法医学杂志

法医学杂志 ›› 2024, Vol. 40 ›› Issue (5): 419-429.DOI: 10.12116/j.issn.1004-5619.2024.440801

• 论著 •    下一篇

基于DeepLabV3+模型的钝性颅脑损伤CT图像智能识别与分割

秦豪杰1(), 刘媛媛1,2, 付恩浩1,2, 刘雅雯2,3, 田志岭2, 董贺文2, 刘太昂4, 邹冬华2, 程亦斌2(), 刘宁国2()   

  1. 1.河南科技大学基础医学与法医学院 特种医学研究院 司法鉴定中心，河南 洛阳 471000
    2.司法鉴定科学研究院 上海市法医学重点实验室 司法部司法鉴定重点实验室 上海市司法鉴定专业技术服务平台，上海 200063
    3.山西医科大学法医学院，山西 晋中 030600
    4.上海维解信息科技有限公司，上海 200444
  • 收稿日期:2024-08-01 发布日期:2025-02-11 出版日期:2024-10-25
  • 通讯作者: 程亦斌,刘宁国
  • 作者简介:秦豪杰（1978—），男，副教授，硕士研究生导师，主要从事法医病理学、法医临床学教学、科研和鉴定；E-mail：herochin@ haust.edu.cn
  • 基金资助:
    国家重点研发计划资助项目(2022YFC3302002);中央级公益性科研院所专项(GY2024D-1);上海市司法鉴定协会立项课题资助项目(SHSFJD2023-008);上海市法医学重点实验室资助项目(21DZ2270800);上海市司法鉴定专业技术服务平台资助项目;国家自然科学基金资助项目(82171872);上海市扬帆专项(23YF1448700);法医病理学公安部重点实验室开放课题资助项目(GAFYBL202308)

Intelligent Recognition and Segmentation of Blunt Craniocerebral Injury CT Images Based on DeepLabV3+ Model

Hao-jie QIN1(), Yuan-yuan LIU1,2, En-hao FU1,2, Ya-wen LIU2,3, Zhi-ling TIAN2, He-wen DONG2, Tai-ang LIU4, Dong-hua ZOU2, Yi-bin CHENG2(), Ning-guo LIU2()   

  1. 1.College of Basic Medicine and Forensic Medicine, Henan University of Science and Technology, Institute of Medical Aspects of Specific Environments, Judicial Expertise Center, Luoyang 471000, Henan Province, China
    2.Shanghai Key Laboratory of Forensic Medicine, Key Laboratory of Forensic Science, Ministry of Justice, Shanghai Forensic Service Platform, Academy of Forensic Science, Shanghai 200063, China
    3.School of Forensic Medicine, Shanxi Medical University, Jinzhong 030600, Shanxi Province, China
    4.Shanghai Wisdom Information Technology Co. Ltd. , Shanghai 200444, China
  • Received:2024-08-01 Online:2025-02-11 Published:2024-10-25
  • Contact: Yi-bin CHENG, Ning-guo LIU

RichHTML

3

PDF (PC)

410

补充材料

1

摘要：

目的 基于钝性颅脑损伤CT图像训练卷积神经网络DeepLabV3+模型，实现对常见颅脑损伤的智能化识别与分割（下文简称“分割”），探索深度学习技术在法医学钝性颅脑损伤自动化诊断中的应用价值。 方法 收集活体5 486张钝性颅脑损伤CT图像作为训练集、验证集和测试集进行模型训练与性能评估，另取活体255张钝性颅脑损伤与156张正常颅脑CT图像作为盲测集，评估模型分割5类颅脑损伤（头皮血肿、颅骨骨折、硬脑膜外血肿、硬脑膜下血肿和脑挫伤）的能力。再收集尸体340张钝性颅脑损伤和120张正常颅脑CT图像作为新的盲测集，探索用活体颅脑损伤CT图像训练的模型在尸体颅脑损伤分割中的应用价值。对除盲测集以外的所有钝性颅脑损伤CT图像中的5类颅脑损伤进行人工标记，再将各数据集输入模型，对模型进行训练后，根据训练集、验证集的损失函数与准确率评估并优化模型性能，根据测试集的Dice值评估模型泛化能力；根据盲测集的准确率、精确率和F1值评价模型对5类颅脑损伤的分割性能。 结果 经过对模型的训练和优化，最终的最优模型对头皮血肿、颅骨骨折、硬脑膜外血肿、硬脑膜下血肿和脑挫伤分割的平均Dice值分别是0.766 4、0.812 3、0.938 7、0.782 7和0.858 1，均大于0.75，达到了预期要求。盲测集的外部验证结果显示，5类颅脑损伤分割的F1值在活体颅脑损伤CT图像中分别是93.02%、89.80%、87.80%、92.93%和86.57%，在尸体颅脑损伤CT图像中分别是83.92%、44.90%、76.47%、64.29%和48.89%，说明该模型在活体CT图像上能准确分割5类颅脑损伤，而在尸体CT图像上的分割能力相对较差，但仍然能够准确分割头皮血肿、硬脑膜外血肿和硬脑膜下血肿。 结论 基于CT图像训练的深度学习模型可用于颅脑损伤的分割，但直接将活体颅脑损伤模型用于尸体颅脑损伤的分割有局限性。本研究为钝性颅脑损伤虚拟解剖数据的智能分割提供了新途径。

关键词: 法医学, 人工智能, DeepLabV3+模型, 钝性颅脑损伤, 深度学习, 计算机体层成像, 图像分割

Abstract:

Objective To achieve intelligent recognition and segmentation of common craniocerebral injuries （hereinafter referred to as “segmentation”） by training convolutional neural network DeepLabV3+ model based on CT images of blunt craniocerebral injury （BCI）, and to explore the value of deep learning in automated diagnosis of BCI in forensic medicine. Methods A total of 5 486 CT images of BCI from living persons were collected as the training set, validation set and test set for model training and performance evaluation. Another 255 CT images of BCI and 156 normal craniocerebral CT images from living persons were collected as the blind test set to evaluate the ability of the model to segment the five types of craniocerebral injuries including scalp hematoma, skull fracture, epidural hematoma, subdural hematoma, and brain contusion. Another 340 BCI and 120 normal craniocerebral CT images from cadavers were collected as the new blind test set to explore the application value of the model trained by living CT images in the segmentation of BCI in cadavers. The five types CT images of all BCI except the blind test set were manually labeled; then, each dataset was inputted into the model to train the model. The performance of the model was evaluated and optimized based on the loss function and accuracy curves of the training set and validation set, and the generalization ability was evaluated based on the Dice value of the test set. According to the accuracy, precision and F1 value of the blind test set, the segmentation performance of the model for five types of BCI was evaluated. Results After training and optimizing the model, the average Dice values of the final optimal model to scalp hematoma, skull fracture, epidural hematoma, subdural hematoma and brain contusion segmentation were 0.766 4, 0.812 3, 0.938 7, 0.782 7 and 0.858 1, respectively, all greater than 0.75, meeting the expected requirements. External validation showed that the F1 values were 93.02%, 89.80%, 87.80%, 92.93% and 86.57% in living CT images, respectively; 83.92%, 44.90%, 76.47%, 64.29% and 48.89% in cadaveric CT images, respectively. The above suggested that the model was able to accurately segment various types of craniocerebral injury on living CT images, while its segmentation ability was relatively poor on cadaveric CT images, but still able to accurately segment scalp hematoma, epidural hematoma and subdural hematoma. Conclusion Deep learning model trained on CT images can be used for BCI segmentation. However, the direct use of living persons’ BCI models for the identification of cadaveric BCI has some limitations. This study provides a new approach for intelligent segmentation of virtual anatomical data for BCI.

Key words: forensic medicine, artificial intelligence, DeepLabV3+ model, blunt craniocerebral injury, deep learning, computed tomography, image segmentation

中图分类号: