Objective To explore the feasibility of constructing a three-classification bone status screening radiomics model on chest CT images.Methods A total of 371 patients who underwent both chest and abdominal plain CT examinations were retrospec-tively selected and randomly divided into training set(296 cases)and test set(75 cases)in a ratio of 8︰2.Additionally,110 patients were included as external validation set using the same criteria.The 120 kVp abdominal images were transmitted to a quantitative compu-ted tomography(QCT)post-processing workstation to measure the bone mineral density(BMD)of the L1-L2 vertebral bodies.Patients were classified into osteoporosis(OP)group(BMD<80 mg/cm3),osteopenia group(80 mg/cm3≤BMD≤120 mg/cm3)and normal bone mass group(BMD>120 mg/cm3)based on QCT BMD results.The automatic segmentation model was used to segment T10-T12 vertebral trabecular bone on chest CT images and the radiomics models based on random forest(RF)and logistic regres-sion(LR)was established to evaluate BMD,enabling it to simultaneously distinguish OP,osteopenia,and normal bone mass.The diag-nostic performance of the two models were evaluated using metrics such as the area under the curve(AUC),sensitivity and specificity.The DeLong test was used to compare the differences between the two models.Results In the test set,the AUC for differentiating normal bone mass were 0.948 and 0.877 for the RF and LR models,respectively;the AUC for differentiating OP were 0.942 and 0.836,respectively;and the AUC for differentiating osteopenia were 0.871 and 0.688,respectively.The performance comparison results of the models showed that there was no statistically significant difference in AUC(0.966 vs 0.907,P>0.05)between RF model and LR model in the external validation set for distinguishing OP,while there was a statistically significant difference in AUC for distinguishing osteopenia(0.895 vs 0.749,P=0.009)and normal bone mass(0.975 vs 0.906,P=0.023).The RF model performance was superior to the LR model.Conclusion The radiomics model developed based on chest plain CT can be used for opportunistic OP screening with good diagnostic efficacy,and the the model based on the RF classifier outperforms the LR model.

Objective To explore the application value of deep learning image reconstruction(DLIR)algorithm combined with a large reconstruction matrix in lung nodules screening using low-dose computed tomography(LDCT)of the chest.Methods Patients who underwent LDCT scans were prospectively enrolled.The control group(group A)used the iterative reconstruction(IR)algorithm(Karl)with a reconstruction level of Karl 5,reconstructed images of 512×512(group A1)matrix,and 1 024 × 1 024(group A2)matrix.The experimental group employed DLIR combined with 512×512(group B)matrix and 1 024 × 1 024(group C)matrix for image reconstruction at levels 1-5,which were recorded as groups B1-5 and groups C1-5.The CT values and standard deviation(SD)values of the lung parenchyma and tracheal air were measured,and the signal-to-noise ratio(SNR)was calculated.The overall lung image quality was scored on a Likert 5-point scale,and the subgroup with the best lung image quality was selected.The lung nodule detec-tion rate and clarity were compared with group A1.Results Under the same reconstruction matrix,the CT values of the tracheal air and lung parenchyma in the experimental group showed no significant difference compared to the control group,while the SD values were lower and SNR were higher(P＜0.05).Within groups B and C,as the DLIR level increased,the SD values of the tracheal air and lung paren-chyma gradually decreased,and SNR gradually improved(P＜0.05).Subjective scores for the image quality in groups B and C initially increased and then decreased,with group B3 and group C4 showed the best image quality.No difference was observed in objective eval-uation between the two groups,but the subjective image quality score of group C4 was superior to group B3(P＜0.05).Subjective eval-uation of lung nodule display in group C4 was better than in group A1(P＜0.05).Conclusion DLIR algorithm combined with a large reconstruction matrix is feasible for lung nodules screening in chest LDCT,reducing image noise while improving lung nodules clarity,demonstrating significant clinical value.

Objective:To explore the application value of the double-low technique of auto-prescription technique combined with iterative reconstruction algorithm in CT pulmonary angiography (CTPA).Methods:A total of 86 patients who were clinically suspected of having pulmonary embolism and underwent CTPA examination in the First Affiliated Hospital of Dalian Medical University were prospectively collected and randomly assigned to a control group ( n = 45) and an observation group ( n = 41) according to the random number table method. In the control group, a tube voltage of 120 kVp was used with a standard iodine contrast agent dose of 60 ml, and images were reconstructed using the 40% adaptive statistical iterative reconstruction algorithm (ASIR-V). In the observation group, the tube voltage was set by auto-prescription technique, and 0.4 ml/kg of personalized low iodine contrast agent was used. Images were reconstructed with 40%, 60%, and 80% ASIR-V, respectively, and designated as observation 1, observation 2, and observation 3 respectively. The volume CT dose index (CTDI vol), dose-length product (DLP), and effective dose ( E) were recorded and compared among the four groups. The CT values and standard deviation (SD) of the main pulmonary artery, left and right pulmonary arteries, as well as the left and right pulmonary lobe arteries were measured, and the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of these arteries were calculated. Additionally, the SD value at the contrast medium concentration in the superior vena cava was measured, and the artifact index (AI) was subsequently calculated. Two observers independently assessed the visibility of the pulmonary arteries, image noise, and sclerosis artifacts in the superior vena cava using a blinded method. Results:The E in the observation group was 3.28 (2.08, 3.93) mSv, which was significantly lower than that in the control group [5.03 (4.86, 5.20)] mSv, and the difference was statistically significant ( Z = 174.00, P < 0.05). The contrast agent dosage in the observation group was 28 (25, 30) ml, which was lower than that in the control group (60 ml), and the difference was statistically significant ( Z = 0, P < 0.05). The CT values for the main pulmonary artery and the left and right pulmonary lobe arteries in the observation group were higher than those in the control group, and the differences were all statistically significant ( t = -3.65 to -3.89, P < 0.05). The SNR and CNR of the observation groups 2 and 3 were greater than those of the control group ( t = -9.20 to -2.98, P < 0.05). The consistency of subjective evaluations between the two observers was good ( Kappa = 0.729 - 0.879, P < 0.05). There was no statistically significant difference in the subjective score of pulmonary artery visibility between the control and observation group ( P > 0.05). The subjective scores for image noise in observation group 2 and group 3 were higher than those in the control group ( U =598.50, 654.00, P < 0.05). The presence of artifacts due to sclerosis in the superior vena cava was significantly lower in the observation group compared to the control group ( χ2 = 46.09, P < 0.001). Conclusions:The combination of auto-prescription technique with ASIR-V reconstruction algorithm and low contrast agent imaging protocol can reduce the radiation dose and contrast agent dose without compromising image quality, and enable personalized double low CTPA imaging.

Objective To develop an automated bone mineral density(BMD)assessment model based on proximal femur images from abdomen-pelvic CT scans and to analyze its application value in opportunistic osteoporosis(OP)screening.Methods A retrospective selection was conducted on 351 patients who underwent abdomen-pelvic plain CT examination.The patients were randomly divided into training set(n=245)and test set(n=106)in a ratio of 7∶3.All images were transferred to a quantitative computed tomography(QCT)post-processing workstation to measure the BMD of the left proximal femur.According to the QCT BMD T-score,the patients were divided into osteoporosis(T-score-2.5),osteopenia(-2.5＜T-score＜-1)and normal bone density(T-score≥-1).The left proximal femur was dissected using an automatic segmentation model,and two three-class BMD assessment radiomics models were constructed using random forest(RF)and logistic regression(LR)classifiers,respectively.The receiver operating characteristic(ROC)curves were generated,and the area under the curve(AUC),sensitivity,specificity and other metrics were calculated to evaluate the diagnostic performance of the two models.The DeLong test was used to compare differences between the models.Results In the test set,the AUC of the RF and LR models for identifying osteoporosis were 0.953 and 0.954,respectively.The AUC for identifying osteopenia were 0.894 and 0.870,and the AUC for identifying normal bone density were 0.975 and 0.982,respectively.The comparison of model performance showed no statistically significant differences between the RF and LR models in identifying the three bone states in both the training and test sets(P＞0.05).Conclusion Both the RF and LR radiomics models,constructed based on abdomen-pelvic plain CT scans,can be used for opportunistic BMD screening with high diagnostic efficiency.

Objective To explore the feasibility of constructing a three-classification bone status screening radiomics model on chest CT images.Methods A total of 371 patients who underwent both chest and abdominal plain CT examinations were retrospec-tively selected and randomly divided into training set(296 cases)and test set(75 cases)in a ratio of 8︰2.Additionally,110 patients were included as external validation set using the same criteria.The 120 kVp abdominal images were transmitted to a quantitative compu-ted tomography(QCT)post-processing workstation to measure the bone mineral density(BMD)of the L1-L2 vertebral bodies.Patients were classified into osteoporosis(OP)group(BMD<80 mg/cm3),osteopenia group(80 mg/cm3≤BMD≤120 mg/cm3)and normal bone mass group(BMD>120 mg/cm3)based on QCT BMD results.The automatic segmentation model was used to segment T10-T12 vertebral trabecular bone on chest CT images and the radiomics models based on random forest(RF)and logistic regres-sion(LR)was established to evaluate BMD,enabling it to simultaneously distinguish OP,osteopenia,and normal bone mass.The diag-nostic performance of the two models were evaluated using metrics such as the area under the curve(AUC),sensitivity and specificity.The DeLong test was used to compare the differences between the two models.Results In the test set,the AUC for differentiating normal bone mass were 0.948 and 0.877 for the RF and LR models,respectively;the AUC for differentiating OP were 0.942 and 0.836,respectively;and the AUC for differentiating osteopenia were 0.871 and 0.688,respectively.The performance comparison results of the models showed that there was no statistically significant difference in AUC(0.966 vs 0.907,P>0.05)between RF model and LR model in the external validation set for distinguishing OP,while there was a statistically significant difference in AUC for distinguishing osteopenia(0.895 vs 0.749,P=0.009)and normal bone mass(0.975 vs 0.906,P=0.023).The RF model performance was superior to the LR model.Conclusion The radiomics model developed based on chest plain CT can be used for opportunistic OP screening with good diagnostic efficacy,and the the model based on the RF classifier outperforms the LR model.

Objective:To explore the application value of the double-low technique of auto-prescription technique combined with iterative reconstruction algorithm in CT pulmonary angiography (CTPA).Methods:A total of 86 patients who were clinically suspected of having pulmonary embolism and underwent CTPA examination in the First Affiliated Hospital of Dalian Medical University were prospectively collected and randomly assigned to a control group ( n = 45) and an observation group ( n = 41) according to the random number table method. In the control group, a tube voltage of 120 kVp was used with a standard iodine contrast agent dose of 60 ml, and images were reconstructed using the 40% adaptive statistical iterative reconstruction algorithm (ASIR-V). In the observation group, the tube voltage was set by auto-prescription technique, and 0.4 ml/kg of personalized low iodine contrast agent was used. Images were reconstructed with 40%, 60%, and 80% ASIR-V, respectively, and designated as observation 1, observation 2, and observation 3 respectively. The volume CT dose index (CTDI vol), dose-length product (DLP), and effective dose ( E) were recorded and compared among the four groups. The CT values and standard deviation (SD) of the main pulmonary artery, left and right pulmonary arteries, as well as the left and right pulmonary lobe arteries were measured, and the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of these arteries were calculated. Additionally, the SD value at the contrast medium concentration in the superior vena cava was measured, and the artifact index (AI) was subsequently calculated. Two observers independently assessed the visibility of the pulmonary arteries, image noise, and sclerosis artifacts in the superior vena cava using a blinded method. Results:The E in the observation group was 3.28 (2.08, 3.93) mSv, which was significantly lower than that in the control group [5.03 (4.86, 5.20)] mSv, and the difference was statistically significant ( Z = 174.00, P < 0.05). The contrast agent dosage in the observation group was 28 (25, 30) ml, which was lower than that in the control group (60 ml), and the difference was statistically significant ( Z = 0, P < 0.05). The CT values for the main pulmonary artery and the left and right pulmonary lobe arteries in the observation group were higher than those in the control group, and the differences were all statistically significant ( t = -3.65 to -3.89, P < 0.05). The SNR and CNR of the observation groups 2 and 3 were greater than those of the control group ( t = -9.20 to -2.98, P < 0.05). The consistency of subjective evaluations between the two observers was good ( Kappa = 0.729 - 0.879, P < 0.05). There was no statistically significant difference in the subjective score of pulmonary artery visibility between the control and observation group ( P > 0.05). The subjective scores for image noise in observation group 2 and group 3 were higher than those in the control group ( U =598.50, 654.00, P < 0.05). The presence of artifacts due to sclerosis in the superior vena cava was significantly lower in the observation group compared to the control group ( χ2 = 46.09, P < 0.001). Conclusions:The combination of auto-prescription technique with ASIR-V reconstruction algorithm and low contrast agent imaging protocol can reduce the radiation dose and contrast agent dose without compromising image quality, and enable personalized double low CTPA imaging.

Objective To explore the application value of deep learning image reconstruction(DLIR)algorithm combined with a large reconstruction matrix in lung nodules screening using low-dose computed tomography(LDCT)of the chest.Methods Patients who underwent LDCT scans were prospectively enrolled.The control group(group A)used the iterative reconstruction(IR)algorithm(Karl)with a reconstruction level of Karl 5,reconstructed images of 512×512(group A1)matrix,and 1 024 × 1 024(group A2)matrix.The experimental group employed DLIR combined with 512×512(group B)matrix and 1 024 × 1 024(group C)matrix for image reconstruction at levels 1-5,which were recorded as groups B1-5 and groups C1-5.The CT values and standard deviation(SD)values of the lung parenchyma and tracheal air were measured,and the signal-to-noise ratio(SNR)was calculated.The overall lung image quality was scored on a Likert 5-point scale,and the subgroup with the best lung image quality was selected.The lung nodule detec-tion rate and clarity were compared with group A1.Results Under the same reconstruction matrix,the CT values of the tracheal air and lung parenchyma in the experimental group showed no significant difference compared to the control group,while the SD values were lower and SNR were higher(P＜0.05).Within groups B and C,as the DLIR level increased,the SD values of the tracheal air and lung paren-chyma gradually decreased,and SNR gradually improved(P＜0.05).Subjective scores for the image quality in groups B and C initially increased and then decreased,with group B3 and group C4 showed the best image quality.No difference was observed in objective eval-uation between the two groups,but the subjective image quality score of group C4 was superior to group B3(P＜0.05).Subjective eval-uation of lung nodule display in group C4 was better than in group A1(P＜0.05).Conclusion DLIR algorithm combined with a large reconstruction matrix is feasible for lung nodules screening in chest LDCT,reducing image noise while improving lung nodules clarity,demonstrating significant clinical value.

Objective To develop an automated bone mineral density(BMD)assessment model based on proximal femur images from abdomen-pelvic CT scans and to analyze its application value in opportunistic osteoporosis(OP)screening.Methods A retrospective selection was conducted on 351 patients who underwent abdomen-pelvic plain CT examination.The patients were randomly divided into training set(n=245)and test set(n=106)in a ratio of 7∶3.All images were transferred to a quantitative computed tomography(QCT)post-processing workstation to measure the BMD of the left proximal femur.According to the QCT BMD T-score,the patients were divided into osteoporosis(T-score-2.5),osteopenia(-2.5＜T-score＜-1)and normal bone density(T-score≥-1).The left proximal femur was dissected using an automatic segmentation model,and two three-class BMD assessment radiomics models were constructed using random forest(RF)and logistic regression(LR)classifiers,respectively.The receiver operating characteristic(ROC)curves were generated,and the area under the curve(AUC),sensitivity,specificity and other metrics were calculated to evaluate the diagnostic performance of the two models.The DeLong test was used to compare differences between the models.Results In the test set,the AUC of the RF and LR models for identifying osteoporosis were 0.953 and 0.954,respectively.The AUC for identifying osteopenia were 0.894 and 0.870,and the AUC for identifying normal bone density were 0.975 and 0.982,respectively.The comparison of model performance showed no statistically significant differences between the RF and LR models in identifying the three bone states in both the training and test sets(P＞0.05).Conclusion Both the RF and LR radiomics models,constructed based on abdomen-pelvic plain CT scans,can be used for opportunistic BMD screening with high diagnostic efficiency.

Objective Based on a novel type of cell death induced by disulfide stress,known as disulfidptosis,this study explores the role of long non-coding RNA(LncRNA)in gastric cancer and establishes a prognosis model re-lated to disulfidptosis,providing a new method for assessing the prognosis of gastric cancer treatment.Methods Transcriptomic data from gastric cancer and normal tissue samples were obtained from the public database TCGA,and disulfidptosis-related LncRNAs were selected through Pearson analysis and LASSO-Cox regression analysis.A relevant prognostic model for gastric cancer was constructed based on the above LncRNAs and validated by function-al enrichment analysis,tumour microenvironment and immune cell infiltration analysis,drug sensitivity analysis and quantitative reverse transcription PCR(RT-qPCR).Results In this study,400 disulfide death-associated LncR-NAs were identified and five of them were screened to construct a prognostic model for assessing the prognosis of gastric cancer patients.The models showed in validation that the survival of the high-risk score group was shorter than that of the low-risk score group(P＜0.05).In addition,the predictive ability of the prognostic model(AUC=0.725)was better than that based only on basic characteristics such as age and gender.The expression levels of disulfide death-associated LncRNAs differed between normal and gastric cancer tissues(P＜0.001).Conclusion The disulfidptosis-related LncRNA prognosis model developed in this study can effectively assess the prognosis of gastric cancer patients and the tumor microenvironment,providing potential targets and a theoretical basis for new immunotherapeutic strategies for gastric cancer.

The purpose of this study was to investigate the role of miR-4281 in immune thrombocytopenic purpura(ITP)and its underlying mechanism.Differentially expressed miRNAs in ITP were searched by using GEO database.Newly diagnosed ITP patients,treatment-effective ITP patients and healthy subjects were collected.Venous blood was obtained and peripheral blood mononuclear cells(PBMCs)and CD4+T cells were separated.Real-time quantitative PCR was used to detect the expression of miR-4281 and other factors in cells.Flow cytometry was conducted to analyze the rate of CD4+CD25+FOXP3+Tregs and Th17 cells.Pearson correlation coefficient method was used to test the relationship between miR-4281 expression level and various factors.Bioinformatics analysis revealed an abnormal expression of miR-4281 in ITP.Compared with healthy subjects and treatment-effective ITP patients,the expression level of miR-4281 in newly diagnosed ITP was decreased.The expression level of miR-4281 in PBMCs of ITP patients was positively correlated with Tregs content and Treg/Th17,while negatively correlated with Th17 cell content.miR-4281 can promote the differentiation of Tregs cells and inhibit the differentiation of Th17.Furthermore,the expression level of miR-4281 is positively correlated with the expression level of FOXP3.Taken together,the expression level of miR-4281 is reduced in ITP.This decrease can inhibit the expression of FOXP3,cause Treg/Th17 immune imbalance,and promote the progression of ITP.