摘要
发育性髋关节发育不良(developmental dysplasia of the hip,DDH)为髋臼和股骨头位置和形态异常的发育性疾病.DDH是儿童最常见的髋部疾病之一,且其发病始于婴幼儿,髋关节解剖结构复杂,临床大规模诊筛工作存在工作量大、误诊率及漏诊率较高以及外科手术有较长的学习曲线等问题.人工智能是用于研究、开发能够模拟人类智能的理论、方法、技术及应用系统的新兴技术科学,主要目标是使机器能够完成需要人类智能的复杂任务.目前人工智能技术在DDH的诊疗方面主要应用于:人工智能结合X线片或超声对DDH进行诊断及筛查,可能一定程度上避免主观性,提高诊断及筛查的准确性;在髋臼周围截骨术和全髋关节置换术的术前规划中,能够更为精确地追踪骨块的位置和设计假体部件的尺寸,使手术规划更加精准;3D导航系统、手术机器人、机械臂等人工智能技术有助于提高手术的准确性和安全性.人工智能具有简便快捷、可重复性高及深度学习能力等特点,一定程度上为DDH诊疗节省了人力物力及时间成本,便于医务工作者共享DDH诊疗经验以及减轻患者、家庭、社会及国家的负担.人工智能在DDH诊疗中的应用日趋广泛,但目前对人工智能在DDH筛查方面仍处于初步阶段,人工智能技术在DDH保髋治疗及其他术式中的应用也鲜有报道,对人工智能在DDH保守治疗、术后并发症和预后预测等方面研究不足.现阶段人工智能在DDH中的应用中仍有一定的局限性,包括图像数据质量难以把控、技术研发及推广难度大、缺少法律法规监管等问题.
Abstract
Developmental dysplasia of the hip(DDH)is a developmental disease with abnormal position and morphology of the acetabulum and femoral head.DDH ranges from mild dysplasia to complete dislocation,including simple acetabular dyspla-sia,subluxation,and total dislocation.DDH is one of the most common hip diseases in children,and its onset begins in infancy.The anatomical structure of the hip joint is complex,and there are some practical problems in the large-scale clinical diagnosis and screening work,such as heavy workload,high misdiagnosis rate and missed diagnosis rate,and a long learning period of sur-gery.Artificial intelligence(AI)is an emerging technology science used to research and develop theories,methods,technologies and application systems that can simulate human intelligence.The main goal is to enable machines to complete complex tasks that require human intelligence.At present,artificial intelligence is mainly used in the diagnosis and treatment of DDH.Artificial intel-ligence combined with X-ray or ultrasound is used to diagnose and screen DDH,which avoids subjectivity to a certain extent and improves the accuracy of diagnosis and screening.In the preoperative planning of periacetabular osteotomy and total hip arthro-plasty,it can more accurately track the position of the bone block and design the size of the prosthesis,which makes the surgical planning more accurate to a certain extent.Artificial intelligence technologies such as 3D navigation system,surgical robot and ro-botic arm are helpful to improve the accuracy and safety of surgery.Artificial intelligence has the characteristics of simplicity,speed,high repeatability and deep learning ability,which to a certain extent saves manpower,material resources and time costs for DDH diagnosis and treatment,facilitates medical workers to share DDH diagnosis and treatment experience and reduces the bur-den of patients,families,society and the country.The application of artificial intelligence in DDH diagnosis and treatment is be-coming more and more extensive,but at present,artificial intelligence in DDH screening is still in its preliminary stage.There are few reports on artificial intelligence technology in DDH hip preservation treatment and other surgical methods,and there is insuffi-cient research on artificial intelligence in DDH conservative treatment,postoperative complications and prognosis prediction.It is worth exploring new ideas by researchers.At present,the application of artificial intelligence in DDH still has certain limitations,including problems such as difficult to control the quality of image data,difficult to develop and promote technology,and lack of laws and regulations.Based on the literature database and relevant data sharing network,this paper reviews the application of arti-ficial intelligence in the diagnosis and treatment of DDH at home and abroad in recent years,summarizes the current application status of artificial intelligence in the diagnosis and treatment of DDH,and provides new ideas for the future application of artificial intelligence in the diagnosis and treatment of DDH.
维普
万方数据