Automatic Identification of Brain Injury Mechanism Based on Deep Learning

doi:10.12116/j.issn.1004-5619.2021.410923

Abstract

Abstract: Objective

To apply the convolutional neural network （CNN） Inception_v3 model in automatic identification of acceleration and deceleration injury based on CT images of brain, and to explore the application prospect of deep learning technology in forensic brain injury mechanism inference.

Methods

CT images from 190 cases with acceleration and deceleration brain injury were selected as the experimental group, and CT images from 130 normal brain cases were used as the control group. The above-mentioned 320 imaging data were divided into training validation dataset and testing dataset according to random sampling method. The model classification performance was evaluated by the accuracy rate, precision rate, recall rate, F1-value and AUC value.

Results

In the training process and validation process, the accuracy rate of the model to classify acceleration injury, deceleration injury and normal brain was 99.00% and 87.21%, which met the requirements. The optimized model was used to test the data of the testing dataset, the result showed that the accuracy rate of the model in the test set was 87.18%, and the precision rate, recall rate, F1-score and AUC of the model to recognize acceleration injury were 84.38%, 90.00%, 87.10% and 0.98, respectively, to recognize deceleration injury were 86.67%, 72.22%, 78.79% and 0.92, respectively, to recognize normal brain were 88.57%, 89.86%, 89.21% and 0.93, respectively.

Conclusion

Inception_v3 model has potential application value in distinguishing acceleration and deceleration injury based on brain CT images, and is expected to become an auxiliary tool to infer the mechanism of head injury.

Key words: forensic medicine, acceleration brain injury, deceleration brain injury, image classification, deep learning, convolutional neural network, receiver operating characteristic （ROC） curve, Inception_v3 model

CLC Number:

DF795.1

Qi-fan YANG, Xue-yang SUN, Yan-bin WANG, Zhi-ling TIAN, He-wen DONG, Lei WAN, Dong-hua ZOU, Xiao-tian YU, Guang-zheng ZHANG, Ning-guo LIU. Automatic Identification of Brain Injury Mechanism Based on Deep Learning[J]. Journal of Forensic Medicine, 2022, 38(2): 223-230.

Figures/Tables 7

Tab. 1 Distribution of training validation dataset and testing dataset

组别	性别		年龄/岁			损伤类型			合计
组别	男性	女性	最大	最小	$x ˉ$ ±s	加速性损伤	减速性损伤	正常颅脑	合计
合计	251	69	97	7	43.53±16.74	109	81	130	320
训练验证集	179	45	97	7	43.57±16.68	80	63	81	224
测试集	72	24	84	11	44.35±16.70	29	18	49	96

Tab. 1 Distribution of training validation dataset and testing dataset

组别	性别		年龄/岁			损伤类型			合计
组别	男性	女性	最大	最小	$x ˉ$ ±s	加速性损伤	减速性损伤	正常颅脑	合计
合计	251	69	97	7	43.53±16.74	109	81	130	320
训练验证集	179	45	97	7	43.57±16.68	80	63	81	224
测试集	72	24	84	11	44.35±16.70	29	18	49	96

Fig. 1 The integral structure diagram of Inception_v3 model

Fig. 2 Schematic diagram of model loss function and accuracy rate change

Fig. 3 The confusion matrices of the model using 117 testing samples

Tab. 2 The classification results of Inception_v3 in the testing dataset

类别	精确率	召回率	F1值
加速性损伤	84.38	90.00	87.10
正常颅脑	88.57	89.86	89.21
减速性损伤	86.67	72.22	78.79

Fig. 4 The ROC curves of the model in the testing dataset

Fig. 5 Original CT image and the corresponding Grad-CAM image

References 30

1	朱传红，李道泉，王海生，等. 致伤方式、致伤工具的法医学复核鉴定分析12例[J].刑事技术，2001（6）：54-56. doi：10.16467/j.1008-3650.2001.06.037 .
	ZHU C H， LI D Q， WANG H S， et al. Analysis of 12 cases of injury methods and injury tools for forensic medicine review and identification[J]. Xingshi Jishu，2001（6）：54-56.
2	LANGLOIS J A， RUTLAND-BROWN W， WALD M M. The epidemiology and impact of traumatic brain injury[J]. J Head Trauma Rehabil，2006，21（5）：375-378. doi：10.1097/00001199-200609000-00001 .
3	郑剑. CT技术在法医学致伤方式鉴定中的应用[D].上海：复旦大学，2010.
	ZHENG J. The application of computed tomography（CT） in forensic appraisal of injury manner[D]. Shanghai： Fudan University，2010.
4	孙晔，胡宇祺，皮之云，等. 虚拟解剖在法医实践中的发展与应用[J].现代医用影像学，2021，30（8）：1427-1431.
	SUN Y， HU Y Q， PI Z Y， et al. Development and application of virtual anatomy in forensic practice[J]. Xiandai Yiyong Yingxiangxue，2021，30（8）：1427-1431.
5	李伟，齐麟. 虚拟解剖技术在法医学鉴定中的研究进展[J].广东公安科技，2020，28（3）：38-39，50.
	LI W， QI L. Research progress of virtual anatomy technique in forensic identification[J]. Guangdong Gongan Keji，2020，28（3）：38-39，50.
6	王宇聪，朱海标，刘冉，等. 虚拟解剖在法医病理学领域的应用现状[J].中国法医学杂志，2020，35（4）：360-368. doi：10.13618/j.issn.1001-5728.2020.04.004 .
	WANG Y C， ZHU H B， LIU R， et al. Application of virtual anatomy in forensic pathology[J]. Zhongguo Fayixue Zazhi，2020，35（4）：360-368.
7	HINTON G E， OSINDERO S， TEH Y W. A fast learning algorithm for deep belief nets[J]. Neural Comput，2006，18（7）：1527-1554. doi：10.1162/neco.2006.18.7.1527 .
8	孙雪瑞. 基于深度学习的病理图像细胞核分割[D].成都：电子科技大学，2020.
	SUN X R. Nuclear segmentation of pathological image based on deep learning[D]. Chengdu： University of Electronic Science and Technology of China，2020.
9	ESTEVA A， KUPREL B， NOVOA R A， et al. Dermatologist-level classification of skin cancer with deep neural networks[J]. Nature，2017，542（7639）：115-118. doi：10.1038/nature21056 .
10	KERMANY D S， GOLDBAUM M， CAI W， et al. Identifying medical diagnoses and treatable diseases by image-based deep learning[J]. Cell，2018，172（5）：1122-1131.e9. doi：10.1016/j.cell.2018.02.010 .
11	GULSHAN V， PENG L， CORAM M， et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs[J]. JAMA，2016，316（22）：2402-2410. doi：10.1001/jama.2016.17216 .
12	SPAMPINATO C， PALAZZO S， GIORDANO D， et al. Deep learning for automated skeletal bone age assessment in X-ray images[J]. Med Image Anal，2017，36：41-51. doi：10.1016/j.media.2016.10.010 .
13	胡婷鸿，火忠，刘太昂，等. 基于深度学习实现维吾尔族青少年左手腕关节骨龄自动化评估[J].法医学杂志，2018，34（1）：27-32. doi：10.3969/j.issn.1004-5619.2018.01.006 .
	HU T H， HUO Z， LIU T A， et al. Automated assessment for bone age of left wrist joint in Uyghur teenagers by deep learning[J]. Fayixue Zazhi，2018，34（1）：27-32.
14	LI Y， HUANG Z， DONG X， et al. Forensic age estimation for pelvic X-ray images using deep learning[J]. Eur Radiol，2019，29（5）：2322-2329. doi：10.1007/s00330-018-5791-6 .
15	彭丽琴，万雷，汪茂文，等. 运用3种卷积神经网络模型对青少年骨盆骨龄评估的比较[J].法医学杂志，2020，36（5）：622-630. doi：10.12116/j.issn.1004-5619.2020.05.004 .
	PENG L Q， WAN L， WANG M W， et al. Comparison of three CNN models applied in bone age assessment of pelvic radiographs of adolescents[J]. Fayixue Zazhi，2020，36（5）：622-630.
16	CAO Y， MA Y， VIEIRA D N， et al. A potential method for sex estimation of human skeletons using deep learning and three-dimensional surface scanning[J]. Int J Legal Med，2021，135（6）：2409-2421. doi：10.1007/s00414-021-02675-z .
17	盖荣丽，蔡建荣，王诗宇，等. 卷积神经网络在图像识别中的应用研究综述[J].小型微型计算机系统，2021，42（9）：1980-1984. doi：10.3969/j.issn.1000-1220.2021.09.030 .
	GAI R L， CAI J R， WANG S Y， et al. Research review on image recognition based on deep learning[J]. Xiaoxing Weixing Jisuanji Xitong，2021，42（9）：1980-1984.
18	HU H， YANG Y. A combined GLQP and DBN-DRF for face recognition in unconstrained environments[C]// Science and Engineering Research Center. Proceedings of the 2017 2nd International Conference on Control， Automation and Artificial Intelligence （CAAI 2017）， Paris， France： Atlantis Press，2017.
19	KRIZHEVSKY A， SUTSKEVER I， HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Commun ACM，2017，60（6）：84-90. doi：10.1145/3065386 .
20	HE K， ZHANG X， REN S， et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Trans Pattern Anal Mach Intell，2015，37（9）：1904-1916. doi：10.1109/TPAMI.2015.2389824 .
21	HU J， SHEN L， ALBANIE S， et al. Squeeze-and-excitation networks[J]. IEEE Trans Pattern Anal Mach Intell，2020，42（8）：2011-2023. doi：10.1109/TPAMI.2019.2913372 .
22	MONTEIRO M， NEWCOMBE V F J， MATHIEU F， et al. Multiclass semantic segmentation and quantification of traumatic brain injury lesions on head CT using deep learning： An algorithm development and multicentre validation study[J]. Lancet Digit Health，2020，2（6）：e314-e322. doi：10.1016/S2589-7500（20）30085-6 .
23	DHAR R， FALCONE G J， CHEN Y， et al. Deep learning for automated measurement of hemorrhage and perihematomal edema in supratentorial intracerebral hemorrhage[J]. Stroke，2020，51（2）：648-651. doi：10.1161/STROKEAHA.119.027657 .
24	FARZANEH N， WILLIAMSON C A， JIANG C， et al. Automated segmentation and severity analysis of subdural hematoma for patients with traumatic brain injuries[J]. Diagnostics （Basel），2020，10（10）：773. doi：10.3390/diagnostics10100773 .
25	YAO H， WILLIAMSON C， GRYAK J， et al. Automated hematoma segmentation and outcome prediction for patients with traumatic brain injury[J]. Artif Intell Med，2020，107：101910. doi：10.1016/j.artmed.2020.101910 .
26	LIU S， UTRIAINEN D， CHAI C， et al. Cerebral microbleed detection using susceptibility weighted imaging and deep learning[J]. Neuroimage，2019，198：271-282. doi：10.1016/j.neuroimage.2019.05.046 .
27	CHILAMKURTHY S， GHOSH R， TANAMALA S， et al. Deep learning algorithms for detection of critical findings in head CT scans： A retrospective study[J]. Lancet，2018，392（10162）：2388-2396. doi：10.1016/S0140-6736（18）31645-3 .
28	AOE J， FUKUMA R， YANAGISAWA T， et al. Automatic diagnosis of neurological diseases using MEG signals with a deep neural network[J]. Sci Rep，2019，9（1）：5057. doi：10.1038/s41598-019-41500-x .
29	SARRAF S， TOFIGHI G. Classification of Alzheimer’s disease using fMRI data and deep learning convolutional neural networks[J]. arXiv，2016. arXiv：.
30	GARLAND J， ONDRUSCHKA B， STABLES S， et al. Identifying fatal head injuries on postmortem computed tomography using convolutional neural network/deep learning： A feasibility study[J]. J Forensic Sci，2020，65（6）：2019-2022. doi：10.1111/1556-4029.14502 .

[1]	Hong-xia HAO, Jie-min CHEN, Rong-rong WANG, Xiao-ying YU, Meng WANG, Zhi-lu ZHOU, Yan-liang SHENG, Wen-tao XIA. The Value of VR-PVEP in Objective Evaluation of Monocular Refractive Visual Impairment [J]. Journal of Forensic Medicine, 2023, 39(4): 382-387.
[2]	Jie BAI, Jing SUN, Xiao-guang CHENG, Fan LIU, Hua LIU, Xu WANG. Construction and Application of Rib Fracture Diagnosis Model Based on YOLOv3 Algorithm [J]. Journal of Forensic Medicine, 2023, 39(4): 343-349.
[3]	Fei FAN, Juan WU, Zhen-hua DENG. Application Progress of Objective Audiological Detection Techniques in Forensic Clinical Medicine [J]. Journal of Forensic Medicine, 2023, 39(4): 360-366.
[4]	Jian XIANG, Xu WANG, Li-li YU, Kang-jia JIN, Ying-kai YANG. Objective Assessment of Visual Field Defects Caused by Optic Chiasm and Its Posterior Visual Pathway Injury [J]. Journal of Forensic Medicine, 2023, 39(4): 350-359.
[5]	Zhang-ming GAO, Jing-yu SHI, Hao ZENG, Xue-jun ZHANG. Rapid Determination of Bucinnazine in Blood by UPLC-MS/MS [J]. Journal of Forensic Medicine, 2023, 39(4): 388-392.
[6]	Yu-qi CAO, Yan SHI, Ping XIANG, Yin-long GUO. Research Progress on Machine Learning Assisted Non-Targeted Screening Strategy for Identification of Fentanyl Analogs [J]. Journal of Forensic Medicine, 2023, 39(4): 406-416.
[7]	Xu WANG. Legal Theories， Disability Models and Principles of Disability Assessment [J]. Journal of Forensic Medicine, 2023, 39(4): 329-336.
[8]	Hao-yang WANG, Jian WU, Qian ZHANG, Xin-feng MIN, Xiu-yan LIU, Yin-long GUO. Structural Analysis and Characterization of 4-F-α-PVP Analog 4-F-3-Methyl-α- PVP Hydrochloride [J]. Journal of Forensic Medicine, 2023, 39(2): 144-150.
[9]	Hang CHEN, Jing HU, Zheng QIAO, Hong-xiao DENG, Min LÜ, Wei LIU. Research Progress on Biological Matrix Reference Materials in Forensic Toxicology [J]. Journal of Forensic Medicine, 2023, 39(2): 176-185.
[10]	Jiao-jiao JI, Duo-qi XU, Ping XIANG, Hui YAN, Min SHEN. Analysis of Forty-Two Psychoactive Substances in a Single Hair by Micro-Segmental Technique [J]. Journal of Forensic Medicine, 2023, 39(2): 151-160.
[11]	Yong-gang MA, Yong-jie CAO, Yi-hua ZHAO, Xin-jun ZHOU, Bin HUANG, Gao-chao ZHANG, Ping HUANG, Ya-hui WANG, Kai-jun MA, Feng CHEN, Dong-chuan ZHANG, Ji ZHANG. Sex Estimation of Medial Aspect of the Ischiopubic Ramus in Adults Based on Deep Learning [J]. Journal of Forensic Medicine, 2023, 39(2): 129-136.
[12]	Zhong-ping CHENG, Yan-fei LIU, Xing-min XU, Yao-nan MO. Progress in the Application of Magnetic Nanoparticles in Forensic Trace Analysis [J]. Journal of Forensic Medicine, 2023, 39(2): 168-175.
[13]	Jiang-wei YAN, Jun-hong SUN, Hong-xing WANG, Zhi-wen WEI, Xiang-jie GUO, Ji LI, Cai-rong GAO, Geng-qian ZHANG, Xin-hua LIANG, Qiang-qiang ZHANG, Hong-wei WANG, Si-jin LI, Ying-yuan WANG, Ke-ming YUN. Exploration and Practice of the “One Combination， Two Highlights， Three Combinations， Four in One” Innovative Talents Training Mode in Forensic Medicine [J]. Journal of Forensic Medicine, 2023, 39(2): 193-199.
[14]	Dong GAO, Pei-pei ZHUO, Dong TIAN, Dan RAN, Qing XIA, Wen-tao XIA. Correlation between Elbow Flexor Muscle Strength and Needle Electromyography Parameters after Musculocutaneous Nerve Injury [J]. Journal of Forensic Medicine, 2023, 39(2): 137-143.
[15]	Hong-xia HAO, Ya-hui WANG, Zhi-lu ZHOU, Tai-ang LIU, Jin CHEN, Yu-heng HE, Lei WAN, Wen-tao XIA. Research Progress of Age Estimation in the Living by Knee Joint MRI [J]. Journal of Forensic Medicine, 2023, 39(1): 66-71.