Stroke ranks as the third leading cause of death and the fourth leading cause of disability worldwide. Its high disability rate and prolonged recovery period not only severely impact patients' quality of life but also impose a significant burden on families and society. Primary prevention is the cornerstone of stroke control, as early intervention on risk factors can effectively reduce its incidence. Therefore, the development of predictive models for first-ever stroke risk holds substantial clinical value. In recent years, advancements in big data and artificial intelligence technologies have opened new avenues for stroke risk prediction. This article reviews the current research status of traditional methods and machine learning models in predicting first-ever stroke risk and outlines future development trends from three perspectives: First, emphasis should be placed on technological innovation by incorporating advanced algorithms such as deep learning and large models to further enhance the accuracy of predictive models. Second, there is a need to diversify data types and optimize model architectures to construct more comprehensive and precise predictive models. Lastly, particular attention should be given to the clinical validation of models in real-world settings. This not only enhances the robustness and generalizability of the models but also promotes physicians' understanding of predictive models, which is crucial for their application and dissemination.

Objective To observe the value of clinical and CT radiomics features for predicting microsatellite instability-high(MSI-H)status of gastric cancer.Methods Totally 150 gastric cancer patients including 30 cases of MSI-H positive and 120 cases of MSI-H negative were enrolled and divided into training set(n=105)or validation set(n=45)at the ratio of 7∶3.Based on abdominal vein phase enhanced CT images,lesions radiomics features were extracted and screened,and radiomics scores(Radscore)was calculated.Clinical data and Radscores were compared between MSI-H positive and negative patients in training set and validation set.Based on clinical factors and Radscores being significant different between MSI-H positive and negative ones,clinical model,CT radiomics model and clinical-CT radiomics combination model were constructed,and their predictive value for MSI-H status of gastric cancer were observed.Results Significant differences of tumor location and Radscore were found between MSI-H positive and negative patients in both training and validation sets(all P＜0.05).The area under the curve(AUC)of clinical model,CT radiomics model and combination model for evaluating MSI-H status of gastric cancer in training set was 0.760,0.799 and 0.864,respectively,of that in validation set was 0.735,0.812 and 0.849,respectively.AUC of clinical-CT radiomics combination model was greater than that of the other 2 single models(all P＜0.05).Conclusion Clinical-CT radiomics combination model based on tumor location and Radscore could effectively predict MSI-H status of gastric cancer.

Objective To observe the reproducibility of radiomics feature(RF)extracted from virtual monoenergetic image(VMI)of rabbit VX2 hepatoma models obtained with 3 different dual-energy CT(DECT)systems,and to explore relationship of reproducibility and diagnostic performance of RF.Methods Fifteen rabbits with VX2 hepatoma were randomly divided into 3 groups(each n=5).Contrast-enhanced abdominal CT scanning under volume CT dose index(CTDIvol)levels of 6,9 and 12 mGy were performed with dual-source DECT(dsDECT),rapid kV switching DECT(rsDECT)and dual-layer detector DECT(dlDECT),respectively.VMI were reconstructed at 10 keV increments from 40 to 140 keV.RF were extracted from VMI,the reproducibility was assessed using intra-class correlation coefficient(ICC),and those with ICC≥0.8 were considered as reproducible RF.The percentage of reproducible features(denoted by R)were compared among different scanner pairings and different CTDIvol levels.Within each CTDIvol group,the reconstruction energy levels yielding the maximum number(denoted by N)of common RF across different scanner pairings were identified.The receiver operating characteristic(ROC)curve was drawn,the area under the curve(AUC)was calculated,and the diagnostic efficacies of reproducible RF and other RF were compared under optimal reproducible conditions.Spearman correlation coefficient between ICC and the corresponding AUC of RF were calculated.Results RrsDECT-dsDECT(6.45％,95％CI[2.36％,8.87％])was higher than RdlDECT-dsDECT(0.72％,95％CI[0.15％,1.79％])and RrsDECT-dlDECT(1.43％,95％CI[0.60％,4.06％])(all adjusted P＜0.05),R9mGy(3.70％,95％CI[1.31％,5.73％])and R12mGy(2.63％,95％CI[0.60％,6.69％])were higher than R6mGy(1.31％,95％CI[0.12％,1.55％])(all adjusted P＜0.05).The optimal reproducible reconstruction energy levels of RF under CTDIvol of 6,9 and 12 mGy concentrated at 50-70 keV.AUC of reproducible RFs were higher than of other RF(all adjusted P＜0.05)and had certain correlation with the reproducibility(rs=0.102-0.516,P＜0.05).Conclusion The reproducibility of RF extracted from contrast-enhanced VMI CT images of rabbit VX2 hepatoma models associated with DECT scanner,CTDIvol level and reconstruction energy level.RF with higher reproducibility might have better diagnostic performance.

Objective To observe the value of clinical data combined with CT radiomics features for evaluating programmed cell death-ligand 1(PD-L1)status in gastric cancer.Methods Totally 277 gastric cancer patients were retrospectively enrolled and randomly divided into training set(n=195)and validation set(n=82)at the ratio of 7:3.There were 88 cases in PD-L1 positive subgroup and 107 cases in negative subgroup of training set,while 37 and 45 cases of validation set,respectively.The clinical and conventional CT features were compared between subgroups in both sets,the independent influencing factors of PD-L1 status in gastric cancer were analyzed,and radiomic features were screened based on CT data.Then clinical model,radiomics model and clinical-radiomics model were established,and the efficacy of each model for evaluating PD-L1 status in gastric cancer was observed.Results In training set,Borrmann type,cN stage,cM stage,clinical stage,maximum diameter and thickness were significant difference between subgroups(all P＜0.05).Borrmann type,clinical stage and the thickness were all independent influencing factors of PD-L1 positivity(all P＜0.05).The area under the curve(AUC)of clinical model,radiomic model and clinical-radiomics model for evaluating PD-L1 status in gastric cancer in training set was 0.748,0.832 and 0.841,respectively,and was 0.657,0.801 and 0.789 in validation set,respectively.AUC of clinical model was lower than the other models(all P＜0.05).Conclusion Clinical data combined with CT radiomics features was helpful for evaluating PD-L1 status in gastric cancer.

Patients with severe trauma require an extremely timely treatment and transfusion plays an irreplaceable role in the emergency treatment of such patients. An increasing number of evidence-based medicinal evidences and clinical practices suggest that patients with severe traumatic bleeding benefit from early transfusion of low-titer group O whole blood or hemostatic resuscitation with red blood cells, plasma and platelet of a balanced ratio. However, the current domestic mode of blood supply cannot fully meet the requirements of timely and effective blood transfusion for emergency treatment of patients with severe trauma in clinical practice. In order to solve the key problems in blood supply and blood transfusion strategies for emergency treatment of severe trauma, Branch of Clinical Transfusion Medicine of Chinese Medical Association, Group for Trauma Emergency Care and Multiple Injuries of Trauma Branch of Chinese Medical Association, Young Scholar Group of Disaster Medicine Branch of Chinese Medical Association organized domestic experts of blood transfusion medicine and trauma treatment to jointly formulate Chinese expert consensus on blood support mode and blood transfusion strategies for emergency treatment of severe trauma patients ( version 2024). Based on the evidence-based medical evidence and Delphi method of expert consultation and voting, 10 recommendations were put forward from two aspects of blood support mode and transfusion strategies, aiming to provide a reference for transfusion resuscitation in the emergency treatment of severe trauma and further improve the success rate of treatment of patients with severe trauma.

Objective:To assess the quality of low-keV monoenergetic images using deep learning image reconstruction (DLIR) algorithm combined with dual energy CT (DECT) in gastric cancer and to compare them with images from the conventional adaptive statistical iterative reconstruction (ASiR-V) algorithm.Methods:In this cross-sectional study, DECT images of 31 gastric cancer patients in the First Affiliated Hospital of Zhengzhou University were prospectively collected from September 2022 to March 2023. The 55 keV monoenergy images were reconstructed using the DLIR algorithm at low-, medium-, and high-intensity levels (DLIR-L, DLIR-M, and DLIR-H) based on arterial phase and venous phase images, respectively. The 70 keV 40% mixing coefficient (ASiR-V40%) images were reconstructed using the ASiR-V algorithm. In the objective evaluation of images, the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) for both lesions and muscle were calculated across four sets of reconstructed images. In the subjective evaluation of images, scores were assigned to the overall image quality, lesion visibility, and diagnostic confidence for each set of reconstructed images. Comparisons of SNR and CNR between the 4 groups were made by One-way repeated-measures ANOVA or Friedman′s test. Comparisons of scores were made by Friedman′s test. The P value of pairwise comparison was adjusted using Bonferroni correction methods. Results:In the objective evaluations, CNR lesion, SNR lesion and SNR muscle were highest on the 55 keV DLIR-H images in the arterial and venous phases, and showed a gradually increasing trend on the 70 keV ASiR-V40%, 55 keV DLIR-L, DLIR-M, DLIR-H images ( P<0.05). In subjective evaluations, compared to the 70 keV ASiR-V40% images, overall image quality scores were numerically higher for the 55 keV DLIR-H ( P>0.05), similar or slightly worse for the 55 keV DLIR-M, and significantly lower for the 55 keV DLIR-L ( P<0.05). The lesion visibility and diagnostic confidence on the 55 keV DLIR reconstruction images were higher in both arterial and venous phases than 70 keV ASiR-V40% images ( P<0.05). Conclusions:Compared to the conventional 70 keV ASiR-V40% images, the 55 keV DLIR-H images had higher lesion contrast and diagnostic confidence with lower image noise. The 55 keV DLIR-M images had comparable overall image quality to 70 keV ASiR-V40% images, but the former had higher lesion contrast and diagnostic confidence. The 55 keV DLIR-L was unable to improve image quality to the level of 70 keV ASiR-V40%.

Objective:To investigate the value of CT quantitative parameters of tumor and spleen in predicting the clinical stage of gastric cancer (Ⅰ/Ⅱ stage and Ⅲ/Ⅳ stage).Methods:This study was a case-control study. The data of 145 patients with gastric cancer confirmed by pathology in the First Affiliated Hospital of Zhengzhou University from February 2019 to June 2021 were retrospectively collected, including 70 cases of Ⅰ/Ⅱ stage and 75 cases of Ⅲ/Ⅳ stage. On the baseline CT images, the tumor related parameters, including tumor thickness, length of tumor, CT attenuation of tumor unenhanced phase, CT attenuation of tumor arterial phase, CT attenuation of tumor venous phase were measured. The spleen related parameters, including splenic thickness, CT attenuation of splenic unenhanced phase, CT attenuation of splenic arterial phase, CT attenuation of splenic venous phase, and standard deviation of CT attenuation (CTsd) in splenic unenhanced phase were also measured. The independent sample t test or Mann-Whitney U test was used to compare the parameters between the Ⅰ/Ⅱ stage and Ⅲ/Ⅳ stage patients. The multi-factor logistic regression analysis was used to find the independent predictors of gastric cancer clinical stage, and establish the combined parameters. The efficiency to the diagnosis of gastric cancer stage of single and combined parameters was evaluated using the operating characteristic curve, and the DeLong test was used to compare the differences of area under the curve (AUC). Results:There were significant differences in tumor thickness, length of tumor, CT attenuation of tumor venous phase, CT attenuation of splenic unenhanced phase, CT attenuation of splenic venous phase, CTsd in splenic unenhanced phase between the Ⅰ/Ⅱ stage and Ⅲ/Ⅳ stage of gastric cancer ( P<0.05). Multivariate analysis showed that tumor thickness ( OR=1.073, 95% CI 1.026-1.123, P=0.002), CT attenuation of splenic venous phase ( OR=1.040, 95% CI 1.011-1.070, P=0.006) and CTsd in splenic unenhanced phase ( OR=1.625, 95% CI 1.330-1.987, P<0.001) were independent risk factors for the clinical stage of gastric cancer and the combined parameters were established. The AUC values of tumor thickness, CT attenuation of splenic venous phase, CTsd in splenic unenhanced phase and combined parameters were 0.655, 0.614, 0.749 and 0.806, respectively. The AUC of combined parameters was higher than those of tumor thickness and CT attenuation of splenic venous phase, and the differences were statistically significant ( Z=3.37, 3.82, both P<0.001). Conclusion:Tumor thickness, CT attenuation of splenic venous phase and CTsd in splenic unenhanced phase are independent risk factors for the clinical stage of gastric cancer, and combined parameters can improve the diagnostic efficiency.

Objective:To evaluate the value of preoperative prediction of vessel invasion (VI) of locally advanced gastric cancer by machine learning model based on the venous phase enhanced CT radiomics features.Methods:A retrospective analysis of 296 patients with locally advanced gastric cancer confirmed by pathology in the First Affiliated Hospital of Zhengzhou University from July 2011 to December 2020 was performed. The patients were divided into VI positive group ( n=213) and VI negative group ( n=83) based on pathological results. The data were divided into training set ( n=207) and test set ( n=89) according to the ratio of 7∶3 with stratification sampling. The clinical characteristics of patients were recorded, and the independent risk factors of gastric cancer VI were screened by multivariate logistic regression. Pyradiomics software was used to extract radiomic features from the venous phase enhanced CT images, and the minimum absolute shrinkage and selection algorithm (LASSO) was used to screen the features, obtain the optimal feature subset, and establish the radiomics signature. Four machine learning algorithms, including extreme gradient boosting (XGBoost), logistic, naive Bayes (GNB), and support vector machine (SVM) models, were used to build prediction models for the radiomics signature and the screened clinical independent risk factors. The efficacy of the model in predicting gastric cancer VI was evaluated by the receiver operating characteristic curve. Results:The degree of differentiation (OR=13.651, 95%CI 7.265-25.650, P=0.003), Lauren′s classification (OR=1.349, 95%CI 1.011-1.799, P=0.042) and CA199 (OR=1.796, 95%CI 1.406-2.186, P=0.044) were independent risk factors for predicting the VI of locally advanced gastric cancer. Based on the venous phase enhanced CT images, 864 quantitative features were extracted, and 18 best constructed radiomics signature were selected by LASSO. In the training set, the area under the curve (AUC) of XGBoost, logistic, GNB and SVM models for predicting gastric cancer VI were 0.914 (95%CI 0.875-0.953), 0.897 (95%CI 0.853-0.940), 0.880 (95%CI 0.832-0.928) and 0.814 (95%CI 0.755-0.873), respectively, and in the test set were 0.870 (95%CI 0.769-0.971), 0.877 (95%CI 0.788-0.964), 0.859 (95%CI 0.755-0.961) and 0.773 (95%CI 0.647-0.898). The logistic model had the largest AUC in the test set. Conclusions:The machine learning model based on the venous phase enhanced CT radiomics features has high efficacy in predicting the VI of locally advanced gastric cancer before the operation, and the logistic model demonstrates the best diagnostic efficacy.

Objective:Evaluate the image quality of dual-layer spectral detector CT pulmonary angiography with low-dose contrast agents, and explore the influence of pulmonary artery diameter on the image quality.Methods:A total of 91 spectral CT pulmonary angiography from March 2022 to August 2022 were retrospectively analyzed. The cases were divided into Group 1 ( n=34, main pulmonary artery diameter ≥ 30 mm) and Group 2 ( n=57, main pulmonary artery diameter<30 mm). The dosage of contrast agent was 30 ml. The CT attenuation values(CT values), standard deviation(SD), signal-to-noise ratio(SNR), and contrast-to-noise ratio(CNR) values of pulmonary artery from the main trunk to the subsegmental pulmonary artery between two groups were compared. The CT dose index volume(CTDI vol) and dose-length production (DLP) were recorded. Two readers evaluated the image quality using three-point method. The inter-reader agreement was performed by Kappa test. Results:The CT values of the pulmonary trunk and left pulmonary artery between two groups was not significantly different ( P>0.05). The CT values of the left upper lobe artery, segmental artery, and subsegmental artery in Group 1 were lower than those in Group 2 ( t=-2.13, -2.17, Z=-2.33, P<0.05). The SD values of pulmonary trunk and segmental artery in Group 1 were higher than those in Group 2 ( t=2.27, Z=-2.23, P<0.05). The SD values of left pulmonary artery, left upper lobe artery, and subsegmental artery between two groups were not significantly different ( P>0.05). The SNR and CNR values of main pulmonary trunk, left pulmonary artery, left superior lobar artery, and segmental artery in Group 2 were higher than those in Group 1 ( Z=-2.45, -2.57, -2.09, -3.58, P<0.05; Z =-2.33, -2.42, -2.07, -3.45, P<0.05), while these values of the subsegmental artery between two groups were not significantly different ( P>0.05). The two readers had good consistency in evaluating image quality (Kappa value>0.75, P<0.05). Conclusions:Spectral CT pulmonary angiography with 30 ml contrast agent would generate good quality images. However, the distal pulmonary artery would be poorly revealed when the diameter of main pulmonary artery more than 30 mm, especially in patients with suspected pulmonary hypertension.

Photoacoustic imaging combines the advantages of conventional optical imaging and ultrasound imaging with high spatial resolution and imaging contrast that can capture molecular information from macroscopic, mesoscopic and microscopic levels in real time and provide highly specific tissue images. Photoacoustic imaging has been successfully applied to the study of clinical diagnosis of breast cancer. It has significant advantages in early detection of breast cancer, benign and malignant diagnosis, identifying different pathological types and molecular subtypes of diagnosis, monitoring the effect of anti-cancer treatment and assessing tumor margins to guide surgery. Based on the published data of clinical researches, this review summarizes the advantages of photoacoustic imaging over conventional breast imaging modalities in breast cancer detection and the progress of clinical studies.