He Guangzhi, Li Zhou, Zhu Zhiyuan, Han Tong, Cao Yan, Chen Chaoyu, Huang Yuhao, Dou Haoran, Liang Lianying, Zhang Fangmei, Peng Jin, Tan Tao, Liu Hongmei, Yang Xin, Ni Dong
Jinan University, Guangzhou, Guangdong, China.
Department of Ultrasound, Shenzhen Guangming District People's Hospital, Songbai Road, Matian Street, Shenzhen, Guangdong, China.
BMC Pregnancy Childbirth. 2025 Jan 7;25(1):10. doi: 10.1186/s12884-024-07108-4.
Early diagnosis of cleft lip and palate (CLP) requires a multiplane examination, demanding high technical proficiency from radiologists. Therefore, this study aims to develop and validate the first artificial intelligence (AI)-based model (CLP-Net) for fully automated multi-plane localization in three-dimensional(3D) ultrasound during the first trimester.
This retrospective study included 418 (394 normal, 24 CLP) 3D ultrasound from 288 pregnant woman between July 2022 to October 2024 from Shenzhen Guangming District People's Hospital during the 11-13 weeks of pregnancy. 320 normal volumes were used for training and validation, while 74 normal and 24 CLP volumes were used for testing. Two experienced radiologists reviewed three standard lip and palate planes (mid sagittal, retronasal triangle, and maxillary axial planes) as ground truth (GT) and the CLP-Net was developed to locate these planes.
In normal test set, mean angle(± SD)° and distance(± SD)mm differences were 6.24 ± 4.83, 9.81 ± 5.48, 15.36 ± 18.14 and 0.86 ± 0.72, 1.36 ± 1.15, 1.96 ± 2.35 for MSP ± SD, RTP ± SD and MAP ± SD, NCC and SSIM were 0.931 ± 0.079, 0.819 ± 0.122, 0.781 ± 0.157 and 0.896 ± 0.058, 0.785 ± 0.076, 0.726 ± 0.088 respectively. In the CLP cases, there were 8.61 ± 5.52, 10.67 ± 5.08, 16.91 ± 17.42 and 1.03 ± 1.20, 1.17 ± 1.08, 1.34 ± 0.95 for mean angle and distance in MSP, RTP, and MAP, respectively. NCC and SSIM were 0.876 ± 0.104, 0.803 ± 0.084, 0.793 ± 0.089 and 0.841 ± 0.105, 0.812 ± 0.085, 0.764 ± 0.100, respectively. CLP-Net predictions had a highly visual acceptance rate among radiologists (MSP: 95%, RTP: 70%, MAP: 70%), with improved localization speed 15s(31.3%) for senior radiologists and 63s(38.9%) for junior radiologists.
CLP-Net accurately locates three planes for CLP screening, aiding radiologists and enhancing the efficiency of ultrasound examinations.
唇腭裂(CLP)的早期诊断需要多平面检查,这对放射科医生的技术水平要求很高。因此,本研究旨在开发并验证首个基于人工智能(AI)的模型(CLP-Net),用于在孕早期对三维(3D)超声进行全自动多平面定位。
这项回顾性研究纳入了2022年7月至2024年10月期间深圳市光明区人民医院288名孕妇在妊娠11-13周时的418例3D超声检查(394例正常,24例CLP)。320例正常容积用于训练和验证,74例正常容积和24例CLP容积用于测试。两名经验丰富的放射科医生将三个标准的唇腭裂平面(正中矢状面、鼻后三角平面和上颌轴位平面)作为金标准(GT)进行评估,同时开发CLP-Net来定位这些平面。
在正常测试组中,正中矢状面(MSP)、鼻后三角平面(RTP)和上颌轴位平面(MAP)的平均角度(±标准差)°差异分别为6.24±4.83、9.81±5.48、15.36±18.14,平均距离(±标准差)mm差异分别为0.86±0.72、1.36±1.15、1.96±2.35,归一化互相关系数(NCC)分别为0.931±0.079、0.819±0.122、0.781±0.157,结构相似性指数(SSIM)分别为0.896±0.058、0.785±0.076、0.726±0.088。在CLP病例中,MSP、RTP和MAP的平均角度分别为8.61±5.52、10.67±5.08、16.91±17.42,平均距离分别为1.03±1.20、1.17±1.08、1.34±0.95。NCC分别为0.876±0.104、0.803±0.084、0.793±0.089,SSIM分别为0.841±0.105、0.812±0.085、0.764±0.100。CLP-Net的预测结果在放射科医生中具有很高的视觉接受率(MSP:95%,RTP:70%,MAP:70%),资深放射科医生的定位速度提高了15秒(31.3%),初级放射科医生提高了63秒(38.9%)。
CLP-Net能准确地定位用于CLP筛查的三个平面,有助于放射科医生并提高超声检查的效率。