Comparison of CT Values between Thrombus and Postmortem Clot Based on Cadaveric Pulmonary Angiography / 法医学杂志

OBJECTIVES@#To explore the difference in CT values between pulmonary thromboembolism and postmortem clot in postmortem CT pulmonary angiography (CTPA) to further improve the application value of virtual autopsy.@*METHODS@#Postmortem CTPA data with the definite cause of death from 2016 to 2019 were collected and divided into pulmonary thromboembolism group (n=4), postmortem clot group (n=5), and control group (n=5). CT values of pulmonary trunk and left and right pulmonary artery contents in each group were measured and analyzed statistically.@*RESULTS@#The average CT value in the pulmonary thromboembolism group and postmortem clot group were (168.4±53.8) Hu and (282.7±78.0) Hu, respectively, which were lower than those of the control group (1 193.0±82.9) Hu (P<0.05). The average CT value of the postmortem clot group was higher than that of the pulmonary thromboembolism group (P<0.05).@*CONCLUSIONS@#CT value is reliable and feasible as a relatively objective quantitative index to distinguish pulmonary thromboembolism and postmortem clot in postmortem CTPA. At the same time, it can provide a scientific basis to a certain extent for ruling out pulmonary thromboembolism deaths.

Pelvic Injury Discriminative Model Based on Data Mining Algorithm / 法医学杂志

OBJECTIVES@#To reduce the dimension of characteristic information extracted from pelvic CT images by using principal component analysis (PCA) and partial least squares (PLS) methods. To establish a support vector machine (SVM) classification and identification model to identify if there is pelvic injury by the reduced dimension data and evaluate the feasibility of its application.@*METHODS@#Eighty percent of 146 normal and injured pelvic CT images were randomly selected as training set for model fitting, and the remaining 20% was used as testing set to verify the accuracy of the test, respectively. Through CT image input, preprocessing, feature extraction, feature information dimension reduction, feature selection, parameter selection, model establishment and model comparison, a discriminative model of pelvic injury was established.@*RESULTS@#The PLS dimension reduction method was better than the PCA method and the SVM model was better than the naive Bayesian classifier (NBC) model. The accuracy of the modeling set, leave-one-out cross validation and testing set of the SVM classification model based on 12 PLS factors was 100%, 100% and 93.33%, respectively.@*CONCLUSIONS@#In the evaluation of pelvic injury, the pelvic injury data mining model based on CT images reaches high accuracy, which lays a foundation for automatic and rapid identification of pelvic injuries.

Research Status of Postmortem MSCT Angiography in Forensic Science / 法医学杂志

Virtual autopsy is a new technique for investigating the morphological changes of cadaveric tissues and organs by medical imaging technology. It has been widely used in the identification of causes of death. Multislice spiral computed tomography （MSCT） has become a routine inspection method in some identification institutions, although it cannot completely replace traditional autopsy, it plays a key auxiliary or substitute role in the identification of certain abnormal causes of death. Plain MSCT scan cannot visualize cadaveric vessels, but can perform cadaveric angiography to determine vascular injury or disease. This technology has entered a rapid development period in recent years, and has made a considerable progress in contrast agent, perfusion methods and imaging methods. This article summarizes several common cadaveric MSCT angiography methods, such as systemic angiography, angiography through cardiopulmonary resuscitation, targeted angiography, and angiography by cardiac puncture, and analyzes and compares the application prospects.

Comparison of Three CNN Models Applied in Bone Age Assessment of Pelvic Radiographs of Adolescents / 法医学杂志

Objective To compare the performance of three deep-learning models （VGG19, Inception-V3 and Inception-ResNet-V2） in automatic bone age assessment based on pelvic X-ray radiographs. Methods A total of 962 pelvic X ray radiographs taken from adolescents （481 males, 481 females） aged from 11.0 to 21.0 years in five provinces and cities of China were collected, preprocessed and used as objects of study. Eighty percent of these X ray radiographs were divided into training set and validation set with random sampling method and used for model fitting and hyper-parameters adjustment. Twenty percent were used as test sets, to evaluate the ability of model generalization. The performances of the three models were assessed by comparing the root mean square error （RMSE）, mean absolute error （MAE） and Bland-Altman plots between the model estimates and the chronological ages. Results The mean RMSE and MAE between bone age estimates of the VGG19 model and the chronological ages were 1.29 and 1.02 years, respectively. The mean RMSE and MAE between bone age estimates of the Inception-V3 model and the chronological ages were 1.17 and 0.82 years, respectively. The mean RMSE and MAE between bone age estimates of the Inception-ResNet-V2 model and the chronological ages were 1.11 and 0.84 years, respectively. The Bland-Altman plots showed that the mean value of differences between bone age estimates of Inception-ResNet-V2 model and the chronological ages was the lowest. Conclusion In the automatic bone age assessment of adolescent pelvis, the Inception-ResNet-V2 model performs the best while the Inception-V3 model achieves a similar accuracy as VGG19 model.

Research Progress and Prospect of Machine Learning in Bone Age Assessment / 法医学杂志

Bone age assessment has always been one of the key issues and difficulties in forensic science. With the gradual development of machine learning in many industries, it has been widely introduced to imageology, genomics, oncology, pathology, surgery and other medical research fields in recent years. The reason why the above research fields can be closely combined with machine learning, is because the research subjects of the above branches of medicine belong to the computer vision category. Machine learning provides unique advantages for computer vision research and has made breakthroughs in medical image recognition. Based on the advantages of machine learning in image recognition, it was combined with bone age assessment research, in order to construct a recognition model suitable for forensic skeletal images. This paper reviews the research progress in bone age assessment made by scholars at home and abroad using machine learning technology in recent years.

Research Progress on Postmortem Changes of Computed Tomography Imaging Characteristics on Corpses / 法医学杂志

Postmortem changes on corpses appear immediately after death, and can transform the original structure characteristics of the corpse to different degrees as well as show specific changes on computed tomography （CT） images, sometimes with false positives and false negatives, influencing the identification of injuries or diseases. This paper systematically summarizes the postmortem changes of computed tomography imaging characteristics on corpses, to further expand the application of virtopsy in the practices of forensic pathology identification, and provide reference for the identification of injuries, diseases and changes after normal death.

Assessment of the Original Height of L1~2 after Vertebral Compression Fracture / 法医学杂志

OBJECTIVES@#To explore the assessment method of original height of L1-2 after vertebral compression fracture and its application value in forensic clinical practice.@*METHODS@#A total of 154 normal thoracic and lumbar X-ray films were collected, and 140 cases were used as experimental group while 14 cases as validation group. The heights of anterior （Ha） and posterior （Hp） vertebral body of T₁₂-L₃ vertebrae in each X-ray image were measured. In the experimental group, the correlation analysis between HaL₁ and HaT₁₂, HpT₁₂, HpL₁, HaL₂ and HpL₂ was carried out, and regression equation was established via fitting. The correlation analysis between HaL₂ and HaL₁, HpL₁, HpL₂, HaL₃, HpL₃ was performed, and the regression equation was also established via fitting. The difference between the predicted and measured values of HaL₁ and HaL₂ in validation group was compared.@*RESULTS@#In the 140 normal subjects, HaL₁ （y₁） was well correlated with HaT₁₂ （x₁） and HaL₂（x₂）, and the multiple linear regression equation was y₁=2.545+0.423 x₁+0.486 x₂ （determining coefficient R²=0.712, P<0.05; F=169.206, P<0.05）. There was no significant difference between the predicted and actual measured values of HaL₁ in the validation group （ P>0.05）. HaL₂ （y₂） was well correlated with HaL₁ （x₃） and HaL₃ （x₄）, and the multiple linear regression equation was y₂=4.354+0.530 x₃+0.349 x₄ （determining coefficient R²=0.689, P<0.05; F=151.575, P<0.05）. There was no significant difference between the predicted and actual measured values of HaL₂ in the validation group （ P>0.05）.@*CONCLUSIONS@#It is more appropriate to evaluate the original height of L₁ or L₂ single vertebrae by comparing with the height of the anterior edge of the upper and lower adjacent vertebral bodies.

Automated Assessment for Bone Age of Left Wrist Joint in Uyghur Teenagers by Deep Learning / 法医学杂志

Objective To realize the automated bone age assessment by applying deep learning to digital radiography(DR)image recognition of left wrist joint in Uyghur teenagers, and explore its practical ap-plication value in forensic medicine bone age assessment. Methods The X-ray films of left wrist joint after pretreatment, which were taken from 245 male and 227 female Uyghur nationality teenagers in Uygur Autonomous Region aged from 13.0 to 19.0 years old, were chosen as subjects. And AlexNet was as a regression model of image recognition. From the total samples above, 60% of male and fe-male DR images of left wrist joint were selected as net train set, and 10% of samples were selected as validation set. As test set, the rest 30%were used to obtain the image recognition accuracy with an error range in ±1.0 and ±0.7 age respectively, compared to the real age. Results The modelling results of deep learning algorithm showed that when the error range was in ±1.0 and ±0.7 age respectively, the accuracy of the net train set was 81.4% and 75.6% in male, and 80.5% and 74.8% in female, respectively. When the error range was in ±1.0 and ±0.7 age respectively, the accuracy of the test set was 79.5% and 71.2% in male, and 79.4% and 66.2% in female, respectively. Conclusion The combination of bone age research on teenagers' left wrist joint and deep learning, which has high accuracy and good feasi-bility, can be the research basis of bone age automatic assessment system for the rest joints of body.

Advantages and Application Prospects of Deep Learning in Image Recognition and Bone Age Assessment / 法医学杂志

Deep learning and neural network models have been new research directions and hot issues in the fields of machine learning and artificial intelligence in recent years. Deep learning has made a breakthrough in the applications of image and speech recognitions, and also has been extensively used in the fields of face recognition and information retrieval because of its special superiority. Bone X-ray images express different variations in black-white-gray gradations, which have image features of black and white contrasts and level differences. Based on these advantages of deep learning in image recognition, we combine it with the research of bone age assessment to provide basic datum for constructing a forensic automatic system of bone age assessment. This paper reviews the basic concept and network architectures of deep learning, and describes its recent research progress on image recognition in different research fields at home and abroad, and explores its advantages and application prospects in bone age assessment.