Volume 101 ·  Issue 07 ·  2021

Artificial intelligence (AI) is a hot point in clinical medicine research. In recent years, AI has played an important role in recognizing the lesion, improving the diagnostic accuracy and assessing the diagnostic efficacy. To accelerate the pace of AI industry, it should be a first thing to improve relevant industrial policies and regulations and to build a transformation platform for industry-university-research. All these will contribute to the standardization in further development of medical imaging AI industry.

To investigate the diagnostic value of radiomics model based on plain CT scan of peripheral coronary artery adipose tissue for non-calcified plaque.

The image data of 461 patients undergoing coronary CT angiography (CCTA) in the Department of Radiology of the First Affiliated Hospital of Suzhou University from August 1,2019 to July 31,2020 were retrospectively analyzed. Two hundred and six cases (355 branches) with non-calcified plaques, and 255 cases (510 branches) with no coronary artery disease were detected by CCTA. The regions of interest (ROI) of the pericoronary adipose tissue were segmented on the plain CT scan images (coronary calcification score (CCS) sequence). The coronary ROI was determined by selecting the coronary artery with a length of 40 mm and starting at 10 mm from the opening of the coronary artery, and the pericoronary adipose ROI was generated automatically. The pericoronary fat attenuation index (FAI) was then performed, and the radiomics features were extracted. The 865 coronary arteries were divided into the training group (n=606) and the testing group (n=259) at a ratio of 7∶3, and the radiomics model was carried out. The receiver operating characteristic (ROC) analysis was used to assess the FAI value and the diagnostic efficacy of the radiomics model for non-calcified plaque.

A total of 1 692 features were extracted from images of pericoronary adipose based on plain scan. All features were screened by using max-relevance and min-redundancy (mRMR) and least absolute shrinkage and selection operator (LASSO), and 14 features were selected for the establishment of the radiomics model. The accuracy, sensitivity, specificity and area under the curve (AUC) of the model in distinguishing patients with non-calcified plaque and those without coronary stenosis in the testing group were 70.3%, 63.2%, 75.2% and 0.75, respectively.

The radiomics model based on plain CT scan of the pericoronary adipose tissue had good diagnostic efficacy for non-calcified plaque.

To investigate the role of artificial intelligence-based coronary CT blood flow reserve score (FFR_CT) in assessing hemodynamic relevance in patients with deep myocardial bridge (MB) of the left anterior descending coronary artery.

A total of 113 patients diagnosed with deep MB of the left anterior descending coronary artery by coronary CT angiography (CCTA) at the Department of Radiology of Tongji Hospital Affiliated to Tongji University from January 2017 to December 2019 were retrospectively analyzed. The location, length, depth, and degree of systolic compression of the MB were measured. The artificial intelligence-based coronary FFR_CT software was employed to calculate the FFR_CT value of the deep MB of the left anterior descending coronary artery. With the boundary of 0.80, all patients were divided into FFR_CT normal group (FFR_CT>0.80) and FFR_CT abnormal group (FFR_CT≤0.80), and the relationship between FFR_CT abnormality and the location, length, depth, and degree of systolic stenosis of the deep MB of the left anterior descending branch was analyzed. The effectiveness of the receiver operating characteristic (ROC) curve in predicting FFR_CT abnormalities was measured by using ROC curve to analyze the length, depth, and degree of systolic stenosis of MB.

There were no significant differences in age, gender and high-risk factors between FFR_CT normal group (n=79) and FFR_CT abnormal group (n=34) (P>0.05). In terms of clinical symptoms, unstable angina, asymptomatic myocardial ischemia, stable angina in the FFR_CT normal group were 15.2%, 41.8%, 32.9%,respectively, while 32.4%, 23.5%, 35.3% in the FFR_CT abnormal group,respectively. Except for unstable angina (χ²=4.32,P=0.038), there were no significant differences in asymptomatic myocardial ischemia and stable angina between the two groups (χ²=3.42, 0.06, P>0.05). The length of deep MB was about (36±5) mm in the FFR_CT normal group and (44±5) mm in the FFR_CT abnormal group, respectively. The difference between the two groups was statistically significant (t=-7.703, P<0.001). The ROC curve showed that the optimal critical value of the length of the deep MB was 39.7 mm, the area under the curve was 0.88 (95%CI:0.81-0.95, P<0.001), and the accuracy rate of diagnosing FFR_CT ≤0.80 was 82.3%.

FFR_CT value is of great value in the evaluation of hemodynamics in patients with deep myocardial bridge of left anterior descending coronary artery, and the length of deep myocardial bridge is an important factor affecting FFR_CT value.

To investigate the segmentation effects of the deep learning method on CT in the arterial phase and venous phase respectively by using subjective and objective evaluation system, and to investigate the factors that affect the difference between arterial phase and venous phase pancreas segmentation and the related factors affecting the venous pancreas segmentation.

A total of 218 cases of pancreatic CT scan data in the Department of Radiology of Peking Union Medical College Hospital from January to November 2019 were retrospectively collected. Each case contained images of arterial and venous phases, and the data were randomly divided into training set (139 cases), validation set (20 cases) and test set (59 cases) according to the ratio of the training and verification set to the test set of 7∶3. The two-stage global local progressive fusion network was trained on the training set, the model parameters of the optimal segmentation effect were found on the validation set, and the test set was predicted and the results were evaluated subjectively and objectively. The Likert 5-point scale was used for subjective evaluation based on the critical regions between pancreas and peripheral organs, while the Dice similarity coefficient (DSC) was used for objective evaluation. The paired t test or Wilcoxon paired rank test was used to compare the differences of subjective and objective scores of the arterial phase and venous phase.

For the critical regions of the pancreas at the duodenum, duodenal jejunal flexure, left adrenal gland, portal vein, superior mesenteric vein, splenic artery and splenic vein, the median number of subjective scores in arterial phase were 4(4, 5), 5(4, 5), 5(4, 5), 4(4, 5), 5(4, 5), 5(5, 5) and 4(3, 5)points respectively, the median number(first quartile, third quartile) of subjective scores in venous phase were 4(4, 4), 5(4, 5), 5(4, 5), 5(4, 5), 5(5, 5), 4(3, 4) and 5(5, 5) points respectively,there were statistically significant differences of the median number(first quartile, third quartile) of the subjective scores between the arterial and venous phase for the critical regions of the pancreas at the organs described above (all P<0.05). DSC in the venous phase was slightly higher than that in the arterial phase and the difference was not statistically significant (DSC: 0.932 vs 0.921,P=0.952). Subjective scores in venous phase of the pancreas and duodenal jejunum, stomach, and left adrenal gland with fat gaps were 4.64,4.68 and 4.63 points respectively, and those of the group without fat gaps were 4.56,4.62 and 4.56 points respectively, there were statistically significant differences of the subjective scores in venous phase of the groups with fat gaps or not between the pancreas and the organs described above (t=2.147, 2.112, 2.277, all P<0.05). Except the spleen, the density differences between the critical regions of the pancreas and other surrounding organs were statistically significant in arterial phase and venous phase segmentation (allP<0.05).

Dual-phase CT was used to construct a deep learning automatic pancreas segmentation model, and the segmentation effect was evaluated subjectively and objectively. Subjective evaluation was helpful to improve the ability to segment the critical regions of the pancreas in the future.

To evaluate the diagnostic value of the lung nodule classification and segmentation algorithm based on deep learning among different CT reconstruction algorithms.

Chest CT of 363 patients from June 2019 to September 2019 in Radiology Department of Tianjin Medical University Chu Hsien-I Memorial Hospital were retrospectively collected in this study, each of which consisted of images by three different reconstruction methods (lung reconstruction, mediastinal reconstruction, bone reconstruction).These collected data were used as testing set and a total of 4 185 Chest CTs including the public data set and the constructed private data set were used as the training set. A model combines 3D deep convolutional neural network and recurrent neural network under a multi-task joint learning algorithm for lung nodule classification and segmentation were constructed. The well-trained method was tested on 363 test cases using two metrics, i.e., the accuracy of the density classification and the Dice coefficient of nodule segmentation. The performances under three reconstruction methods were statistically analyzed according to the variance analysis among three different reconstruction methods.

The average classification accuracies of the nodule under three reconstruction methods were 98.67%±5.70%, 98.38%±6.61% and 97.89%±7.32%. Specifically, the accuracies of the solid nodules under three reconstruction methods were 98.79%±5.58%, 98.49%±6.89% and 97.90%±7.41% and the accuracies of the sub-solid nodules were 97.57%±10.19%, 98.52%±7.77% and 98.52%±7.77%. There was no significant difference in the classification accuracy of pulmonary nodules under three different reconstruction algorithms (all P>0.05). The average Dice coefficients of nodule segmentation was 79.87%±5.78%, 79.02%±6.04% and 79.31%±5.95%. There was no significant difference in the average Dice coefficients of nodule segmentation under three different reconstruction algorithms (allP>0.05).

Deep learning algorithm which combined with 3D convolutional neural network and recurrent neural network has demonstrated relatively stable in classification and segmentation of lung nodules under different CT reconstruction method.

To investigate the diagnostic efficacy and potential application value of deep learning-based chest CT auxiliary diagnosis system in emergency trauma patients.

A total of 403 patients, including 254 males and 149 females aged from 16 to 100 (50±19) years, who received emergency treatment for trauma and chest CT examination in the Eastern Theater General Hospital from September 2019 to November 2019 were retrospectively analyzed. Dr. Wise Lung Analyzer′s chest CT auxiliary diagnosis system was applied to detect 5 types of injuries, including pneumothorax, pleural effusion/hemothorax, pulmonary contusion (shown as consolidation and ground glass opacity), rib fractures, and other fractures (including thoracic vertebrae, sternum, scapula and clavicle, etc.) and 6 other abnormalities (bullae, emphysema, pulmonary nodules, stripe, reticulation, pleural thickening). The diagnostic reference standards were labeled by two radiologists independently. The sensitivity and specificity of the auxiliary diagnosis system were evaluated. The imaging diagnostic reports were compared with the results of the auxiliary diagnosis system, and the diagnostic consistency between the two was calculated by using the Kappa test.

According to the reference standards, among the 403 patients, 29 were pneumothorax, 75 were pleural effusion/hemothorax, 131 were pulmonary contusion, 124 were rib fractures, and 63 were other fractures. The sensitivity and specificity of the auxiliary diagnosis system for detection of pneumothorax, pleural effusion/hemothorax, rib fractures, and other fractures were 96.6%, 97.6%, 80.0%, 99.7%, 99.2%, 83.9%, 84.1%, and 99.7%, respectively. The sensitivity of detecting lung contusion was 97.7%. There was a high consistency between the auxiliary diagnosis system and imaging diagnosis in the diagnosis of injuries, in which the kappa values of pneumothorax, pleural effusion, rib fracture and other fractures were 0.783, 0.821, 0.706 and 0.813, respectively (all P<0.001). Two cases of pneumothorax, three cases of pleural effusion/hemothorax, nine cases of rib fractures, and six cases of other fractures missed by imaging diagnosis were all detected by the auxiliary diagnosis system. The detection sensitivity of the auxiliary diagnosis system was higher for emphysema, pulmonary nodules and stripe (all>85%), but lower for bullae, reticulation and pleural thickening.

The deep learning-based chest CT auxiliary diagnosis system could effectively assist chest CT to detect injuries in emergency trauma patients, which was expected to optimize the clinical workflow.

To investigate the outcome of hip fracture patients associated with hyponatremia.

From January 2012 to December 2016, the data of 1 001 elderly patients with hip fracture treated in the Seventh Medical Center of PLA General Hospital were analyzed retrospectively. According to the level of serum sodium, the patients were divided into hyponatremia group (sodium<135 mmol/L) and non-hyponatremia group (sodium≥135 mmol/L), in which≥130-135 mmol/L was mild hyponatremia, ≥125-130 mmol/L was moderate hyponatremia, and<125 mmol/L was severe hyponatremia. The length of hospital stay, incidence of complications and mortality were compared between patient with hyponatremia and without; and the above three indexes between patients with mild hyponatremia and moderate severe hyponatremia were also analyzed.

There were 126 patients with hyponatremia, including 98 with mild hyponatremia (9.8%, 98/1 001), 18 with moderate hyponatremia (1.8%, 18/1 001), and 10 with severe hyponatremia (1.0%, 10/1001); of those patients, there were 53 males and 73 females, aged 82(78, 87) years. There were 875 patients in non-hyponatremia group, including 274 males and 601 females, and aged 81(75, 85) years. The mortality of 30 days, 1 year and overall were 9.5% (12/126), 26.2% (33/126) and 40.5% (51/126) in hyponatremia group, 3.5% (31/875), 14.2% (124/875) and 27.7% (242/875) in non-hyponatremia group, respectively; the differences between the two groups were all statistically significant (χ²=9.583, 12.031, 9.839, all P<0.05). After adjusting the age, sex, fracture type and coexisting diseases, hyponatremia was an independent risk factor for 30 days, 1 year and overall mortality, theOR(95%CI) was 2.744(1.331-5.654), 1.975(1.251-3.119), 1.637(1.099-2.440), respectively (all P<0.05). The 30 days, 1 year and overall mortality for mild hyponatremia were 6.1%, 24.5% and 37.8%, respectively; and those were 21.4%, 32.1% and 50.0% in patients with moderate and severe hyponatremia, respectively; only the difference for 30 days mortality was statistically different between two groups (χ²=4.278,P=0.039). The length of hospital stay for mild hyponatremia patients were 11 (9,16) d, and it was 12(10,18) d in patients with moderate and severe hyponatremia patients, and there was no significant difference between the two groups (Z=1.613, P=0.107). The incidence of complications was 22.9% (200/875) in non-hyponatremia group and 32.5%(41/126) in hyponatremia group, and there was significant difference between the two groups (χ²=5.649, P=0.017).

Compared with non-hyponatremia, patients with hyponatremia have higher incidence of perioperative complications, longer hospital stay and higher mortality. With the increasing degree of hyponatremia, the above indicators tend to be serious.

To explore the factors that affect the fluctuation of intraoperative regional cerebral oxygen saturation (rSctO₂) in elderly patients undergoing laparoscopic hysterectomy.

A retrospective analysis of 39 elderly patients undergoing elective laparoscopic hysterectomy in Yale New Haven Hospital from October 2016 to February 2017 was performed. Factors including patients′ demographic data, past medical history, intraoperative monitoring index and rSctO₂ index (baseline, maximum, minimum, maximum-baseline, baseline-minimum) were recorded. Pearson correlation analysis was used to analyze the correlation between rSctO₂ indexes and preoperative and intraoperative factors. Independent sample t-test was used to compare the differences of rSctO₂ indexes between hypertension group and the group without hypertension, as well as diabetes group and the group without diabetes. Taking diabetes as the stratification factor, the relationship between rSctO₂and factors including age, body mass index, hypertension were analyzed using multivariate analysis.

The age of the patients was (67.6±5.6) years old. The baseline value of rSctO₂ was (71.2±3.9)% in the left frontal lobe and (70.8±4.1)% in the right frontal lobe. Results from univariate analysis showed that the left and right maximum-baseline value of rSctO₂ were (12.9±4.5)% and (12.8±4.4)%, which were significantly lower than (16.3±4.2)% and (16.2±4.9)% in the group without hypertension (t=2.329, 2.266, all P<0.05). The right baseline-minimum in the group with hypertension was (5.3±3.7)%, which was significantly higher than (2.7±4.2)% in the group without hypertension (t=-2.061, P<0.05). The left minimum value of rSctO₂ in diabetic group was (64.4±4.2)%, and the maximum value of rSctO₂on the left and right sides were (81.7±4.0)% and (81.9±4.5)%, which were lower than (68.4±4.1)%, (86.7±4.2)% and (86.0±3.7)% in the group without diabetes (t=2.623, 3.297, 2.850, all P<0.05). The results of multivariate analysis showed the influence of age (t=2.866, P<0.05) and hypertension on left maximum-baseline (t=-4.530, P<0.01) was statistically significant. The influence of hypertension on right maximum-baseline was statistically significant (t=-4.629,P<0.01).

Preoperative diabetes and hypertension are factors significantly affecting the intraoperative rSctO₂ of elderly patients with laparoscopic hysterectomy.

To analyze the characteristic changes of corneal nerve fibers in patients with Parkinson's disease (PD) by corneal confocal microscopy (CCM) and investigate the association of corneal nerve fiber parameters with disease severity and motor symptoms.

Forty-two patients with PD were recruited from the Department of Neurology, Henan University People′s Hospital from June 2018 to October 2019. Meanwhile, 40 healthy controls who visited the hospital for physical examination at the same period were enrolled. Corneal nerve fibers in both eyes of all participants were detected by using CCM. The differences of corneal nerve fibers were comparatively analyzed between PD group and healthy controls. Associations of corneal nerve parameters with clinical characteristics such as course of disease, Hoehn and Yahr stage (H-Y stage), unified Parkinson disease rating scale (UPDRS), levodopa equivalent daily dosage (LEDD) were analyzed by using partial correlations. The receiver operating characteristic (ROC) curve was used to analyze the capability of corneal nerve fibers for distinguishing patients with PD from healthy controls.

Corneal nerve fiber density (CNFD) in PD group ((19±3)/mm²) was significantly decreased compared with healthy controls ((28±4)/mm²) (t=10.798, P<0.001). However, corneal nerve branch density (CNBD) was significantly increased in PD group ((25±11)/mm²) compared with healthy controls ((18±6)/mm²) (t=-3.427, P=0.001). Meanwhile, corneal nerve fiber length (CNFL) was decreased in PD group ((11.0±2.5) mm/mm²) in comparison with healthy controls ((12.5±1.6) mm/mm²) (t=3.139, P=0.002). ROC curve analysis revealed that CNFD could discriminate PD patients from healthy controls, with an area under the curve of 0.961 3 (95%CI: 92.42-99.84, P<0.000 1). CNFD was negatively correlated with H-Y stage and UPDRS-Ⅲ (r=-0.501 and -0.399, both P<0.05). CNBD was significantly negatively associated with H-Y stage, UPDRS-Ⅲ and UPDRS-Total (r=-0.622, -0.394 and -0.354, respectively, all P<0.05). CNFL was negatively correlated with H-Y stage, UPDRS-Ⅲ and UPDRS-total (r=-0.574, -0.484 and -0.422, respectively, all P<0.05).

Small nerve fiber injuries exist in PD patients. Corneal nerve fibers negatively correlates with motor symptoms. CNFD have a good discriminative power to distinguish PD patients from healthy controls and may serve as a marker for PD.

To investigate the role of Orai1-mediated store-operated calcium entry in the immune damage of CD4⁺ T cells in septic mice.

Sepsis mouse model was established by cecal ligation and puncture(CLP). Balb/c mice of clean grade were sacrificed 1, 3, and 5 days after operation. Spleen samples were harvested at given intervals. Splenic CD4⁺ T cells were selected by immunomagnetic beads and the expression of Orai1 protein was detected by western blotting, the storage operated calcium entry (SOCE) was detected by flow cytometry, the apoptosis of CD4⁺ T cells was detected by flow cytometry, the proliferation of CD4⁺ T cells was detected by CCK-8, and the IFN-γ and IL-4 were detected by enzyme-linked immunosorbent assay (ELISA). Then the expression of Orai1 protein was regulated to further detect the SOCE and immune function of splenic CD4⁺ T cells in mice. The experiment was divided into 4 groups, sham group, CLP3 group, Orai1 down group (Orai1-down group) and Orai1 up regulation group (Orai1-up group).

The relative expression of Orai1 protein in splenic CD4⁺ T cells in sham group was 1.03±0.16. Compared with sham group, Orai1 protein levels in CLP Group were all significantly lower (F=19.64, P=0.000 5). The increased value of splenic CD4⁺ T cells fluorescence intensity in sham group was 494±41. Compared with sham group, the levels of SOCE in CLP Group were all lower (F=30.01, P=0.001). The ratio of early and late apoptosis of CD4⁺ T cells in sham group was 8.7%±1.5%. Compared with sham group, the early and late apoptosis rates of CLP Group were significantly higher (F=32.29, P=0.000 1). The OD of sham group was 0.81±0.10 at 450 nm. Compared with sham group, the proliferation ability of splenic CD4⁺ T cells in CLP Group were significantly decreased (F=7.26, P=0.001 8). Compared with sham group, the secretion of IFN-γ and IL-4 by CD4⁺ T cells and the ratio of IFN-γ/IL-4 in CLP Group were all significantly decreased (F=19.690, 6.183, 11.230, all P<0.05). Compared with CLP3 group, the increased value of fluorescence intensity of CD4⁺ T cells was significantly decreased, the early and late apoptosis ratio of CD4⁺ T cells was significantly increased, the OD450 nm value of CD4⁺ T cells was decreased, the multiplication capacity of splenic CD4⁺ T cells were decreased, the level of IFN-γ and IL-4 secreted by T cells were decreased, and the value of IFN-γ/IL-4 in orai1-down group was decreased (t=4.819, 7.952, 2.988, 28.760, 3.140, 7.670, all P<0.05). However, Orail-up group showed the opposite trend.

Orai1-mediated store-operated calcium entry can alleviate the immune dysfunction of CD4⁺ T cells in septic mice.

This paper reports a patient with Miller-Fisher syndrome (MFS) secondary to Brucella meningoencephalitis. The patient was exposed to cattle and sheep for a long time. The first manifestation was muscle soreness, headache, vomiting and confusion of limbs. Combined with clinical and imaging data, the patient was diagnosed with Brucella meningoencephalitis. During the treatment of anti-brucellosis, the patient suffered from sudden limb weakness, facial paralysis and fixed pupil; Secondary MFS was finally diagnosed by neuroelectrophysiology, neuroultrasound and CSF-related antibody detection. This case emphasizes the strong clinical heterogeneity of brucellosis, which can cause simultaneous damage to the peripheral and central nervous systems, and also provides ideas for the etiology of MFS and reduces the rate of misdiagnosis and misdiagnosis in clinical work.

Gastric glomus tumor (GGT) is a rare benign tumor that occurs in the gastric antrum. Glomus tumors of the gastric body with undetermined malignant potential are even rarer. Combined with previous literature, although GGT has relative characteristics and needs to be differentiated from gastric tumors, the diagnosis still needs pathological examination. This paper reports a case of glomus tumor with undetermined malignant potential in the body of the stomach, in order to improve the diagnosis and understanding of this disease, and to reduce misdiagnosis and misdiagnosis.

Herpes zoster is a skin disease caused by varicella-zoster virus (VZV), which is often accompanied by severe pain. Some patients, especially the elderly, are prone to postherpetic neuralgia (PHN), which seriously affects the quality of life of patients. Integrated Chinese and Western medicine has made some progress in the treatment of herpes zoster and the prevention of PHN; The development and marketing of recombinant herpes zoster vaccine provides the possibility for high-risk groups to prevent herpes zoster, and has broad development prospects. This article reviews the antiviral, analgesic and controversial use of glucocorticoids in western medicine, traditional Chinese medicine, and the application of vaccines, so as to provide reference for clinical prevention and treatment.

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease with short survival time after clinical symptoms, and there is no effective treatment at present. The study found that ALS is a gradually developing multi-stage disease. Animal models and clinical studies have revealed the pathophysiological characteristics and possible mechanisms of the pre-symptomatic stage of the disease, which provides inspiration for the formulation of more effective treatment plans.

Ectopic pregnancy is one of the main causes of death in early pregnancy women, and the incidence population is a younger trend. In the past, surgical treatment was mainly performed for patients in clinic. However, with the increase of the demand for reproductive function preservation, the clinical treatment of ectopic pregnancy gradually changed to conservative treatment. Expected treatment is a commonly used conservative treatment, which is a treatment that does not take treatment measures such as hormonal drugs or surgery within a certain period of time, but only gives clinical follow-up. This article reviews the indications, contraindications and efficacy of expectant treatment of ectopic pregnancy, and provides reference for the treatment of ectopic pregnancy.

Pancreatic segmentation is widely used in surgical planning, diabetes assessment and detection and analysis of pancreatic tumors. Because the shape, size and density of the pancreas vary greatly, and it is closely related to adjacent organs such as intestine and blood vessels, it is still a difficult problem to accurately segment the pancreas on CT images. In order to understand the factors affecting the deep learning pancreatic segmentation algorithm, this study evaluated the segmentation effect of 82 venous abdominal CT images from public datasets, analyzed from three aspects: population and clinical data, CT technical factors and CT imaging performance, and used Dice similarity coefficient (DSC) to evaluate the segmentation accuracy. The results showed that the mean DSC of pancreatic segmentation was 78% ± 8%. Factors significantly associated with pancreatic segmentation effects included patient body mass index (r=0.34,P<0.01), abdominal visceral fat (r=0.51,P<0.000 1), pancreatic volume (r=0.41,P=0.001), standard deviation of intrapancreatic CT values (r=0.30,P=0.01), the median and mean of CT values at the critical 5 mm of the pancreas (r=-0.53,P<0.000 1;r=-0.52,P<0.000 1), CT Image pixel values (r=0.31,P=0.004)。 There was no significant correlation between DSC and height, sex and mean CT values of pancreas.