胰腺自动分割与区域定量及糖尿病评估研究

doi:10.11938/cjmr20253155

波谱学杂志 ›› 2025, Vol. 42 ›› Issue (4): 378-389.doi: 10.11938/cjmr20253155cstr: 32225.14.cjmr20253155

胰腺自动分割与区域定量及糖尿病评估研究

李英豪¹^,², 王丽辉³, 王苏成¹, 朱中旗¹, 黄长栋¹, 李仁峰³, 曹开明³, 胡海洋³, 贾一鸣³, 梁松涛³, 杨光¹^,²^,^§(), 路青³^,^#(), 汪红志¹^,²^,^*()

1.上海市磁共振重点实验室，华东师范大学物理与电子科学学院，上海 200062
2.华东师范大学医学磁共振与分子影像技术研究院，上海 200062
3.同济大学附属上海市东方医院放射科，上海 200123

收稿日期:2025-03-27 出版日期:2025-12-05 在线发表日期:2025-04-23
通讯作者: § Tel: 18521510757, E-mail: gyang@phy.ecnu.edu.cn; # Tel: 021-38804518, E-mail: Drluqingsjtu@163.com; * Tel: 13916346546, E-mail: hzwang@phy.ecnu.edu.cn.
基金资助:
上海市卫生健康委员会卫生行业临床研究专项(202340212);上海市东方医院引进人才项目(DFRC2023015)

Study on Pancreas Automatic Segmentation, Regional Quantification, and Diabetes Assessment

LI Yinghao¹^,², WANG Lihui³, WANG Sucheng¹, ZHU Zhongqi¹, HUANG Changdong¹, LI Renfeng³, CAO Kaiming³, HU Haiyang³, JIA Yiming³, LIANG Songtao³, YANG Guang¹^,²^,^§(), LU Qing³^,^#(), WANG Hongzhi¹^,²^,^*()

1. Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai 200062, China
2. Institute of Magnetic Resonance and Molecular Imaging in Medicine, East China Normal University, Shanghai 200062, China
3. Department of Radiology, Shanghai East Hospital, Tongji University, Shanghai 200123, China

Received:2025-03-27 Published:2025-12-05 Online:2025-04-23
Contact: § Tel: 18521510757, E-mail: gyang@phy.ecnu.edu.cn; # Tel: 021-38804518, E-mail: Drluqingsjtu@163.com; * Tel: 13916346546, E-mail: hzwang@phy.ecnu.edu.cn.

摘要/Abstract

摘要：

胰腺健康与糖尿病等疾病密切相关，准确检测胰腺脂肪含量对疾病的早期诊断和干预具有重要意义．本文提出了一种基于深度学习的胰腺自动分割与脂肪定量方法．首先，使用nnU-Net训练分割模型，实现对m-Dixon序列中胰腺整体的高精度分割，测试集DSC系数（Dice Similarity Coefficient，DSC）达0.92．随后，提出一种自动分区与脂肪定量评估方法，实现胰腺头、体、尾的精准划分，并定量分析其体积及脂肪含量．基于256例受试者的研究结果表明，胰腺尾部脂肪含量与2型糖尿病显著相关（p < 0.05）．进一步利用随机森林分类模型进行糖尿病风险预测，其中基于尾部脂肪含量的分类曲线下面积（Area Under the Curve，AUC）为0.68，而结合多区域脂肪信息构建的组合脂肪含量的分类AUC达0.73．研究结果表明，该方法可为糖尿病的早期诊断提供有效的技术支持.

关键词: 磁共振成像, 2型糖尿病预测, 深度学习, 胰腺, nnU-Net

Abstract:

Pancreatic health is closely linked to diabetes, making accurate fat quantification crucial for early diagnosis. This study proposes a deep learning-based method for automatic pancreatic segmentation and fat quantification. The nnU-Net model achieves high-precision segmentation on m-Dixon Imaging, with a Dice similarity coefficient (DSC) of 0.92. A novel sub-region partitioning and quantification method enables precise delineation of the pancreatic head, body, and tail. Analysis of 256 subjects (healthy, prediabetic, diabetic) reveals a significant association between pancreatic tail fat and type 2 diabetes (p < 0.05). Using random forest classifiers, diabetes risk was effectively predicted based on tail fat content and a composite fat index, yielding an area under the curve (AUC) of 0.68 and 0.73, respectively. This method offers a promising tool for the early diagnosis of diabetes.

Key words: magnetic resonance imaging (MRI), type 2 diabetes mellitus (T2DM) prediction, deep learning, pancreas, nnU-Net

中图分类号:

O482.53

李英豪, 王丽辉, 王苏成, 朱中旗, 黄长栋, 李仁峰, 曹开明, 胡海洋, 贾一鸣, 梁松涛, 杨光, 路青, 汪红志. 胰腺自动分割与区域定量及糖尿病评估研究[J]. 波谱学杂志, 2025, 42(4): 378-389.

LI Yinghao, WANG Lihui, WANG Sucheng, ZHU Zhongqi, HUANG Changdong, LI Renfeng, CAO Kaiming, HU Haiyang, JIA Yiming, LIANG Songtao, YANG Guang, LU Qing, WANG Hongzhi. Study on Pancreas Automatic Segmentation, Regional Quantification, and Diabetes Assessment[J]. Chinese Journal of Magnetic Resonance, 2025, 42(4): 378-389.

图/表 11

图1

图2

图3

图4

表1

图5

图6

图7

图8

图9

图10

参考文献 25

[1]	ONG K L, STAFFORD L K, MCLAUGHLIN S A, et al. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021[J]. Lancet, 2023, 402(10397): 203-234. doi: 10.1016/S0140-6736(23)01301-6 pmid: 37356446
[2]	TURNER C D, BAGNARA J T. General endocrinology[M]. 6th ed. Philadelphia: W. B. Saunders Company, 1976.
[3]	TAYLOR R. Understanding the cause of type 2 diabetes[J]. Lancet Diabetes Endocrinol, 2024, 12(9): 664-673. doi: 10.1016/S2213-8587(24)00157-8
[4]	WANG X, MISAWA R, ZIELINSKI M C, et al. Regional differences in islet distribution in the human pancreas-preferential beta-cell loss in the head region in patients with type 2 diabetes[J]. PloS one, 2013, 8(6): e67454. doi: 10.1371/journal.pone.0067454
[5]	SARMA M K, SAUCEDO A, DARWIN C H, et al. Noninvasive assessment of abdominal adipose tissues and quantification of hepatic and pancreatic fat fractions in type 2 diabetes mellitus[J]. Magn Reson Imaging, 2020, 72: 95-102. doi: S0730-725X(20)30326-X pmid: 32668273
[6]	NADARAJAH C, FANANAPAZIR G, CUI E, et al. Association of pancreatic fat content with type II diabetes mellitus[J]. Clin Radiol, 2020, 75(1): 51-56. doi: S0009-9260(19)30269-7 pmid: 31711639
[7]	GO VLW, DIMAGNO E P, GARDNER J D, et al. The Pancreas: An Integrated Textbook of Basic Science, Medicine, and Surgery[M]. 2nd ed. Hoboken: Blackwell Publishing, 2004.
[8]	SHEN J, BAUM T, CORDES C, et al. Automatic segmentation of abdominal organs and adipose tissue compartments in water-fat MRI: application to weight-loss in obesity[J]. Eur J Radiol, 2016, 85(9): 1613-1621. doi: 10.1016/j.ejrad.2016.06.006 pmid: 27501897
[9]	WOLZ R, CHU C, MISAWA K, et al. Automated abdominal multi-organ segmentation with subject-specific atlas generation[J]. IEEE Trans Med Imaging, 2013, 32(9): 1723-1730. doi: 10.1109/TMI.2013.2265805
[10]	CHU C, ODA M, KITASAKAI T, et al. Multi-organ segmentation based on spatially-divided probabilistic atlas from 3D abdominal CT images[C]// Medical Image Computing and Computer-Assisted Intervention-MICCAI 2013: 16th International Conference, Nagoya, Japan, September 22-26, 2013, Proceedings, Part II 16. Springer Berlin Heidelberg, 2013: 165-172.
[11]	SAITO A, NAWANO S, SHIMIZU A. Joint optimization of segmentation and shape prior from level-set-based statistical shape model, and its application to the automated segmentation of abdominal organs[J]. Med Image Anal, 2016, 28: 46-65. doi: 10.1016/j.media.2015.11.003 pmid: 26716720
[12]	王鑫. 基于统计模型的胰腺分割算法的研究与实现[D]. 沈阳: 东北大学, 2013.
[13]	DAI J L, HE C, WU J, et al. Pancreatic cystic neoplasms segmentation network combining dual decoding and global attention upsampling modules[J]. Chinese J Magn Reson, 2024, 41(2): 151-161.
	戴俊龙, 何聪, 武杰, 等. 融合双解码和全局注意力上采样模块的胰腺囊性肿瘤分割网络[J]. 波谱学杂志, 2024, 41(2): 151-161. doi: 10.11938/cjmr20233073
[14]	CHEN L, WAN L. CTUNet: automatic pancreas segmentation using a channel-wise transformer and 3D U-Net[J]. Vis Comput, 2023, 39(11): 5229-5243. doi: 10.1007/s00371-022-02656-2
[15]	PAITHANE P, KAKARWA S. LMNS-Net: Lightweight multiscale novel semantic-net deep learning approach used for automatic pancreas image segmentation in CT scan images[J]. Expert Syst Appl, 2023, 234: 121064. doi: 10.1016/j.eswa.2023.121064
[16]	GHORPADE H, JAGTA J, PATIL S, et al. Automatic segmentation of pancreas and pancreatic tumor: a review of a decade of research[J]. IEEE Access, 2023, 11: 108727-108745. doi: 10.1109/ACCESS.2023.3320570
[17]	ISENSEE F, JAEGER P F, KOHL S A A, et al. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation[J]. Nat Methods, 2021, 18(2): 203-211. doi: 10.1038/s41592-020-01008-z pmid: 33288961
[18]	SULOCHANA S, SIVAKAMI T. A gross morphological study of the pancreas in human cadavers[J]. Natl J Clin Anat, 2012, 1(2): 55-60. doi: 10.4103/2277-4025.298007
[19]	KUKU G M, HITTATIYA K, SPRINKART A M, et al. Comparison between modified Dixon MRI techniques, MR spectroscopic relaxometry, and different histologic quantification methods in the assessment of hepatic steatosis[J]. Eur Radiol, 2015, 25: 2869-2879. doi: 10.1007/s00330-015-3703-6 pmid: 25903702
[20]	FEDOROV A, BEICHEL R, KALPATHY-CRAMER J, et al. 3D Slicer as an image computing platform for the Quantitative Imaging Network[J]. Magn Reson Imaging, 2012, 30(9): 1323-1341. doi: 10.1016/j.mri.2012.05.001 pmid: 22770690
[21]	BI X L, LU M, XIAO B, et al. Pancreas segmentation based on dual-decoder U-Net convolutional neural network[J]. J Softw, 2022, 33(5): 1947-1958.
	毕秀丽, 陆猛, 肖斌, 等. 基于双解码U型卷积神经网络的胰腺分割[J]. 软件学报, 2022, 33(5): 1947-1958.
[22]	ZHANG Z, KELES E, DURAK G, et al. Large-scale multi-center CT and MRI segmentation of pancreas with deep learning[J]. Med Image Anal, 2025, 99: 103382. doi: 10.1016/j.media.2024.103382
[23]	QU T, LI X, WANG X, et al. Transformer guided progressive fusion network for 3D pancreas and pancreatic mass segmentation[J]. Med Image Anal, 2023, 86: 102801. doi: 10.1016/j.media.2023.102801
[24]	CHEN J, CHEN W, ZHU Z, et al. MAFE-Net: A multi-level attention feature extraction network for pancreas segmentation[J]. J Mach Learn, 2024,10: 1-19.
[25]	SKUDDE-HILL L, SEQUEIRA I R, CHO J, et al. Fat distribution within the pancreas according to diabetes status and insulin traits[J]. Diabetes, 2022, 71(6): 1182-1192. doi: 10.2337/db21-0976

胰腺自动分割与区域定量及糖尿病评估研究

Study on Pancreas Automatic Segmentation, Regional Quantification, and Diabetes Assessment

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 25

相关文章 15

编辑推荐

Metrics

本文评价

[1]	陈茜, 刘思婕, 蔡悦, 程琳琳, 王旭霞, 康彦, 林富春, 雷皓. 致痫剂量尼古丁对青少年雌性大鼠海马结构的影响[J]. 波谱学杂志, 2025, 42(4): 345-354.
[2]	高照耀, 张展, 胡亮亮, 许光宇, 周胜, 胡雨欣, 林子捷, 周超. 基于虚拟线圈和GRAPPA增强网络的PMRI方法[J]. 波谱学杂志, 2025, 42(4): 390-401.
[3]	张明玉, 肖洒, 石胜杰, 张学成, 周欣. 超极化¹²⁹Xe MRI的多模态增强去噪扩散模型研究[J]. 波谱学杂志, 2025, 42(4): 364-377.
[4]	马滢雪, 赵晏强, 杨晓冬, 蒋滨, 陶诚. 我国高场及超高场磁共振成像设备研制和市场化的机遇与挑战[J]. 波谱学杂志, 2025, 42(3): 334-344.
[5]	李鹏, 纪雨萍, 胡悦. 基于流形结构正则化的快速高质量磁共振指纹定量成像[J]. 波谱学杂志, 2025, 42(3): 249-264.
[6]	舒炜. B超和MRI在胎儿骨骼异常中的诊断价值分析[J]. 波谱学杂志, 2025, 42(3): 265-274.
[7]	隋美菊, 张磊, 王瑞芳, 骆盈盈, 李莎, 丘茂松, 徐秋怡, 陈代钦, 陈世桢, 周欣. MRI示踪的纳米酶用于级联反应增强的免疫治疗[J]. 波谱学杂志, 2025, 42(3): 231-248.
[8]	孟靖欣, 王远军. 基于扩散磁共振的大脑浅表白质纤维束研究进展[J]. 波谱学杂志, 2025, 42(2): 205-220.
[9]	孙灏芸, 王丽嘉. 融合注意力机制和空洞卷积的3D ELD_MobileNetV2在肝结节分类中的应用[J]. 波谱学杂志, 2025, 42(2): 130-142.
[10]	曹飞, 徐芊芊, 陈浩, 祖洁, 李晓文, 田锦, 鲍磊. 基于交叉自监督和DWI的NIID智能诊断方法[J]. 波谱学杂志, 2025, 42(2): 154-163.
[11]	陈群, 杨子剑, 程心怡, 郏思奕, 杜小霞, 王梦星. 磁共振成像技术在儿童运动干预研究中的应用[J]. 波谱学杂志, 2025, 42(2): 195-204.
[12]	杨佳铖, 王远军. 改进约束球面反卷积模型的脑灰质微结构成像[J]. 波谱学杂志, 2025, 42(1): 67-79.
[13]	薛培阳, 耿辰, 李郁欣, 鲍奕仿, 鲁宇澄, 戴亚康. 基于3D ResNet50改进模型的TOF-MRA脑动脉瘤分类方法[J]. 波谱学杂志, 2025, 42(1): 56-66.
[14]	宁欣宙, 黄臻, 陈西曲, 刘鑫杰, 陈罡, 张志, 鲍庆嘉, 刘朝阳. 用于超快时空编码MRI的Transformer超分辨率重建算法研究[J]. 波谱学杂志, 2024, 41(4): 454-468.
[15]	逄奇凡, 王志超, 武玉朋, 李建奇. K空间填充策略对基于FLASH序列的APT图像脂肪伪影的影响[J]. 波谱学杂志, 2024, 41(4): 443-453.