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 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.

Purpose To explore the application value of auto-prescription technique in CT portal venography and liver volume measurement.Materials and Methods A total of 180 patients who underwent contrast-enhanced abdominal CT in the First Affiliated Hospital of Dalian Medical University from April to November 2022 were prospectively collected and randomly divided into two groups,with 90 cases in each group.Group A(experimental group)was scanned with auto-prescription technique.Group B(conventional group)was scanned with 120 kVp,and the other imaging parameters kept the same as group A.In group A,the correlation between body mass index and tube voltage(kVp)recommended by auto-prescription technique was analyzed.The following values were compared between group A and group B,including the volume CT dose index,dose length product,CT value of portal vein and liver,standard deviation,contrast-to-noise ratio(CNR),image quality and subjective score of automatic segmentation.The liver volumes of group A and B were measured by Liver Segmentation software.The volumes measured by automatic segmentation software(Vsoft)were recorded,and then were compared with those segmented manually(gold standard)(Vstnd).Results Compared with group B,the volume CT dose index and dose length product in group A decreased by 22%and 25%(t=-4.654,-5.403,both P<0.05),respectively.The CT value and CNR of portal vein and liver in group A were higher than those in group B(t=4.857-6.011,all P<0.05).There was no significant difference in standard deviation,Vstnd and Vsoft between group A and group B(P>0.05).The consistency of subjective scores between the two observers was good(ICC=0.773-0.856,all P<0.05).The subjective scores of two-dimensional and three-dimensional images,as well as the automatic software segmentation scores in group A were higher than those in group B(Z=897.500,1 287.500,912.721,all P<0.05).The recommended kVp value in group A was positively correlated with body mass index(r=0.674,P<0.001).Conclusion The application of auto-prescription technique in CT portal venography can ensure the image quality,measure the liver volume accurately,and reduce the patient's radiation dose individually.

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.

Purpose To explore the application value of auto-prescription technique in CT portal venography and liver volume measurement.Materials and Methods A total of 180 patients who underwent contrast-enhanced abdominal CT in the First Affiliated Hospital of Dalian Medical University from April to November 2022 were prospectively collected and randomly divided into two groups,with 90 cases in each group.Group A(experimental group)was scanned with auto-prescription technique.Group B(conventional group)was scanned with 120 kVp,and the other imaging parameters kept the same as group A.In group A,the correlation between body mass index and tube voltage(kVp)recommended by auto-prescription technique was analyzed.The following values were compared between group A and group B,including the volume CT dose index,dose length product,CT value of portal vein and liver,standard deviation,contrast-to-noise ratio(CNR),image quality and subjective score of automatic segmentation.The liver volumes of group A and B were measured by Liver Segmentation software.The volumes measured by automatic segmentation software(Vsoft)were recorded,and then were compared with those segmented manually(gold standard)(Vstnd).Results Compared with group B,the volume CT dose index and dose length product in group A decreased by 22%and 25%(t=-4.654,-5.403,both P<0.05),respectively.The CT value and CNR of portal vein and liver in group A were higher than those in group B(t=4.857-6.011,all P<0.05).There was no significant difference in standard deviation,Vstnd and Vsoft between group A and group B(P>0.05).The consistency of subjective scores between the two observers was good(ICC=0.773-0.856,all P<0.05).The subjective scores of two-dimensional and three-dimensional images,as well as the automatic software segmentation scores in group A were higher than those in group B(Z=897.500,1 287.500,912.721,all P<0.05).The recommended kVp value in group A was positively correlated with body mass index(r=0.674,P<0.001).Conclusion The application of auto-prescription technique in CT portal venography can ensure the image quality,measure the liver volume accurately,and reduce the patient's radiation dose individually.

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 effects of the auto-prescription technique combined with noise index (NI) and adaptive statistical iterative reconstruction algorithm-veo (ASIR-V) on the radiation dose and image quality in the computed tomographic angiography (CTA) of the deep inferior epigastric artery (DIEA).Methods:The data of 150 cases who underwent DIEP CTA in the First Affiliated Hospital of Dalian Medical University were prospectively collected, and were randomized into groups A, B, and C, with 50 cases in each group. For group A, the conventional tube voltage was 120 kVp, and images were reconstructed using 40% ASIR-V. For group B, the tube voltage based on the auto-prescription technique was adopted, and images were reconstructed using 40%, 60%, and 80% ASIR-V (corresponding to the three subgroups B1-B3, respectively). For group C, the tube voltage based on the auto-prescription technique was employed, NI = 13, and images were reconstructed by 40%, 60%, and 80% ASIR-V (corresponding to subgroups C1, C2, and C3, respectively). Subsequently, the CT and SD values of the originating femoral artery of DIEA and the rectus abdominis at the same level as DIEA were measured on axial images of each group, the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated, and the radiation doses of the three groups were recorded. Finally, the image quality and noise of DIEA in each group were independently and subjectively scored by two radiologists in a blinded manner.Results:Compared to group A, groups B and C exhibited higher CT values of femoral artery ( t= -5.18, -5.17, -5.06, -6.08, -6.08, -6.07, P < 0.05) and higher image SNR and CNR ( P < 0.05), which increased gradually with the upgrading of ASIR-V ( F = 55.45, 49.70, 53.47, 68.89, P < 0.05). The two radiologists offered consistent subjective evaluations (Kappa = 0.76-0.92, P < 0.05). The images of groups B and C displayed more distinct perforating branches and emerging points of DIEA and received higher subjective scores of intramuscular course compared to those of group A. The subjective scores of image noise in subgroups B2, B3, and C2 were not significantly different from those in group A. The radiation doses in groups B and C decreased by 15.10% and 52.85%, respectively compared to those in group A ( H = 75.21, P < 0.05). Conclusion:The combination of the auto-prescription technique with NI = 13 and 60% ASIR-V can reduce the radiation dose by 52.85% while ensuring a clear display of DIEA.

Objective To explore the application value of deep learning image reconstruction(DLIR)algorithm in low-dose deep inferior epigastric artery(DIEA)computed tomography angiography(CTA).Methods A total of 59 patients undergoing DIEA CTA were prospectively selected and randomly divided into two groups:group A(29 cases)and group B(30 cases).Group A was the conventional radiation dose group(tube voltage 120 kVp),the tube current was dose modulation 3,and the iterative reconstruction algo-rithm was Karl 5.Group B was the low radiation dose group(tube voltage 120 kVp),the tube current was dose modulation 2,with DLIR reconstruction algorithm,and four levels of DLIR(1-4)were reconstructed and labeled as groups B1 to B4.The mean tube current,vol-ume CT dose index(CTDIvol),and dose length product(DLP)of group A and group B were recorded,and the effective dose(ED)was calculated.The CT and standard deviation(SD)values of the external iliac artery and psoas major muscle were measured on axial images of each group,and signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)were calculated.Two observers using double-blind method independently evaluated the DIEA on volume rendering(VR)and maximum intensity projection(MIP)images of each group using a 3-point scale,and a 5-point scale was used to evaluate the overall image quality on axial images.Results Compared with group A,the mean tube current and ED in group B were decreased by 39.33％and 44.09％,respectively(P＜0.05).With the increasing of DLIR level from groups B1 to B4,the SD value of the image gradually decreased,while SNR and CNR gradually increased(P＜0.05).The CNR in groups B3 and B4 was higher than that in group A(P＜0.05).The subjective scores of the two observers were consistent(Kappa value 0.779-0.889,P＜0.05),and there was no statistical significance between group A and group B in the perforating vessels from the DIEA,intramuscular course,and the point of emergence(P＞0.05).With the increase in DLIR level,the subjective score of the overall image quality from group B1 to group B4 showed a trend of first increasing and then decreasing,and the score of group B3 was the highest(4.50±0.51)points,which had no statistical significance compared with group A(4.45±0.51)points(P＞0.05).Conclusion DLIR can effectively reduce the radiation dose of the DIEA CT A scan and ensure the image quality,among which DLIR 3 is the recommended best reconstruction grade.

Objective To investigate the influence of spectral CT monochromatic imaging technique on image quality and liver vol-ume measurement in computed tomography portal venography(CTPV).Methods A total of 120 patients who underwent contrast-enhanced abdominal CT examination were prospectively selected and randomly divided into group A and group B.The group A(n=60)was scanned with conventional parameters such as 120 kVp,Smart mA mode,and image reconstruction of 60％adaptive statistical iterative reconstruction-Veo(ASIR-V);while the group B(n=60)was with instantaneous switching 80/140 kVp,gemstone spectral imaging(GSI)Assist mode,and image reconstruction of 60％ASIR-V.In group B,six subgroups of images from 75 to 50 keV(with 5 keV interval)were recorded as subgroups B1 to B6.On the axial images,the CT values and standard deviation(SD)values of por-tal vein,liver and erector spinae were measured,and then the contrast-to-noise ratio(CNR)of portal vein and liver were calculated.The Liver Segmentation software was used to segment the liver in groups A and B,the liver volumes by automatic segmentation and manual segmentation(golden standard)were recorded,and then the volume difference rate was calculated.The overall image quality and automatic liver segmentation results were evaluated by two radiologists using a 5-point scale.Results In terms of overall image quality,subgroup B6 achieved the highest score and was superior to group A(P＜0.001).In terms of liver segmentation,subgroup B3 had the highest score and was superior to group A(P＜0.001).With the decrease of keV,the CT values,SD values and CNR of portal vein and liver in group B were gradually increased(P＜0.05),in which subgroup B6 was higher than that in group A(P＜0.001).The volume difference rate initially decreased and then increased with the decrease of keV.Except for subgroups B2 and B3,the differences were statistically significant between other subgroups and group A(P＜O.001),and the subgroup B3 had the lowest volume difference rate.Conclusion Spectral CT monochromatic ima-ging technique has an influence on CTPV image quality and liver volume measurement.The 50 keV images are the best for displaying portal vein,and the automatic liver segmentation volume of 65 keV images is closest to the real liver volume.

Objective To explore the application value of 1 024×1 024 reconstruction matrix combined with iterative reconstruc-tion algorithm(Karl)in deep inferior epigastric artery(DIEA)computed tomography angiography(CTA).Methods A total of 40 patients who underwent DIEA CTA were prospectively selected and the original data were reconstructed by grouping.Group A was reconstructed using a conventional 512×512 matrix combined with Karl 5 grade.Group B was reconstructed using 1 024×1 024 recon-struction matrix combined with Karl 5,7,and 9 grades,respectively,and 3 subgroups B1-B3 were obtained.The CT and standard devia-tion(SD)values of the external iliac artery and psoas major muscle were measured on axial images,and signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)were calculated.A 3-point scale was used to evaluate the perforating vessels from the DIEA,intramuscular course,point of emergence,superficial inferior epigastric artery(SIEA)and superficial inferior epigastric vein(SIEV)on volume ren-dering(VR)and maximum intensity projection(MIP)images by two observers,and a 5-point scale was used to evaluate the overall image quality on axial images.Results With the increase of Karl grade in groups B1 to B3,the SD value of the external iliac artery decreased gradually(P＜0.05),while SNR and CNR increased gradually(P＜0.05).The SD values of the external iliac artery in group B2 and group B3 were lower than those in group A(P＜0.05),and SNR and CNR were higher than those in group A(P＜0.05).There was a good consistency in the subjective evaluation between the two observers(Kappa values=0.773-0.872,P＜0.05).The perforating vessels from the DIEA,intramuscular course,point of emergence,SIEA and SIEV display and overall image quality subjective scores of group B2 and group B3 were better than those of group A(P＜0.05),and the scores of group B2 showed the greatest improvement.Conclusion The 1 024 × 1 024 reconstruction matrix combined with the Karl 7 reconstruction algorithm can optimize the image quality and improve the display of the DIEA and perforator microvessels.