[inventory] summary of scientific research progress of radiology in 2020 (8)

< br / > < br / >? The overall survival time (OS) of acute myeloid leukemia (AML) is relatively shortThe diffusion weighted MRI of IVIM can separate the microvascular perfusion and tissue diffusion, which may be used to evaluate the infiltration of bone marrow in addition to solid massTherefore, the study of overall survival will be helpful to elucidate the value of IVIM in evaluating the long-term prognosis of AML< br / > < br / > objective < br / > < br / > the purpose of this study is to evaluate whether IVIM indexes using bone marrow infiltration are correlated with OS in newly diagnosed AML patients< br / > < br / > materials and methods < br / >MRI of lumbar spine was performed by IVIM sequenceThe median follow-up was 20 months (range 1-56 months)IVIM parameters (false perfusion fraction: F; diffusion coefficient: D; false diffusion coefficient: D *) were collectedThe IVIM threshold of OS was evaluated by nonparametric log level testSingle variable Kaplan Meier test and multivariate Cox proportional hazard regression analysis were used to evaluate the predictive value of potential indicatorsResults 53 patients were evaluatedThe nonparametric log rank test showed that the thresholds of F and D were 31.0% and 0.2 × 10-3 mm2 / sec, respectivelyUnivariate analysis showed that high F value (> 31.0%) and low D value (≤ 0.2 × 10-3 mm2 / sec) were associated with shorter OS (P =003 and01)More than 31.0% of F value was correlated with OS and independent of clinical covariates (age, karyotype and leukocyte count)< br / > < br / > conclusion < br / > https：// article.do ? Id = 67d31934e4bd < br / > Objective to compare the value of DW MRI and MR elastography in the evaluation of liver fibrosis < br / > < br / > materials and methods < br / > The shear modulus (μ MRE) was measured in the region of interest of liver cirrhosis in MR elastography The ADC was measured on DW MRI (b = 200 and 1500 sec / mm2) and converted to the virtual shear modulus (μ diff) based on DW MRI The liver fibrosis was graded by μ MRE and μ diff value (f0-f4) based on MRI, and the indexes of serum fibrosis were analyzed Results: 74 patients were included in this study The inter group reliability coefficients of DW MRI and MR elastography were 0.86 kPa and 1.2 kPa, respectively There was a significant correlation between displacement ADC and μ MRE (R2 = 0.81; P < 0.001), which indicated that there was a high consistency between μ MRE and μ diff (mean difference, - 0.02 kPa ± 0.88; P < 0.001) A total of 55% of the patients (41 / 74) based on DW MRI and MR elastography had the same fibrosis classification, and about 35% of the patients (26 / 74) had a grade 1 difference When the fibrosis grade was divided into two grades (f0-f1), 85% of the patients had the same fibrosis grade Compared with the serum index, μ diff has a better ability to distinguish f0-f2 and f3-f4 (AUC, 0.79; 95% CI: 0.69, 0.90), while the serum index is slightly better than the former in distinguishing f0-f1 and f2-f4 (liver fibrosis grade is evaluated by MR elastography) Vs f3-f4, 0.81 [95% CI： 0.71, 0.91], P = .08)。 < br / > < br / > conclusion < br / > DW MRI can be used as an alternative method for noninvasive liver fibrosis grading in MR elastography https：// article.do ? Id = 8ed61934e68e < br / > Such a model needs to be validated in clinical practice < br / > < br / > objective < br / > < br / > the purpose of this study is to establish a multi task deep learning model to grade hip plain film according to the signs of hip osteoarthritis, and compare with the results of diagnosis < br / > by radiologist < br / > < br / > materials and methods < br / > < br / > in this study, the anterior and posterior radiographs of pelvis in load-bearing position were analyzed retrospectively to evaluate the hip joint The patients were followed up 48 months later According to the diagnostic criteria of osteoarthritis, femoral osteophyte (FO), acetabulum osteophyte (AO) and hip joint space stenosis (JSN) were classified as none, mild, severe and severe Subchondral sclerosis and subchondral cyst were classified as none or none Subjects were divided into training group (80%) (n = 3494), verification group (10%) (n = 437) and test group (10%) (n = 437) The neural network is based on densenet-161, which is built by using the shared convolution feature extractor trained by multi task loss function Statistical analysis was used to evaluate the effectiveness of the model Results 4368 subjects were included in the study In the internal data, the accuracy of Fos, AOS, JSN, subchondral sclerosis and subchondral cystic features were 86.7% (1333 of 1538), 69.9% (1075 of 1538), 81.7% (1257 of 1538), 95.8% (1473 of 1538) and 97.6% (1501 of 1538) respectively In the external data, the accuracy of Fos, AOS, JSN, subchondral sclerosis and subchondral cystic features were 82.7% (86 of 104), 65.4% (68) respectively of 104) 、80.8% (84 of 104) 、88.5% (92 of 104) 、91.3% (95 of 104)。 < br / > < br / > conclusion < br / > < br / > multi task depth model can be used to evaluate the characteristics of osteoarthritis of hip joint on plain film https：// article.do ? Id = 7f8a19348426 < br / >? < br / > < br / > background < br / > < br / > intracranial atherosclerosis is an important cause of ischemic brain < br / > Stroke < br /, which is related to several risk factors of < br / > blood vessels < br / / At present, the evaluation of intracranial arterial lesions is mainly based on the stenosis of the lumen rather than the vessel wall < br / > < br / > objective < br / > < br / > < br / > materials and methods < br / > The number, location and enhancement of vascular wall lesions were evaluated in the anterior circulation artery Using the generalized evaluation equation to evaluate the correlation between the vascular risk index and the number and intensity of vascular wall lesions Results: 90 subjects were included in this study The older the age, the higher the score of hypertension, < br / > diabetes, < br / > and multi vessel risk index, the more the number of vascular wall lesions in anterior circulation There was a correlation between the enhancement of vascular wall lesions and age There was no significant correlation between smoking and the number of vascular wall lesions < br / > < br / > conclusion < br / > https：// article.do ? Id = 8a2c1938585d < br / >? Background: there is a significant correlation between pulmonary dysfunction and obesity, but there is no prospective CT study to evaluate the effect of obesity on lung and bronchus The purpose of this study was to evaluate the morphological changes of lung parenchyma and trachea in 6 months before and after obesity surgery and the correlation between the function and symptoms < br / > < br / > materials and methods < br / > Before and after the operation, all patients underwent restrictive inspiratory and expiratory chest CT scan, pulmonary function examination, and body mass index computer modified Medical Research Committee dyspnea scale and Epworth score * 2 chest radiologists rated the density of mosaics, end expiratory gas retention and tracheal morphology The cross axial area of trachea at the end of inspiratory and expiratory was measured T-test or Wilcoxon's sign grade test were used to compare the price before and after operation Using Spearman regression and logistic regression analysis to evaluate the correlation between CT imaging performance, function and symptom index The results were evaluated in 51 patients Before surgery, the degree of gas retention was significantly correlated with the decrease of total vital capacity (Spearman rank correlation coefficient [rs] = -0.40 * P =.004) After surgery, the density of mosaics (0% [interquartile range {IQR}:0%-2.5%] vs 0% [IQR:0%-0%], P <.001), gas retention (9.6% [IQR:5.8%-15.8%] vs 2.5% [IQR:0%-6.7%], P < VS) and tracheal collapse (201, 201, 229, 229, and *) decreased The postoperative mMRC dyspnea score was positively correlated with the degree of gas retention (rs = 0.46, P =.001) and the morphology of the end trachea (rs = 0.40 * P =.01) In multivariate analysis, gas retention was the major determinant of postoperative dyspnea improvement (odds ratio, 1.2; 95% confidence interval:1.1 * 1.2; P =.03) Conclusion * conclusion obese patients will have respiratory symptoms after bariatric surgery, which is associated with the decrease of CT and tracheal collapse degree https：// article.do ? Id = f11a93860ef < br / > < br / > radiology: machine learning helps you to find and calculate the acute ischemic cerebral infarction < br / > < br / > background < br / > < br / > it is very important for the treatment of patients with acute ischemic stroke (AIS) to be able to identify the acute infarct and its degree at the time of onset Patients with large infarction are less likely to benefit from thrombolysis or mechanical thrombectomy < br / > < br / > objective < br / > < br / > the purpose of this study is to use non enhanced CT to establish an automatic calculation method that can evaluate the acute infarct focus and volume of AIS patients < br / > < br / > materials and methods < br / > The volume of ischemic cerebral infarction was drawn as a reference standard An automatic segmentation method including machine learning (ML) was established to identify cerebral infarction 157 patients were randomly selected to train and verify the ML model with non enhanced images of infarction lesions on DW MRI, and the remaining 100 patients were tested as an independent group Quantitative comparison of ML calculation with Bland Altman plots and Pearson correlation analysis