ABSTRACT
Objective:To investigate the value of a deep learning method based on MobileNet in classification of bedside chest radiograph and improvement of the work efficiency.Methods:A total of 6, 320 bedside chest radiographs from January 2017 to December 2022 in the Second Peoples′ Hospital of Changzhou were retrospectively collected. The included cases were divided into normal group (885 images), pneumonia group (1 927 images), pleural effusion group (373 images), and pneumonia with pleural effusion group (3 135 images). Three hundred and fifty images were selected as a validation set, while the remaining images were divided into a train set (4 775 images) and a test set (1 195 images) using simple randomization, by 8∶2 ratio. Two lightweight convolutional neural network models (MobileNetV1 and MobileNetV2) were used to construct a bedside chest radiograph classification model, based on which two fine-tuning strategies were designed. Four models were generated namely MobileNetV1_False (V1_False), MobileNetV1_True (V1_True), MobileNetV2_False (V2_False) and MobileNetV2_True (V2_True). In the first stage, a binary classification model was established to divide the images into normal and lesion groups; then a four-class classification model was established in the second stage, with which the images were divided into four groups: normal, pneumonia, pleural effusion and pneumonia with pleural effusion. Metrics for model performance evaluation including accuracy (Ac), precision (Pr), recall rate (Rc), F1 score (F1) and area under the receiver operating characteristic curve (AUC) were calculated.Results:In both the first and second stages, V1_True and V2_True had higher Ac, Pr, Rc, and F1 than V1_False and V2_False in both the training set and validation set; and the V1_True model outperformed the other three models in classification. The classification Ac of the V1_True model in the validation set was higher than that of radiologists in the first stage [95.71% (335/350) vs. 90.29% (316/350)] and in the second stage [93.43% (327/350) vs. 87.14% (305/350)]. The recognition time of V1_True model′s in the validation set of 350 bedside chest radiographs was significantly less than that of the radiologists (mean: 17 s vs. 300 min).Conclusions:V1_True is an optimal MobileNet model for classifying bedside chest radiographs. The application of this model in clinical practice may help to accurately identify the information of lung lesions from bedside chest radiographs in time, and may improve the work efficiency in the radiology department.
ABSTRACT
Objective:To investigate the relationship between the changes of default network topology properties of brain function and cognitive function in patients with end-stage renal disease (ESRD).Methods:A total of 31 patients with ESRD were enrolled in the Department of Nephrology, Changzhou Second Hospital Affiliated to Nanjing Medical University from January 2019 to December 2020, and 18 healthy persons were included in the same period as the control group.The cognitive function was evaluated with the Montreal cognitive assessment (MoCA) and trail making tests, and then the subjects were examined by resting-state functional magnetic resonance imaging (rs-fMRI). After preprocessing, the brain functional network was constructed and the topology properities of the network were calculated.The SPSS 20.0 software was used for statistical analysis.Independent sample t-test, chi square test and Pearson correlation analysis were used for data statistics. Results:(1) The score of MoCA in the ESRD group(23.37±1.77) was significantly lower than that in the healthy control group(27.94±1.13)( t=9.537, P<0.001). (2) The levels of Eglobal, Elocal, Cp and Sigma in ESRD group ((0.129±0.025), (0.148±0.040), (0.188±0.046), (1.593±0.650)) were significantly lower than those in healthy control group ((0.160±0.040), (0.212±0.024), (0.276±0.049), (2.004±0.864))( t=3.591, 7.474, 7.058, 2.034, all P<0.05). The Lp value of the ESRD group (8.131±1.905) was significantly higher than that of the control group (6.777±2.150)( t=2.583, P< 0.05). The node efficiency values of bilateral dorsolateral superior frontal gyrus, left middle frontal gyrus, bilateral posterior cingulate gyrus, right hippocampus, left superior marginal gyrus, bilateral angular gyrus and bilateral cuneate anterior lobe in ESRD group ((0.133±0.071), (0.201±0.047), (0.211±0.106), (0.175±0.066), (0.276±0.113), (0.122±0.146), (0.042±0.075), (0.171±0.027), (0.154±0.078), (0.240±0.095), (0.161±0.056))were lower than those in the healthy control group((0.312±0.075), (0.289±0.091), (0.277±0.132), (0.284±0.053), (0.368±0.063), (0.231±0.227), (0.120±0.162), (0.296±0.064), (0.310±0.186), (0.318±0.066), (0.286±0.103))( t=2.107-9.436, all P<0.05). (3)Pearson correlation analysis showed that the node efficiency values of bilateral posterior cingulate gyrus and right hippocampus in ESRD group were positively correlated with the score of MoCA( r=0.36, 0.49, 0.53, all P<0.05). Conclusion:The topological structure of brain functional network is abnormal in ESRD patients, which can affect the cognitive function of patients.
ABSTRACT
Objective:To evaluate the objective imaging markers of cognitive impairment in patients with end-stage renal disease by MRI intravoxel incoherent motion.Methods:A total of 40 patients with ESRD were enrolled in the Department of Nephrology, Changzhou Second Hospital Affiliated to Nanjing Medical University from January 2019 to August 2020, and 24 healthy controls were prospectively enrolled at the same time.All subjects performed with MRI scan were collected, and the slow apparent diffusion coefficient (ADC slow) of the corresponding brain regions were obtained .The cognitive function was evaluated by the Montreal cognitive assessment scale (MoCA). Two-sample t test was used to analyze the difference of ADC slow and cognitive score between the two groups.Pearson correlation analysis was performed among the cognitive function score of end-stage renal disease and ADC slow value. Results:(1) The score of the intelligence test scale in the ESRD group (23.30±1.76) was significantly lower than that of the healthy control group (27.92±1.00) ( P<0.01). The ADC slow values of bilateral frontal lobe, hippocampus, and insula brain areas (respectively(0.648±0.035), (0.633±0.043), (0.762±0.043), (0.756±0.042), (0.792±0.048), (0.776±0.054))in the ESRD group were significantly higher than those in the healthy control group ((0.600±0.039), 0.610±0.037, (0.725±0.059), (0.711±0.054), (0.740±0.063), (0.716±0.051)) ( P<0.01). (2) Pearson correlation analysis showed that the ADC slow values of bilateral insula and right hippocampus in the ESRD group were negatively correlated with MoCA scales ( r=-0.38, -0.38, -0.66, all P<0.05). Conclusion:ADC slow value in IVIM can better reflect the changes of cognitive function impairment in ESRD patients.
ABSTRACT
Three dimensional (3D) imaging and 3D printing have been applied in pancreatic malignancy management.The 3D imaging can not only be applied in pre-procedural evaluation and planning,but also it can be used for residents and fellows for training and education,and families of patients advsing.A model of pancreatic and its surrounding structures can be gotten by the 3D printing technique based on 3D imaging,which can not only recognize the anatomical structure,but also can make the guidance for the procedure.3D imaging and 3D printing technique have an important role in the management of pancreatic malignancy.In this review,we summarized the application of 3D) imaging and 3D printing in malignant neoplasm of pancreas.
ABSTRACT
Objective To investigate the altered spontaneous cerebral activity in patients with type 2 diabetic retinopathy (T2DR). Methods Twenty-one patients with T2DR and sixteen healthy control subject underwent rs-fMRI scans,and the data were analyzed statistically using regional homogeneity(ReHo)method to observe the change of ReHo value.Results Compared to the control group,the T2DR group showed significantly increased ReHo value in the right occipital gyrus,occipital gyrus,inferior occipital gyrus and lingual gyrus regions (t=5.30,P<0.05,voxel>30,AlphaSim corrected),and significantly decreased ReHo value in the left posterior cingulate,margin lobe,right inferior parietal lobule,superior temporal gyrus and hippocampus (t=-4.01,-4.86,P<0.05,voxel>30, AlphaSim corrected).Conclusion The patients with T2DR showed significantly increased ReHo values in the brain visual cortex and visual pathway that were associated with the injury of brain function regions.It is of important value to evaluate brain dysfunction in patients with T2DR using ReHo method of rs-fMRI.
ABSTRACT
Objective To explore the feasibility of “double low ”(low tube voltage and low contrast agent concentration)imaging technology of dual-source computed tomography (DSCT)in coronary angiography for low-weight patients.Methods 60 patients (BMI≤25 kg/m2 )who underwent coronary CT angiography (CCTA)on DSCT scanner were divided randomly into 2 groups:group A was “double low”group (n = 30,tube voltage = 80 kVp,using sinogram affirmed iterative reconstruction technology and 270 mg I/mL concentration of contrast agent);Group B was conventional group (n=30,tube voltage=120 kVp,using filtered back projection technology and 350 mg I/mL concentration of contrast agent).Adaptive cardiosequence prospective ECG-gated technology (CorAdSeq)was applied in both the groups,The collecting phase was 65%-75% RR interval (when heart rate<70 beats per mi-nute)or 40%-50% RR interval (when heart rate≥70 beats per minute).Subjective and objective methods were applied to evaluate the image quality.The image quality and.radiation dose were compared between the “double low”group and the conventional group by using SPSS 1 9.0 software.The differences between measurement data were analyzed by using independent samples t-test.Results The effective dose with size specific dose estimates (EDssde)of group A(1.22±0.31)mGy/cm was significantly lower than that of group B (3.44±0.80)mGy/cm with P <0.001.Contrast-noise ratio,signal-noise ratio and CT value of group A were signifi-cantly higher than those of group B (all P <0.05 ).Conclusion Compared with conventional DSCT coronary angiography,“double low”DSCT coronary scanning proposal for low-weight patients can significantly reduce the radiation dose and the amount of contrast agent,and the image quality can meet the needs of clinical diagnosis.
ABSTRACT
Objective To explore the image quality and radiation dose for coronary angiography using DSCT (dual-source computed tomography) under lower tube voltage and lower contrast agent concentration ('double low').Methods A total of 160 patients with suspected coronary artery disease underwent dual-source CCTA.80 patients (BMI < 25 kg/m2) were randomly assigned to group A (270 mg I/ml-80 kV) and B (350 mg I/ml-120 kV).The other 80 patients (25 ≤ BMI <30 kg/m2) were randomly assigned to group C (270 mg I/ml-100 kV) and D (350 mg I/ml-120 kV),sonogram affirmed iterative reconstruction technology and 270 mg I/ml contrast agent were applied to group A and C ; FBP technology and 350 mg I/ml contrast agent were applied to group B and D.Adaptive cardio-sequence prospective ECG-gated technology was applied in all patients,while the acquisition phase was 65%-75% RR interval (when heart rate < 75 beats / min) or 40%-50% RR interval at when heart rate ≥ 75 beats/min.Volume CT dose index (CTDIvol),dose length product (DLP),effective dose (E),size specific dose estimates (SSDE),image noise,CT value,signal-to-noise ratio (SNR),and contrast-to-noise ratio(CNR) were all evaluated for each patient.Results There was not statistical difference in scan length among groups A and B,C and D.There was statistical difference in tube current between group A and B (t =8.932,P <0.05).There was not statistical difference in tube current between group C and D (tCD =-1.212,P > 0.05).There was statistical difference in CTDIvol,DLP,E and SSDE between group A and B,C and D (tCTDIvol =-16.141,-11.440 ; tDLP =-17.454,-10.521 ; t =-17.444,-10.523 ;tSSDE =-13.032,-9.119,P <0.05).Compared to group B,the SSDE in group A decreased by 64.4%.Compared to group D,the SSDE in group C decreased by 39.3%.There was no statistical difference in image noise between group A and B,C and D (P > 0.05).There was statistical difference in CT value,SNR and CNR between group A and B (t =0.436,4.623,3.272,P < 0.05).There was no statistical difference in CT value and CNR between group C and D,while there was statistical difference in SNR between group C and D (t =2.981,P <0.05).There was no statistical difference in image quality scores between group A and B,C and D (P > 0.05).Conclusion Compared with conventional DSCT coronary angiography,'double low' DSCT coronary scanning solution can significantly reduce the radiation dose and contrast agent,and could provide the effective coronary images which meet the clinical diagnostic needs.
ABSTRACT
Objective To compare the iodine intake,image quality and radiation dose of dual-source CT coronary artery angiography between the low concentration isotonic contrast agent with iterative reconstruction technique of sinogram affirmed iterative rEconstruction (SAFIRE) and common concentration contrast agent with filtered back projection (FBP).Methods One hundred patients [58 men,42 women; age:(62± 11)y] were enrolled in this study.Fifty consecutive patients (Group A) were scanned with FBP and common concentration contrast agent and another 50 consecutive patients (Group B) were scanned with low concentration isotonic contrast agent and iterative reconstruction technique.The image quality were assessed by two experienced radiologists with a double blinded fashion in a five score scale.The attenuation of coronary artery,scan time,imaging noise,signal-to-noise ratio (SNR),contrast-to-noise ratio (CNR),radiation dose and iodine intake between these two groups were compared using two sample t test.The observer agreement for image quality was calculated using Kappa statistics.Results All examinations were completed successfully.There was no statistical difference (P>0.05) between two groups in clinical characteristics and scan parameters.There was no statistical difference for the mean subjective image quality of group A (4.4±0.7) and group B (4.3±0.8) (t=0.924,P>0.05).The Kappa value between two radiologists was 0.887 (P<0.01)on image quality.The mean attenuation of coronary artery segments was higher than 300 HU,especially in group A (386.1±51.5) HU,which was slightly higher than the group B (384.1±77.1) HU,but there was no statistical difference (t=0.157,P>0.05).The SNR and CNR in two groups did not have significant differences (P>0.05).The total iodine and iodine injection rates were 21.0 g,17.5 g/s in Group A and 16.2 g,13.5 g/s in Group B,respectively.The iodine intake was decreased by 22.9% in Group B.The effective radiation dose in Group B (1.09±0.19) mSv was significantly (t=20.260,P<0.01) lower than that in Group A (2.85±0.59) mSv,with the radiation dose reduced by 61.8%.Conclusion Low concentration isotonic contrast agent and iterative reconstruction technique can significantly reduce the radiation dose and iodine intake in CT coronary artery angiography and achieve the same image quality using common concentration contrast agent with FBP.
ABSTRACT
Objective To explore the application of CARE kV technique in the adult chest CT and the value of reducing radiation dose.Methods Sixty-nine patients were divided into two groups by random number generators:group A(39 cases) and group B(30 cases).Group A was examined by using CARE kV technique and group B was examined at routine 120 kV.CT dose index(CTDIvol),dose length product (DLP) and effective dose (E) were compared between the two groups,and analyzed the correlation between tube voltage selection and patient body mass index (BMI) of group A was analyzed.Results The average CTDIvol [(11.00 ± 3.89) mGy],DLP[(294.05 ± 91.17) mGy·cm] and E[(4.12 ± 1.28) mSv] of group A were lower than those of group B (16.64 ± 1.20) mGy,[(475.99 ± 41.16) mGy · cm],[(6.66 ±0.58) mSv].With statistically significant difference (t =-7.653,-10.151,-10.150,P < 0.05).Compared with routine 120 kV technique (group B),the CARE kV technique (group A) could reduce the total radiation dose about 38.14%.Compared obese patients(BMI≥28 kg/m2) with non-obese patients in group A and B,the mean E of non-obese patients was lower than that of obese patients in group A,which reduced the total E about 31.74% (t =4.322,P <0.05),while E in group B was no significant different between non-obese patients and obese patients.Conclusions In adult chest CT,CARE kV technique can select optimum scanning voltage automatically according to the patients with different BMI and anatomical regions,which can reduce the overall radiation dose while maintaining image quality.
ABSTRACT
Objective To study the image quality and radiation dose in dual-source CT cerebral angiography with prospective ECG-triggered sequence mode (step-and-shoot,SAS).Methods A total of forty-three patients with clinically suspected cerebral vascular disease underwent cerebral CT angiography with prospective ECG-triggering (step-and-shoot,SAS).Data acquisition was at 60% R-R interval of the ECG presentation mode.The post-processing included maximum intensity projection (MIP),multiplanar reformation (MPR) and volume rendering (VR).The CTA image quality,radiation dose and rates of excellent images were evaluated.Results The CTA image quality score was 4.72 ± 0.50 and 97.7% (42/43) patients had excellent CTA images.The average effective dose of SAS-CTA was (0.22 ± 0.01 )mSv,which was lower by 76.31% than that of DE-CTA.Conclusions Prospective ECG-triggering sequence could be used in cerebral angiography with a significant reduction in radiation dose and diagnostic image quality.
ABSTRACT
Objective To analyze the clinical application of monoenergetic technique of dual-energy CT in removing metal artifacts for patients with fractures fixed with metal fixer.Methods Fofly-five patients with fractures fixed with metal fixer underwent dual-energy CT scanning for the fractures.Two different data were collected in one-time scanning using dual-energy scanning sequence.With monoenergetic technique,two different data at 100 and 140 kilovolts were used for subtraction to removing metal artifacts based on different densities.Raw data were reconstructed with monoenergetic technique(group A)and conventional simulation method(group B),respectively.And,all data were reconstructed with multiplanar reconstruction (MPR),volume rendering(VR)and maximum intensity projection(MIP),respectively.Wilcoxon signed rank test was applied for the comparison of imaging quality and artifacts between the two groups.Results There were fewer artifacts on the images due to the application of monoenergetic technique in dual-energy CT scanning.In group A,the rate of high-quality images reached to 91.9%(124/135);and,in group B,it was 59.3%(80/135).There were statistical diference between the two groups(Z=-12.541.P<0.01).The images without artifact reached to 89.6%(121/135)in group A;whereas,it was 45.2%(61/135)in group B.There was statistical difference between the two groups(Z=-11.910,P<0.01).Conclusion Using monoenergetic technique,metal artifacts were removed effectively and the fine structure of fracture was clearly displayed.
ABSTRACT
Objective To compare the quality and radiation doses of coronary artery angiography under the natural heart rate condition between Flash spiral heart mode and prospective electrocardiogramtriggering sequence mode using dual-source,in order to choose personalized low doses of coronary artery scanning mode.Methods Sixty patients who underwent coronary angiography(CTA)on a 128-slice,dualsource CT scanner were divided into 2 group i.e,group A(27cases)and group B(33 cases).Flash spiral heart scan mode was employed for group A.Inclusion criteria included:heart rate75 bpm),date acquisition was set at 30%-50%of the R-R interval. (3)At the arrhythmias,premature beat,fibrillation atrial,date acquisition was set at 20%-90%of the R-R interval.In both gronps,patients with a BMI≥25.0kg/m2 were examined with a tube voltage of 120 kV.while the other patients with a BMI<25.0 kg/m2 were examined with a tube voltage of 100 kV.The BMl was(24.6±1.0)kg/m2 in group A,while that was (24.6±0.9)kg/m2 in group B.In both groups,all images were transferred to the workstation for further processing and analysis.The imaging quality of coronary artery segments and the radiation dose were compared with t test.Results A total of 336 coronary artery segments were evaluated in group A and 412 segments were evaluated in group B.The imaging quality of coronary artery segments were scored.Excellent or good was achieved in 98.2%(330 of 336)artery segments in group A,and that was 98.1%(404 of 412)in group B.There was no statistical difference in imaging quality between the two groups(t=0.513,P=0.608).The average effective dose was(0.74±0.29)mSv in group A,whereas that was(3.67±1.37)mSv in group B.There was a significant difference between the two groups(t=-10.858,P=0.000).Conclusions The personalized low doses coronary artery scanning mode can substantially reduce radiation damage while preserving good imaging quality.
ABSTRACT
Objective To observe the CT features of ovary Brenner tumor. Methods CT manifestations of 5 patients with ovary Brenner tumor confirmed with pathology and clinical follow-up were retrospectively analyzed, and the masses were described for location, size, configuration, enhancement pattern, presence of calcification and metastatic spread. Results There were 7 tumors in 5 patients, 3 patients had unilateral tumors and 2 had bilateral Brenner tumors (left side 3 and right side 4), with tumor size ranging from 1.52 to 16.25 cm (mean 7.36 cm). Five masses in 4 patients were benign, 2 (bilateral tumors in 1 patient) were malignant. All tumors had well-defined margin. One patient with bilateral benign tumors had large pleural effusion and seroperitoneum. Five tumors in 4 patients (5/7, 71.43%) were solid and had calcification, 2 tumors in 1 patients (2/7, 28.57%) were mainly cystic, with septa in the tumors. The solid part of all tumors were inhomogeneous and had mild enhancement. Conclusion CT findings of ovary Brenner tumor have some characteristics. Combining with clinical manifestations, CT is helpful for the diagnosis of ovary Brenner tumor.
ABSTRACT
Objective To evaluate the role of dynamic contrast-enhanced magnetic resonance imaging (MRI) in the diagnose of prostatic cancer and benign prostatic hyperplasia (BPH), and to determine the correlation between dynamic MRI findings with angiogenesis.Methods Thirty-two cases of prostatic cancer and 40 cases of BPH underwent dynamic contrast-enhanced MRI.All the patients in this study were diagnosed by histopathology.The results of dynamic contrast-enhanced MRI were evaluated by early-phase enhancement parameters and time-signal intensity curves (SI-T curves), and the curves were classified according to their shapes as type Ⅰ, which had steady enhancement; type Ⅱ, plateau of signal intensity; and type Ⅲ, washout of signal intensity.The pathologic specimens of region of interest (ROI) were obtained, and HE staining, immunohistochemical vascular endothelial growth factor (VEGF), and microvessel density (MVD) measurements were performed.The relationships among dynamic contrast-enhanced MRI features, VEGF, and MVD expression were analyzed.Results In the early-phase enhancement parameters of dynamic contrast-enhanced MRI, onset time,maximum signal intensity, and early-phase enhancement rate differed between prostatic cancer and BPH(P
ABSTRACT
Objective To investigate the MR imaging findings of colonic carcinoma and the diagnostic value of MRI. Methods Multi-planar and multi-sequence MRI scanning, before and after contrast enhancement, were performed in 40 patients with colonic cancer. The patients were fasted for 12 hours, prepared with clean clysis or senna at night before study, given 10 mg of anisodamine 10 minutes before study, and then infused with 800~1 000 ml physiological saline immediately before study by anus. Dukes staging and resectability evaluation were made in 32 patients before surgery and meanwhile the results were compared with pathology. Results Colonic anatomy and surrounding organs were clearly demonstrated on MRI in 40 patients with colonic cancer, particularly in recta and sigmoid flexure. The tumours showed iso-intensity on T 1WI, iso-intensity or slight high-intensity signal on T 2WI, and high-intensity signal on SPIR. Remarkable enhancement was seen in 35/40 (87.5%). Invasion of surrounding organs occurred in 8/40(20.0%)and MRI revealed 6; Meanwhile, MRI revealed lymph node metastasis in 8 out of 12 cases. 32 patients were regarded as resectable before surgery, and 8 patients as unresectable. Four patients were overestimated, the accuracy of preoperative evaluation for the resectability was 87.5%, and the detecting rate of colonic cancer was 100.0%. Conclusion MRI can clearly show the colonic wall thickness, anatomic structure and surrounding anatomy. For the diagnosis of colonic cancer, MRI can not only demonstrate all its morphologic features, such as mass, thickened wall, and invasion of adjacent organs, but also swollen lymph node and metastasis in abdominal cavity. MRI is very helpful in the diagnosis, staging, and respectability evaluation of colonic cancer.