki.com.cn/Article/CJFDTOTAL-CTMR202212063.htm [18] 向作林, 吴国彪, 谭必会, 等. 磁共振成像技术在膝关节前交叉韧带损伤中的应用[J]. 现代医用影像学, 2023, 32(7): 1234-6, 1250. https://www.cnki.com.cn/Article/CJFDTOTAL-XDYY202307010.htm [19] 王文河, 陈伟, 李克军. 前交叉韧带损伤的MRI征象分析[J]. 影像研究与医学应用, 2020, 4(7): 53-4. https://www.cnki.com.cn/Article/CJFDTOTAL-YXYY202007029.htm [20] Van DP. Pitfalls in MR imaging of acute anterior cruciate ligament injuries[J]. Knee Surg Sports Traumatol Arthrosc, 2020, 28(11): 3622-4. [21] 柏瑞, 黄俊成, 何小华, 等. 低剂量双能量CT诊断膝关节前交叉韧带损伤的临床对照分析[J]. 中国CT和MRI杂志, 2022, 20 (10): 157-9. https://www.cnki.com.cn/Article/CJFDTOTAL-CTMR202210061.htm [22] 牛金榜. 健美操运动员前交叉韧带损伤的康复方案研究: Ⅰ度损伤的保守治疗[J]. 体育科技文献通报, 2020, 28(2): 150-1, 154. https://www.cnki.com.cn/Article/CJFDTOTAL-SPOR202002078.htm [23] 谢羽婕, 季玉秀, 张驰. 前交叉韧带损伤诊治策略[J]. 西南医科大学学报, 2023, 46(6): 483-8. https://www.cnki.com.cn/Article/CJFDTOTAL-LXYB202306005.htm [24] Stone AV, Marx S, Conley CW. Management of partial tears of the anterior cruciate ligament: a review of the anatomy, diagnosis, and treatment[J]. J Am Acad Orthop Surg, 2021, 29(2): 60-70. [25] Carulli C, Innocenti M, Roselli G, et al. Partial rupture of anterior cruciate ligament: preliminary experience of selective reconstruction[J]. J Orthop Traumatol, 2020, 21(1): 5. [26] 黄瑜亮, 黄斌, 严俊. 结直肠癌的早期诊断技术进展[J]. 分子影像学杂志, 2019, 42(1): 77-80. doi: 10.12122/j.issn.1674-4500.2019.01.18 [27] 杨剑锋, 乔佩蕊, 李永梅, 等. 机器学习分类问题及算法研究综述[J]. 统计与决策, 2019, 35(6): 36-40. https://www.cnki.com.cn/Article/CJFDTOTAL-TJJC201906009.htm -
点击查看大图
计量
- 文章访问数: 189
- HTML全文浏览量: 69
- PDF下载量: 25
- 被引次数: 0
