MobileNetV4：面向产前超声的主动脉弓分支异常智能诊断研究

doi:10.3877/cma.j.issn.1672-6448.2025.08.005

中华医学超声杂志（电子版） ›› 2025, Vol. 22 ›› Issue (08) : 711 -720. doi: 10.3877/cma.j.issn.1672-6448.2025.08.005

妇产科超声影像学

MobileNetV4：面向产前超声的主动脉弓分支异常智能诊断研究

陈明朗¹, 许凯², 黄稚熙³, 梁博诚⁴, 贺杰¹, 黄海珊⁵, 马微波⁶, 谭莹⁴, 邹志英⁴, 刘晓棠⁷, 彭桂艳⁷, 陈家希¹, 钟晓红⁸^,^†(

)

¹　543002　梧州学院广西机器视觉与智能控制重点实验室
²　650091　昆明，云南大学软件学院
³　518028　深圳，南方医科大学妇女儿童医学中心深圳市妇幼保健院新生儿科
⁴　518028　深圳，南方医科大学妇女儿童医学中心深圳市妇幼保健院超声科
⁵　519000　珠海，中山大学软件工程学院
⁶　201318　上海健康医学院护理与健康管理学院
⁷　518052　深圳，南山区妇幼保健院超声科
⁸　361000　厦门市妇幼保健院超声医学科

收稿日期:2025-05-28 出版日期:2025-08-01
通信作者: 钟晓红
基金资助:
国家自然科学基金(62162054); 广西自然科学基金项目(2025GXNSFAA069497，2025GXNSFAA069688); 梧州学院校级科研项目青年项目(2023C004); 厦门市医疗卫生指导性项目(3502Z20244ZD1224)

MobileNetV4: an intelligent tool for diagnosis of aortic arch branching anomalies based on prenatal ultrasound images

Minglang Chen¹, Kai Xu², Zhixi Huang³, Bocheng Liang⁴, Jie He¹, Haishan Huang⁵, Weibo Ma⁶, Ying Tan⁴, Zhiying Zou⁴, Xiaotang Liu⁷, Guiyan Peng⁷, Jiaxi Chen¹, Xiaohong Zhong⁸^,^†()

¹ Guangxi Key Laboratory of Machine Vision and Intelligent Control, Wuzhou University, Wuzhou 543002, China
² School of Software, Yunnan University, Kunming 650091, China
³ Department of Neonatology,Shenzhen Maternity and Child Healthcare Hospital, Women and Children's Medical Center, Southern Medical University, Shenzhen 518028, China
⁴ Department of Ultrasound, Shenzhen Maternity and Child Healthcare Hospital, Women and Children's Medical Center, Southern Medical University, Shenzhen 518028, China
⁵ School of Software Engineering, Sun Yat-sen University, Zhuhai 519000, China
⁶ School of Nursing and Health Management, Shanghai University of Medicine and Health Sciences, Shanghai 201318, China
⁷ Department of Ultrasound, Nanshan Maternity and Child Healthcare Hospital, Shenzhen 518052, China
⁸ Department of Ultrasound, Xiamen Maternity & Child Healthcare Hospital, Xiamen 361000, China

Received:2025-05-28 Published:2025-08-01
Corresponding author: Xiaohong Zhong

全文 ( ) 下载PDF ( ) 导出引用

引用本文：

陈明朗, 许凯, 黄稚熙, 梁博诚, 贺杰, 黄海珊, 马微波, 谭莹, 邹志英, 刘晓棠, 彭桂艳, 陈家希, 钟晓红. MobileNetV4：面向产前超声的主动脉弓分支异常智能诊断研究[J/OL]. 中华医学超声杂志（电子版）, 2025, 22(08): 711-720.

Minglang Chen, Kai Xu, Zhixi Huang, Bocheng Liang, Jie He, Haishan Huang, Weibo Ma, Ying Tan, Zhiying Zou, Xiaotang Liu, Guiyan Peng, Jiaxi Chen, Xiaohong Zhong. MobileNetV4: an intelligent tool for diagnosis of aortic arch branching anomalies based on prenatal ultrasound images[J/OL]. Chinese Journal of Medical Ultrasound (Electronic Edition), 2025, 22(08): 711-720.

目的

探讨MobileNetV4模型在产前超声图像中识别主动脉弓及其分支异常的应用价值。

方法

回顾性收集2019年3月至2024年3月在深圳市妇幼保健院确诊的血管环产前超声图像，共计12 284张，其中血管环异常图像8913张，正常对比图像为3371张，按疾病类型以8∶1∶1比例将图像划分为训练集、测试集和验证集。构建基于MobileNetV4的深度学习模型，用于主动脉弓分支异常的自动诊断，并与MSPAnet-50、MedMamba、RepVit、ResNet-50和ViT 5种不同应用场景的主流模型进行性能对比。模型评估指标为准确性、敏感度、特异度、精确度、ROC曲线下面积（AUC）及F1分数。通过绘制ROC曲线、混淆矩阵、Grad-CAM热力图和t-SNE特征可视化图，从多个维度评估模型在血管环畸形影像诊断中的综合表现。

结果

在双主动脉弓（DAA）、右主动脉弓（RAA）和右锁骨下动脉迷走（ARSA）三类血管环及正常图像的诊断任务中，MobileNetV4模型均表现出优越性能，其准确性、特异度、敏感度、精确度、AUC和F1分数分别为94.10%、98.03%、94.09%、94.21%、99.12%和94.04%。在热力图中，只有MobileNetV4模型的关注区域高度集中于主动脉弓及其分支等关键解剖结构，显示出明确的关注区域和清晰边界。混淆矩阵显示MobileNetV4模型的预测结果在所有类别上均呈现明显的对角线聚集，表明各类别的诊断任务中MobileNetV4模型的整体准确性较高，尤其在正常对照样本上实现100%的准确性。ROC曲线显示，MobileNetV4模型识别DAA、RAA、ARSA及正常影像的AUC值分别为99.49%、98.47%、100%、98.51%，优于其他对比模型。MobileNetV4模型在t-SNE显示的特征分布中呈现出更清晰的聚类边界。

结论

MobileNetV4模型在主动脉弓分支异常的超声影像智能识别中表现优异，可为产前超声检查提供有效的技术支持，推动血管环异常的智能化筛查与流程标准化。

关键词: 血管环, 产前超声, MobileNet, 人工智能

Objective

To evaluate the value of the MobileNetV4 model in identifying aortic arch and its branch anomalies in prenatal ultrasound images.

Methods

A total of 12 284 prenatal ultrasound images of vascular rings diagnosed at Shenzhen Maternity and Child Healthcare Hospital from March 2019 to March 2024 were retrospectively collected, including 8913 images with vascular ring anomalies and 3371 normal control images. The images were divided into training, testing, and validation sets at a ratio of 8∶1∶1 according to disease type. A deep learning model based on MobileNetV4 was constructed for the automatic diagnosis of aortic arch branch anomalies, and its performance was compared with five mainstream models applied in different scenarios, including MSPAnet-50, MedMamba, RepVit, ResNet-50, and ViT. Evaluation metrics included accuracy, sensitivity, specificity, precision, area under the receiver operating characteristic (ROC) curve (AUC), and F1 score. The model's comprehensive performance in diagnosing vascular ring anomalies was assessed from multiple perspectives by plotting ROC curves, confusion matrices, Grad-CAM heatmaps, and t-SNE feature visualizations.

Results

In the diagnostic tasks involving double aortic arch (DAA), right aortic arch (RAA), aberrant right subclavian artery (ARSA), and normal images, the MobileNetV4 model demonstrated superior performance, with accuracy, specificity, sensitivity, precision, AUC, and F1 score reaching 94.10%, 98.03%, 94.09%, 94.21%, 99.12%, and 94.04%, respectively. In the heatmaps, only the MobileNetV4 model showed highly concentrated attention on key anatomical structures such as the aortic arch and its branches, exhibiting clearly defined attention regions and sharp boundaries. The confusion matrix revealed that the MobileNetV4 model's predictions were strongly clustered along the diagonal for all categories, indicating high overall diagnostic accuracy, with 100% accuracy achieved on normal control samples. The ROC curves showed that the AUC values for DAA, RAA, ARSA, and normal images were 99.49%, 98.47%, 100%, and 98.51%, respectively, outperforming other comparison models. In the t-SNE visualization, the feature distribution of MobileNetV4 presented more distinct clustering boundaries.

Conclusion

The MobileNetV4 model demonstrates excellent performance in the intelligent recognition of aortic arch branch anomalies in ultrasound images, providing effective technical support for prenatal ultrasound examinations and promoting the intelligent screening and standardized workflow of vascular ring anomalies.

Key words: Vascular ring, Prenatal ultrasound, MobileNet, Artificial intelligence

表1 不同血管环疾病超声图像及正常对照图像数据集分布

数据集	DAA	RAA	ARSA	正常	总计
训练集	2588	2615	1926	2696	9825
验证集	325	328	242	338	1233
测试集	323	326	240	337	1226
总计	3236	3269	2408	3371	12284

表1 不同血管环疾病超声图像及正常对照图像数据集分布

数据集	DAA	RAA	ARSA	正常	总计
训练集	2588	2615	1926	2696	9825
验证集	325	328	242	338	1233
测试集	323	326	240	337	1226
总计	3236	3269	2408	3371	12284

图1 血管环疾病及正常对照超声图像。图a、b分别为右锁骨下动脉迷走血管环彩色多普勒超声和灰阶超声图像；图c、d分别为双主动脉弓血管环彩色多普勒超声和灰阶超声图像；图e、f分别为右主动脉弓血管环彩色多普勒超声和灰阶超声图像；图g、h分别为正常对照彩色多普勒超声和灰阶超声图像

表2 不同诊断模型训练策略参数对比

模型名称	优化器	初始学习率	Epoch	影像尺寸（像素）
MSPAnet-50	SGD	10^-1	200	224×224
MedMamba	AdamW	10^-3	200	640×640
RepVit	AdamW	10^-3	300	224×224
ResNet-50	Adam	10^-3	200	224×224
ViT	Adam	10^-3	200	224×224
MobileNetV4	AdamW	10^-3	300	640×640

表2 不同诊断模型训练策略参数对比

模型名称	优化器	初始学习率	Epoch	影像尺寸（像素）
MSPAnet-50	SGD	10^-1	200	224×224
MedMamba	AdamW	10^-3	200	640×640
RepVit	AdamW	10^-3	300	224×224
ResNet-50	Adam	10^-3	200	224×224
ViT	Adam	10^-3	200	224×224
MobileNetV4	AdamW	10^-3	300	640×640

表3 各模型在血管环标准切面诊断任务中的总体性能表现（%）

模型名称	准确性	敏感度	特异度	AUC	精确度	F1分数	第一命中率
MSPAnet-50	85.29	85.42	95.09	96.48	85.57	85.43	86.10
MedMamba	85.92	85.92	95.32	96.92	86.40	86.13	85.37
RepVit	83.57	83.57	94.51	94.55	83.65	83.44	83.59
ResNet-50	87.09	87.09	95.61	97.29	86.47	86.60	86.74
ViT	90.55	90.54	96.79	98.43	90.27	90.39	90.38
MobileNetV4	94.10	94.09	98.03	99.12	94.21	94.04	94.10

表3 各模型在血管环标准切面诊断任务中的总体性能表现（%）

模型名称	准确性	敏感度	特异度	AUC	精确度	F1分数	第一命中率
MSPAnet-50	85.29	85.42	95.09	96.48	85.57	85.43	86.10
MedMamba	85.92	85.92	95.32	96.92	86.40	86.13	85.37
RepVit	83.57	83.57	94.51	94.55	83.65	83.44	83.59
ResNet-50	87.09	87.09	95.61	97.29	86.47	86.60	86.74
ViT	90.55	90.54	96.79	98.43	90.27	90.39	90.38
MobileNetV4	94.10	94.09	98.03	99.12	94.21	94.04	94.10

图2 各模型诊断血管环疾病依据的解剖结构位置Grad-CAM热力图。图a~d、图e~h、图i~l、图m~p、图q~t、图u~x分别为MedMamba、MSPANe、RepVit、ResNet50、ViT和MobelNetV4模型诊断正常对照组、右锁骨下动脉迷走、双主动脉弓、右主动脉弓的超声影像解剖结构位置热力图

图3 各模型在血管环诊断任务中的混淆矩阵对比图。图a~f分别为MedMamba、MSPANet-50、RepVit、ResNet50、ViT和MobileNetV4模型诊断的混淆矩阵注：DAA为双主动脉弓；RAA为右主动脉弓；ARSA为右锁骨下动脉迷走；Normal为正常

图4 各模型诊断血管环疾病及识别正常影像的ROC曲线。图a~f分别为MedMamba、MSPANet-50、RepVit、ResNet50、ViT和MobileNetV4模型的ROC曲线图注：DAA为双主动脉弓；RAA为右主动脉弓；ARSA为右锁骨下动脉迷走；Normal为正常

表4 各模型在血管环标准切面诊断任务中的性能表现（%）

类别	准确性	敏感度	特异度	ROC曲线下面积	精确度	F1分数	第一命中率
MSPAnet-50
DAA	94.99	85.89	97.19	97.48	88.09	86.97	85.89
RAA	87.95	77.23	91.78	94.20	76.99	77.11	77.23
ARSA	99.68	98.82	100.00	99.99	100.00	99.40	98.82
正常	88.12	79.28	91.37	94.25	77.19	78.22	79.28
MedMamba
DAA	94.10	87.60	95.68	97.58	83.14	85.31	87.60
RAA	90.47	82.15	93.43	95.51	81.65	81.90	82.15
ARSA	99.68	99.11	99.89	99.94	99.70	99.41	99.11
正常	89.26	78.08	93.37	95.24	81.25	79.63	78.08
RepVit
DAA	93.38	85.12	95.38	94.98	81.75	83.40	85.12
RAA	88.29	68.31	95.40	91.72	84.09	75.38	68.31
ARSA	98.30	98.82	98.11	99.93	95.16	96.95	98.82
正常	87.08	81.68	89.06	91.59	73.32	77.27	81.68
ResNet-50
DAA	94.43	92.98	94.78	98.62	81.23	86.71	92.98
RAA	90.15	75.69	95.29	95.46	85.12	80.13	75.69
ARSA	99.76	99.11	100.00	99.99	100.00	99.55	99.11
正常	89.01	80.18	92.27	94.92	79.23	79.70	80.18
ViT
DAA	96.69	93.78	97.39	99.36	89.68	91.68	93.78
RAA	92.32	84.00	95.29	97.43	86.39	85.18	84.00
ARSA	99.51	98.82	99.78	99.99	99.40	99.11	98.82
正常	92.24	85.59	94.69	96.94	85.59	85.59	85.59
MobileNetV4
DAA	97.58	94.63	98.29	99.49	93.09	93.85	94.63
RAA	95.88	90.77	97.70	98.47	93.35	92.04	90.77
ARSA	99.76	100.00	99.67	100.00	99.12	99.56	100.00
正常	94.99	90.99	96.46	98.51	90.45	90.72	90.99

表4 各模型在血管环标准切面诊断任务中的性能表现（%）

类别	准确性	敏感度	特异度	ROC曲线下面积	精确度	F1分数	第一命中率
MSPAnet-50
DAA	94.99	85.89	97.19	97.48	88.09	86.97	85.89
RAA	87.95	77.23	91.78	94.20	76.99	77.11	77.23
ARSA	99.68	98.82	100.00	99.99	100.00	99.40	98.82
正常	88.12	79.28	91.37	94.25	77.19	78.22	79.28
MedMamba
DAA	94.10	87.60	95.68	97.58	83.14	85.31	87.60
RAA	90.47	82.15	93.43	95.51	81.65	81.90	82.15
ARSA	99.68	99.11	99.89	99.94	99.70	99.41	99.11
正常	89.26	78.08	93.37	95.24	81.25	79.63	78.08
RepVit
DAA	93.38	85.12	95.38	94.98	81.75	83.40	85.12
RAA	88.29	68.31	95.40	91.72	84.09	75.38	68.31
ARSA	98.30	98.82	98.11	99.93	95.16	96.95	98.82
正常	87.08	81.68	89.06	91.59	73.32	77.27	81.68
ResNet-50
DAA	94.43	92.98	94.78	98.62	81.23	86.71	92.98
RAA	90.15	75.69	95.29	95.46	85.12	80.13	75.69
ARSA	99.76	99.11	100.00	99.99	100.00	99.55	99.11
正常	89.01	80.18	92.27	94.92	79.23	79.70	80.18
ViT
DAA	96.69	93.78	97.39	99.36	89.68	91.68	93.78
RAA	92.32	84.00	95.29	97.43	86.39	85.18	84.00
ARSA	99.51	98.82	99.78	99.99	99.40	99.11	98.82
正常	92.24	85.59	94.69	96.94	85.59	85.59	85.59
MobileNetV4
DAA	97.58	94.63	98.29	99.49	93.09	93.85	94.63
RAA	95.88	90.77	97.70	98.47	93.35	92.04	90.77
ARSA	99.76	100.00	99.67	100.00	99.12	99.56	100.00
正常	94.99	90.99	96.46	98.51	90.45	90.72	90.99

图5 各模型特征分离能力的t-SNE分布可视化图。图a~f分别为MedMamba、MSPANet-50、RepVit、ResNet50、ViT和MobileNetV4模型的t-SNE 特征分布可视化图注：DAA为双主动脉弓；RAA为右主动脉弓；ARSA为右锁骨下动脉迷走；Normal为正常

1	中华医学会超声医学分会妇产超声学组, 国家卫生健康委妇幼司全国产前诊断专家组医学影像组.超声产前筛查指南[J]. 中华超声影像学杂志, 2022, 31(1): 1-12.
2	Bae SB, Kang EJ, Choo KS, et al. Aortic arch variants and anomalies: embryology, imaging findings, and clinical considerations[J]. J Cardiovasc Imaging, 2022, 30(4): 231-262.
3	Xie Y, Weng Z, Wang R, et al. Prenatal ultrasound diagnosis, associated anomalies and pregnancy outcomes of fetal right aortic arch[J]. Front Cardiovasc Med, 2025, 12: 1521338.
4	顾怡君, 李奕冉, 钱艺, 等.基于超声造影定量指标预测肝细胞癌微血管侵犯及评估其复发的研究[J/OL]. 中华医学超声杂志(电子版), 2025, 22(5): 451-461.
5	Priya S, Thomas R, Nagpal P, et al. Congenital anomalies of the aortic arch[J]. Cardiovasc Diagn Ther, 2018, 8(Suppl 1): S26-S44.
6	Hernanz-Schulman M. Vascular rings: a practical approach to imaging diagnosis[J]. Pediatr Radiol, 2005, 35(10): 961-979.
7	Lorentzen MJ, Kristensen AH, Kaldan FP, et al. Handheld ultrasound devices used by newly certified operators for pneumonia in the emergency department-a diagnostic accuracy study[J]. Diagnostics(Basel), 2024, 14(17): 1921.
8	余翔, 袁鹰, 李胜利. 产前超声人工智能应用研究[J/OL].中华医学超声杂志(电子版), 2025, 22(4): 300-304.
9	Jiang C, Ling W, Peng L, et al. Evaluation of tracheal diameter and angles in fetuses with double aortic arch using prenatal ultrasound: implications for postnatal management[J]. Front Med(Lausanne), 2024, 11: 1398623.
10	He J, Yang L, Liang B, et al. Fetal cardiac ultrasound standard section detection model based on multitask learning and mixed attention mechanism[J]. Neurocomputing, 2024, 579: 127443.
11	姜玉新, 王红燕, 李建初, 等.人工智能与超声质控的深度融合:赋能精准医疗与同质化发展[J/OL].中华医学超声杂志(电子版), 2025, 22(5): 383-387.
12	Pu B, Li K, Li S, et al. Automatic fetal ultrasound standard plane recognition based on deep learning and ⅡoT[J]. IEEE Transactions on Industrial Informatics, 2021, 17(11): 7771-7780.
13	Pu B, Zhu N, Li K, et al. Fetal cardiac cycle detection in multi-resource echocardiograms using hybrid classification framework[J]. Future Gener Comput Syst, 2021, 115: 825-836.
14	何冠南, 谭莹, 路玉欢, 等.人工智能在胎儿超声心动图标准切面质量控制中的多中心应用研究[J/OL].中华医学超声杂志(电子版), 2025, 22(5): 388-396.
15	Licari A, Manca E, Rispoli GA, et al. Congenital vascular rings: a clinical challenge for the pediatrician[J]. Pediatr Pulmonol, 2015, 50(5): 511-524.
16	Liang B, Peng F, Luo D, et al. Automatic segmentation of 15 critical anatomical labels and measurements of cardiac axis and cardiothoracic ratio in fetal four chambers using nnU-NetV2[J]. BMC Med Inform Decis Mak, 2024, 24(1): 128
17	Lu Y, Li K, Pu B, et al. A YOLOX-based deep instance segmentation neural network for cardiac anatomical structures in fetal ultrasound images[J]. IEEE/ACM Trans Comput Biol Bioinform, 2024, 21(4): 1007-1018.
18	Lu Y, Tan G, Pu B, et al. SKGC: A general semantic-level knowledge guided classification framework for fetal congenital heart disease[J]. IEEE J Biomed Health Inform, 2024, 28(10): 6105-6116.
19	Pu B, Lu Y, Chen J, et al. MobileUNet-FPN: A semantic segmentation model for fetal ultrasound four-chamber segmentation in edge computing environments[J]. IEEE J Biomed Health Inform, 2022, 26(11): 5540-5550.
20	Yue Y, Li Z. Medmamba: Vision mamba for medical image classification[DB/OL]. (2024-09-29)[2025-05-28].
21	Wang A, Chen H, Lin Z, et al. Repvit: Revisiting mobile CNN from Vit perspective[DB/OL].(2024-03-14)[2025-05-28].
22	Yu Y, Zhang Y, Cheng Z, et al. Multi-scale spatial pyramid attention mechanism for image recognition: An effective approach[J]. Engineering Applications of Artificial Intelligence, 2024, 133: 108261.
23	He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[DB/OL].(2015-11-10)[2025-05-28].
24	Han K, Wang Y, Chen H, et al. A survey on vision transformer[J]. IEEE Trans Pattern Anal Mach Intell, 2023, 45(1): 87-110.
25	Annetta R, Nisbet D, O'Mahony E, et al. Aberrant right subclavian artery: embryology, prenatal diagnosis and clinical significance[J]. Ultrasound, 2022, 30(4): 284-291.
26	裴秋艳, 何怡华, 李胜利. 胎儿心脏超声横切面正常与异常特征专家共识[J].中国超声医学杂志, 2023, 39(6): 601-608.
27	Ramirez Zegarra R, Ghi T. Use of artificial intelligence and deep learning in fetal ultrasound imaging[J]. Ultrasound Obstet Gynecol, 2023, 62(2): 185-194.
28	Axt-Fliedner R, Nazar A, Bedei I, et al. Associated anomalies and outcome in patients with prenatal diagnosis of aortic arch anomalies as aberrant right subclavian artery, right aortic arch and double aortic arch[J]. Diagnostics(Basel), 2024, 14(3): 238.
29	Ramirez P, Uus A, van Poppel M, et al. Multi-task learning for joint weakly-supervised segmentation and aortic arch anomaly classification in fetal cardiac MRI[DB/OL]. (2023-11-13)[2025-05-28].

[1]	江瑶, 蒋程, 余翔, 谭莹, 温昕, 温慧莹, 彭桂艳, 李胜利. 基于注意力机制改进的子宫解剖结构检测与分割多任务模型的性能评估[J/OL]. 中华医学超声杂志（电子版）, 2025, 22(08): 703-710.
[2]	傅小芳, 杨青翰, 孙昌琴, 豆梦杰, 胡峻溥, 孙灏, 吕发勤. 基于YOLO 11的肢体长骨骨折断端超声检测模型的临床价值[J/OL]. 中华医学超声杂志（电子版）, 2025, 22(06): 541-546.
[3]	何冠南, 谭莹, 路玉欢, 蒲斌, 扬水华, 张仁铁, 陈明, 石智红, 钟晓红, 陈曦, 燕柳屹, 李胜利. 人工智能在胎儿超声心动图标准切面质量控制中的多中心应用研究[J/OL]. 中华医学超声杂志（电子版）, 2025, 22(05): 388-396.
[4]	王彦, 张晓航, 冉素真, 钟春燕, 张晋炜, 王希. 双胎贫血-红细胞增多序列征的产前超声特征分析[J/OL]. 中华医学超声杂志（电子版）, 2025, 22(05): 462-469.
[5]	陈茵, 谭莹, 谭渤瀚, 何冠南, 王磊, 温昕, 朱巧珍, 梁博诚, 李胜利. 基于YOLO V8 的胎儿脐膨出超声智能质量评估与诊断[J/OL]. 中华医学超声杂志（电子版）, 2025, 22(04): 305-310.
[6]	郑仁杰, 张尊庶, 黄陈. 病理组学在胃癌诊治中的应用与挑战[J/OL]. 中华普通外科学文献(电子版), 2025, 19(04): 274-280.
[7]	王明媚, 李勇. 肾盂癌的影像诊断及进展[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2025, 19(04): 412-417.
[8]	詹彧鸣, 张翔, 翁山耕. 人工智能在腹膜后肿瘤精准诊疗中的研究进展[J/OL]. 中华疝和腹壁外科杂志(电子版), 2025, 19(04): 371-376.
[9]	希龙夫, 薛荣泉. 人工智能在肝胆胰肿瘤诊治中应用与进展[J/OL]. 中华腔镜外科杂志(电子版), 2025, 18(03): 166-171.
[10]	赵婷, 易晓芳. 人工智能在子宫腺肌病微创治疗中的应用进展[J/OL]. 中华腔镜外科杂志(电子版), 2025, 18(03): 172-176.
[11]	施薇薇, 楼微华, 狄文, 严斌, 张楠, 王酉. 人工智能驱动的腔镜外科发展：研究进展与未来趋势[J/OL]. 中华腔镜外科杂志(电子版), 2025, 18(03): 177-183.
[12]	郭寒川, 王乾宇, 吴斌. 人工智能在神经识别的研究进展及直肠癌自主神经保护的应用[J/OL]. 中华结直肠疾病电子杂志, 2025, 14(03): 273-276.
[13]	李媛媛, 李荣山. 机器学习：肾脏疾病研究与诊疗的新前沿[J/OL]. 中华肾病研究电子杂志, 2025, 14(04): 181-187.
[14]	卓莉. 人工智能和多组学：肾脏病诊疗的新方向[J/OL]. 中华肾病研究电子杂志, 2025, 14(03): 180-180.
[15]	王玲洁, 王瑷萍, 李朝军, 丁跃有, 杨德业, 赵清, 崔兆强, 王京昆, 王宏宇. 心脏和血管健康技术创新研发策略专家共识（2024第一次报告，上海）[J/OL]. 中华临床医师杂志(电子版), 2025, 19(05): 323-336.

阅读次数

全文

摘要

选择文件类型/文献管理软件名称

选择包含的内容

MobileNetV4: an intelligent tool for diagnosis of aortic arch branching anomalies based on prenatal ultrasound images

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

MobileNetV4: an intelligent tool for diagnosis of aortic arch branching anomalies based on prenatal ultrasound images