| 506 | 4 | 404 |
| 下载次数 | 被引频次 | 阅读次数 |
目的：探讨基于人工智能（AI）深度学习算法的CT-MRI多模态影像自动融合分割技术，开发并构建前交叉韧带重建术（ACLR）术前自动规划系统的可行性，以实现ACLR高效准确的术前规划。方法：通过中国人民解放军总医院第四医学中心影像中心得到2023年4月至2024年1月就诊的200例前交叉韧带（ACL）、后交叉韧带（PCL）及半月板正常的膝关节疼痛患者的膝关节CT和MRI影像，由运动医学专业医师对骨皮质、ACL、PCL、半月板等结构进行手工标注，并使用AI深度学习算法对标注图像进行学习，构建CT-MRI多模态影像自动融合分割系统。基于CT-MRI配准融合图像，再次使用AI深度学习技术，强化ACLR股骨、胫骨骨道内外口关键点位的识别，构建ACLR术前自动规划系统。招募12例ACL损伤患者并使用其CT影像，3D打印技术打印其假骨模型并使用ACLR术前自动规划系统对胫骨及股骨骨道位置进行规划，以此为依据在假骨模型上钻取骨道，并对股骨和胫骨骨道长度、关节腔内间距的差异进行统计学分析。结果：CT及MRI多模态影像融合分割后可形成包含骨骼及软组织结构的个体化3D膝关节模型，多模态影像融合精度Dice指数为0.864。ACLR术前自动规划系统进行术前规划的平均时间为（3.0±0.5）min。假骨模拟手术中股骨骨道、胫骨骨道长度及关节腔内间距与术前规划的差异均无统计学意义（P均>0.05）。结论：基于AI深度学习的CT-MRI多模态影像自动融合分割技术的ACLR术前自动规划系统更为智能、快速、精准，可显著提高ACLR术前规划能力，有望降低ACLR术后并发症，提升手术效果。
Abstract：Objective: To explore the feasibility of developing and constructing an automatic preoperative planning system for anterior cruciate ligament reconstruction(ACLR) based on artificial intelligence(AI) deep learning algorithm and multimodal CT-MRI images automatic fusion and segmentation, in order to achieve efficient and accurate preoperative planning for ACLR. Methods:CT and MRI images of the knee joints from 200 patients with normal anterior cruciate ligament(ACL), posterior cruciate ligaments(PCL), and meniscus admitted from April 2023 to January 2024 were obtained from the imaging center of the Fourth Medical Center of Chinese PLA General Hospital. Sports medicine specialists manually annotated structures such as the cortical bone, ACL, PCL, and meniscus,and used AI deep learning algorithm to learn the annotated images to build a multimodal CT-MRI image automatic fusion and segmentation system. Based on the CT-MRI registration and fusion images, AI deep learning technology was further applied to enhance the identification of key points at the entry and exit of the femoral and tibial tunnels in ACLR, constructing an automatic preoperative planning system for ACLR. Twelve patients with ACL injuries were recruited, and their CT images were used to guide the creation of pseudo bone models using 3D printing technology. The ACLR automatic preoperative planning system was used to plan the positions of tibial and femoral tunnel. which were then drilled on the pseudo bone models. Statistical analysis was performed to assess the differences in the lengths of the femoral and tibial tunnel and intra-articular spacing. Results: The CT-MRI multimodal image fusion and segmentation generated an individualized 3D knee joint model containing bone and soft tissue structures, with the Dice index of multimodal image fusion accuracy was 0.864. The mean time for preoperative planning using the ACLR automatic preoperative planning system was(3.0±0.5) min. In the pseudo-bone experiments, the differences between the planned and actual lengths of the femoral and tibial tunnels, as well as the intra-articular spacing, were not statistically significant(all P>0.05). Conclusions: The ACLR automatic preoperative planning system based on AI deep learning and multimodal CT-MRI images fusion and segmentation is more intelligent, rapid, and more accurate. It significantly enhances the preoperative planning ability of ACLR, and potentially reducing postoperative complications and revision rate, and improving surgical outcomes.
[1] Whittaker JL, Losciale JM, Juhl CB, et al. Risk factors for knee osteoarthritis after traumatic knee injury:a systematic review and meta-analysis of randomised controlled trials and cohort studies for the OPTIKNEE Consensus[J]. Br J Sports Med, 2022, 56(24):1406-1421.
[2] Liu Y, Li C, Ma N, et al. Proprioceptive and clinical outcomes after remnant preserved anterior cruciate ligament reconstruction:assessment with minimal confounding factors[J]. Orthop Surg, 2022, 14(1):44-54.
[3] Hong IS, Pierpoint LA, Hellwinkel JE, et al. Clinical outcomes after ACL reconstruction in soccer(football, futbol)players:a systematic review and meta-analysis[J]. Sports Health, 2023, 15(6):788-804.
[4] Hayb?ck G, Raas C, Rosenberger R. Failure rates of common grafts used in ACL reconstructions:a systematic review of studies published in the last decade[J]. Arch Orthop Trauma Surg, 2022, 142(11):3293-3299.
[5]王云鹭,李锡勇,刘伦,等.前交叉韧带重建术后移植物愈合的研究进展[J].实用骨科杂志, 2023, 29(1):47-51.
[6] Diquattro E, Jahnke S, Traina F, et al. ACL surgery:reasons for failure and management[J]. EFORT Open Rev, 2023, 8(5):319-330.
[7]陆莉霞,邹俊忠,郭玉成,等.多模态融合的膝关节损伤预测[J].计算机工程与应用, 2021, 57(9):225-232.
[8] Azam MA, Khan KB, Salahuddin S, et al. A review on multimodal medical image fusion:compendious analysis of medical modalities, multimodal databases, fusion techniques and quality metrics[J]. Comput Biol Med, 2022(144):105253.
[9] Fang D, Yang C, Zhou H, et al. Effects of CT/MRI image fusion on cerebrovascular protection, postoperative complications, and limb functional recovery in patients with anterior and middle skull base tumors:based on a retrospective cohort study[J]. Contrast Media Mol Imaging, 2022, 2022:7855576.
[10]靳能皓,孟凡皓,乔波,等.多模态图像自动配准融合技术联合手术机器人在面侧深区肿瘤诊疗中的应用[J].中国医学影像学杂志, 2023, 31(6):572-576.
[11]吴东,刘星宇,张逸凌,等.人工智能辅助全髋关节置换术三维规划系统的研发及临床应用研究[J].中国修复重建外科杂志, 2020, 34(9):1077-1084.
[12] Gurung B, Liu P, Harris PDR, et al. Artificial intelligence for image analysis in total hip and total knee arthroplasty:a scoping review[J]. Bone Joint J, 2022, 104-B(8):929-937.
[13] Pearle AD, McAllister D, Howell SM. Rationale for strategic graft placement in anterior cruciate ligament reconstruction:IDEAL femoral tunnel position[J]. Am J Orthop(Belle Mead NJ), 2015, 44(6):253-258.
[14] Parkar AP, Adriaensen M, Vindfeld S, et al. The anatomic centers of the femoral and tibia l insertions of the anterior cruciate ligament:a systematic review of imaging and cadaveric studies reporting normal center locations[J]. Am J Sports Med, 2017, 45(9):2180-2188.
[15] Lertwanich P, Martins CA, Asai S, et al. Anterior cruciate ligament tunnel position measure ment reliability on 3-dimensional reconstructed computed tomography[J]. Arthroscopy, 2011, 27(3):391-398.
[16] Li X, Yan L, Li D, et al. Failure modes after anterior cruciate ligament reconstruction:a systematic review and metaanalysis[J]. Int Orthop, 2023, 47(3):719-734.
[17] Keuning MC, Robben BJ, Brouwer RW, et al. Young men are at higher risk of failure after ACL hamstring reconstructions:a retrospective multivariate analysis[J]. BMC Musculoskelet Disord, 2022, 23(1):598.
[18] Wang Y, Liu X, Xiang L, et al. Risk factors of anterior cruciate ligament accumulated damage among young male patients undergoing routine exercise[J]. Orthop Surg, 2022, 14(6):1109-1114.
[19]李春宝,刘玉杰,冯勇,等.前十字韧带双束重建的临床生物力学研究进展[J].中华骨科杂志, 2013, 33(12):1240-1247.
[20] Diquattro E, Jahnke S, Traina F, et al. ACL surgery:reasons for failure and management[J]. EFORT Open Rev, 2023, 8(5):319-330.
[21]廖正俭,陈忠仪,刘宇清,等.多模态图像融合联合3D打印在中央区窦镰旁肿瘤手术中的应用效果[J].河南医学研究, 2022, 31(15):2689-2692.
[22] Feng K, Wang T, Tang J, et al. Application of CT-MRI fusion-based three-dimensional reconstruction technique in the anatomic study of posterior cruciate ligament[J]. Orthop Surg, 2022, 14(11):2845-2853.
[23] Hefzy MS, Grood ES, Noyes FR. Factors affecting the region of most isometric femoral attachments. PartⅡ:The anterior cruciate ligament[J]. Am J Sports Med, 1989, 17(2):208-216.
[24] Yang G, Liu D, Zhou G, et al. Robot-assisted anterior cruciate ligament reconstruction based on three-dimensional images[J]. J Orthop Surg Res, 2024, 19(1):246.
[25] Zhang L, Liang Q, Zhao Z, et al. Robot-assisted allepiphyseal anterior cruciate ligament reconstruction in skeletally immature patients:a retrospective study[J]. Int Orthop, 2023, 47(2):429-435.
[26] Cho WJ, Kim JM, Kim DE, et al. Accuracy of the femoral tunnel position in robot-assisted anterior cruciate ligament reconstruction using a magnetic resonance imaging-based navigation system:a preliminary report[J]. Int J Med Robot, 2018, 14(5):e1933.
基本信息:
中图分类号:R687.4;TP18;TP391.41
引用信息:
[1]于浩淼,董继祥,李海鹏,等.基于人工智能深度学习的CT-MRI多模态影像自动融合分割技术在前交叉韧带重建术前规划中的应用[J].中华骨与关节外科杂志,2025,18(01):27-35.