Improved detection of aortic dissection in non-contrast-enhanced chest CT using an attention-based deep learning model.

Rationale and objectives: This study investigated the effects of implementing an attention-based deep learning model for the detection of aortic dissection (AD) using non-contrast-enhanced chest computed tomography (CT). Materials and methods: We analysed the records of 1300 patients who underwent contrast-enhanced chest CT at 2 medical centres between January 2015 and February 2023. We considered an internal cohort of 200 patients with AD and 200 patients without AD and an external test cohort of 40 patients with AD and 40 patients without AD. The internal cohort was divided into training and test sets, and a deep learning model was trained using 9600 CT images. A convolutional block attention module (CBAM) and a traditional deep learning architecture (namely, You Only Look Once version 5 [YOLOv5]) were combined into an attention-based model (i.e., YOLOv5-CBAM). Its performance was measured against the unmodified YOLOv5 model, and the accuracy, sensitivity, and specificity of the algorithm were evaluated by two independent radiologists. Results: The CBAM-based model outperformed the traditional deep learning model. In the external testing set, YOLOv5-CBAM achieved an area under the curve (AUC) of 0.938, accuracy of 91.5 %, sensitivity of 90.0 %, and specificity of 92.9 %, whereas the unmodified model achieved an AUC of 0.844, accuracy of 83.6 %, sensitivity of 71.2 %, and specificity of 96.0 %. The sensitivity results of the unmodified algorithms were not significantly different from those of the radiologists; however, the proposed YOLOv5-CBAM algorithm outperformed the unmodified algorithms in terms of detection. Conclusions: Incorporating the CBAM attention mechanism into a deep learning model can significantly improve AD detection in non-contrast-enhanced chest CT. This approach may aid radiologists in the timely and accurate diagnosis of AD, which is important for improving patient outcomes.

Improving diagnostic performance of rib fractures for the night shift in radiology department using a computer-aided diagnosis system based on deep learning: A clinical retrospective study.

OBJECTIVE: To investigate the application value of a computer-aided diagnosis (CAD) system based on deep learning (DL) of rib fractures for night shifts in radiology department. METHODS: Chest computed tomography (CT) images and structured reports were retrospectively selected from the picture archiving and communication system (PACS) for 2,332 blunt chest trauma patients. In all CT imaging examinations, two on-duty radiologists (radiologists I and II) completed reports using three different reading patterns namely, P1 = independent reading during the day shift; P2 = independent reading during the night shift; and P3 = reading with the aid of a CAD system as the concurrent reader during the night shift. The locations and types of rib fractures were documented for each reading. In this study, the reference standard for rib fractures was established by an expert group. Sensitivity and false positives per scan (FPS) were counted and compared among P1, P2, and P3. RESULTS: The reference standard verified 6,443 rib fractures in the 2,332 patients. The sensitivity of both radiologists decreased significantly in P2 compared to that in P1 (both p <  0.017). The sensitivities of both radiologists showed no statistical difference between P3 and P1 (both p >  0.017). Radiologist I's FPS increased significantly in P2 compared to P1 (p <  0.017). The FPS of radiologist I showed no statistically significant difference between P3 and P1 (p >  0.017). The FPS of Radiologist II showed no statistical difference among all three reading patterns (p >  0.05). CONCLUSIONS: DL-based CAD systems can be integrated into the workflow of radiology departments during the night shift to improve the diagnostic performance of CT rib fractures.

Nomogram based on prognostic nutrition index and Chest CT imaging signs predicts lymph node metastasis in NSCLC patients.

OBJECTIVE: To establish and validate a model capable of predicting lymph node metastasis (LNM) of non-small cell lung cancer (NSCLC) patients. METHODS: Preoperative clinical and CT imaging data on patients with NSCLC undergoing surgery were retrospectively analyzed. A model was developed using a training cohort of 290 patients. The univariate analysis followed by dichotomous logistic regression was performed to estimate different risk factors of lymph node metastasis, and a nomogram was constructed. Using another testing cohort of 120 patients, the performance of the nomogram was validated using several evaluation methods and indices and evaluated including via the area under the curve (AUC), calibration curve, Hosmer-Lemeshow test and decision curve analysis (DCA). RESULTS: CT-based imaging signs were important independent risk factors for lymph node metastasis in NSCLC patients. The possible risk factors also included four other independent risk factors through dichotomous logistic regression, i.e., age, SIRI, PNI and CEA, which were filtered and included in the nomogram. Nomogram yields AUC values of 0.828 [95% confidence interval (CI): 0.778-0.877] in the training cohort and 0.816 (95% CI: 0.737-0.895) in the validation cohort, respectively. The calibration curves showed high agreement in both the training and validation cohorts. At the threshold probability of 0-0.8, the nomogram increases the net outcomes compared to the treat-none and treat-all lines in the decision curve. CONCLUSIONS: The nomogram based on the PNI and CT images signs holds promise as a novel and accurate tool for predicting the LNM in NSCLC patients and guiding intraoperative lymph node dissection.

CT-based radiomics for differentiating intracranial contrast extravasation from intraparenchymal haemorrhage after mechanical thrombectomy.

OBJECTIVE: To develop a nonenhanced CT-based radiomic signature for the differentiation of iodinated contrast extravasation from intraparenchymal haemorrhage (IPH) following mechanical thrombectomy. METHODS: Patients diagnosed with acute ischaemic stroke who underwent mechanical thrombectomy in 4 institutions from December 2017 to June 2020 were included in this retrospective study. The study population was divided into a training cohort and a validation cohort. The nonenhanced CT images taken after mechanical thrombectomy were used to extract radiomic features. The maximum relevance minimum redundancy (mRMR) algorithm was used to eliminate confounding variables. Afterwards, least absolute shrinkage and selection operator (LASSO) logistic regression was used to generate the radiomic signature. The diagnostic performance of the radiomic signature was evaluated by the area under the curve (AUC), accuracy, specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV). RESULTS: A total of 166 intraparenchymal areas of hyperattenuation from 101 patients were used. The areas of hyperattenuation were randomly allocated to the training and validation cohorts at a ratio of 7:3. The AUC of the radiomic signature was 0.848 (95% confidence interval (CI) 0.780-0.917) in the training cohort and 0.826 (95% CI 0.705-0.948) in the validation cohort. The accuracy of the radiomic signature was 77.6%, with a sensitivity of 76.7%, a specificity of 78.9%, a PPV of 85.2%, and a NPV of 68.2% in the validation cohort. CONCLUSIONS: The radiomic signature constructed based on initial post-operative nonenhanced CT after mechanical thrombectomy can effectively differentiate IPH from iodinated contrast extravasation. KEY POINTS: • Radiomic features were extracted from intraparenchymal areas of hyperattenuation on initial post-operative CT scans after mechanical thrombectomy. • The nonenhanced CT-based radiomic signature can differentiate IPH from iodinated contrast extravasation early. • The radiomic signature may help prevent unnecessary rescanning after mechanical thrombectomy, especially in cases where contrast extravasation is highly suggestive.

An Illustrated Guide to the Imaging Evolution of COVID in Non-Epidemic Areas of Southeast China.

Purpose: By analyzing the CT manifestations and evolution of COVID in non-epidemic areas of southeast China, analyzing the developmental abnormalities and accompanying signs in the early and late stages of the disease, providing imaging evidence for clinical diagnosis and identification, and assisting in judging disease progression and monitoring prognosis. Methods: This retrospective and multicenter study included 1,648 chest CT examinations from 693 patients with laboratory-confirmed COVID-19 infection from 16 hospitals of southeast China between January 19 and March 27, 2020. Six trained radiologists analyzed and recorded the distribution and location of the lesions in the CT images of these patients. The accompanying signs include crazy-paving sign, bronchial wall thickening, microvascular thickening, bronchogram sign, fibrous lesions, halo and reverse-halo signs, nodules, atelectasis, and pleural effusion, and at the same time, they analyze the evolution of the abovementioned manifestations over time. Result: There were 1,500 positive findings in 1,648 CT examinations of 693 patients; the average age of the patients was 46 years, including 13 children; the proportion of women was 49%. Early CT manifestations are single or multiple nodular, patchy, or flaky ground-glass-like density shadows. The frequency of occurrence of ground-glass shadows (47.27%), fibrous lesions (42.60%), and microvascular thickening (40.60%) was significantly higher than that of other signs. Ground-glass shadows increase and expand 3-7 days after the onset of symptoms. The distribution and location of lesions were not significantly related to the appearance time. Ground-glass shadow is the most common lesion, with an average absorption time of 6.2 days, followed by consolidation, with an absorption time of about 6.3 days. It takes about 8 days for pure ground-glass lesions to absorb. Consolidation change into ground glass or pure ground glass takes 10-14 days. For ground-glass opacity to evolve into pure ground-glass lesions, it takes an average of 17 days. For ground-glass lesions to evolve into consolidation, it takes 7 days, pure ground-glass lesions need 8 days to evolve into ground-glass lesions. The average time for CT signs to improve is 10-15 days, and the first to improve is the crazy-paving sign and nodules; while the progression of the disease is 6-12 days, the earliest signs of progression are air bronchogram signs, bronchial wall thickening, and bronchiectasis. There is no severe patient in this study. Conclusion: This study depicts the CT manifestation and evolution of COVID in non-epidemic origin areas, and provides valuable first-hand information for clinical diagnosis and judgment of patient's disease evolution and prediction.

Non-contrast CT-based radiomic signature for screening thoracic aortic dissections: a multicenter study.

OBJECTIVE: To develop a non-contrast CT-based radiomic signature to effectively screen for thoracic aortic dissections (ADs). METHODS: We retrospectively enrolled 378 patients who underwent non-contrast chest CT scans along with CT angiography or MRI from 4 medical centers. The training and validation sets were from 3 centers, while the external test set was from a 4th center. Radiomic features were extracted from non-contrast CT images. The radiomic signature was created on the basis of selected features by a logistic regression algorithm. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve, accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were conducted to assess the predictive ability of radiomic signature. RESULTS: The radiomic signature demonstrated AUCs of 0.91 (95% confidence interval [CI], 0.86-0.95) in the training set, 0.92 (95% CI, 0.86-0.98) in the validation set, and 0.90 (95% CI, 0.82-0.98) in the external test set. The predicted diagnosis was in good agreement with the probability of thoracic AD. In the external test group, the diagnostic accuracy, sensitivity, specificity, PPV, and NPV were 90.5%, 85.7%, 91.7%, 70.6%, and 96.5%, respectively. CONCLUSIONS: A radiomic signature based on non-contrast CT images can effectively predict thoracic ADs. This method may serve as a potential screening tool for thoracic ADs. KEY POINTS: • The non-contrast CT-based radiomic signature can effectively predict the thoracic aortic dissections. • This radiomic signature shows better predictive performance compared to the current clinical model. • This prediction method may be a potential tool for screening thoracic aortic dissections.

Diffusion tensor imaging in evaluating testicular injury after unilateral testicular torsion and detorsion in rat model: A preliminary study.

Diffusion tensor imaging (DTI) is a functional magnetic resonance sequence based on the movement of water molecules. This study attempted to investigate the feasibility of DTI in evaluating testicular injury after testicular torsion and detorsion. Seventy-two rats were randomly divided into the sham group, torsion group and detorsion group. The left testis in the sham group was brought out through a scrotal incision for 1 hr, and that of the torsion group was twisted 720o clockwise for 1 hr and fixed to the scrotum, while the detorsion group was restored after being twisted 720° for 1 hr. Rats were further divided into four subgroups according to the set time, then performed DTI and histology analysis. The mean diffusion of the torsion and detorsion groups increased within 24 hr (p <.01), while it in the detorsion-1-week-group was lower than that in the detorsion-24-hr-group (p <.05). The fraction anisotropy of both experimental groups decreased in the acute phase (p <.01), while that of the detorsion-1-week-group increased (p <.01). Cosentino score in both experimental groups showed an increasing trend (p <.05). Besides, the spermatogenic ability of the detorsion-1-week-group decreased (p <.05). In conclusion, DTI was able to evaluate the injury after testicular torsion and detorsion.

Temporal changes of COVID-19 pneumonia by mass evaluation using CT: a retrospective multi-center study.

BACKGROUND: Coronavirus disease 2019 (COVID-19) has widely spread worldwide and caused a pandemic. Chest CT has been found to play an important role in the diagnosis and management of COVID-19. However, quantitatively assessing temporal changes of COVID-19 pneumonia over time using CT has still not been fully elucidated. The purpose of this study was to perform a longitudinal study to quantitatively assess temporal changes of COVID-19 pneumonia. METHODS: This retrospective and multi-center study included patients with laboratory-confirmed COVID-19 infection from 16 hospitals between January 19 and March 27, 2020. Mass was used as an approach to quantitatively measure dynamic changes of pulmonary involvement in patients with COVID-19. Artificial intelligence (AI) was employed as image segmentation and analysis tool for calculating the mass of pulmonary involvement. RESULTS: A total of 581 confirmed patients with 1,309 chest CT examinations were included in this study. The median age was 46 years (IQR, 35-55; range, 4-87 years), and 311 (53.5%) patients were male. The mass of pulmonary involvement peaked on day 10 after the onset of initial symptoms. Furthermore, the mass of pulmonary involvement of older patients (>45 years) was significantly severer (P<0.001) and peaked later (day 11 vs. day 8) than that of younger patients (≤45 years). In addition, there were no significant differences in the peak time (day 10 vs. day 10) and median mass (P=0.679) of pulmonary involvement between male and female. CONCLUSIONS: Pulmonary involvement peaked on day 10 after the onset of initial symptoms in patients with COVID-19. Further, pulmonary involvement of older patients was severer and peaked later than that of younger patients. These findings suggest that AI-based quantitative mass evaluation of COVID-19 pneumonia hold great potential for monitoring the disease progression.

[Applied research of CT angiography of lower extremities by using 100 kVp and iodixanol(270 mg I/ml) combined with iDose(4) iterative reconstruction technique].

OBJECTIVE: To evaluate the feasibility of CT angiography of lower extremities by using 100 kVp as tube voltage and Iodixanol(270 mg I/ml) as contrast medium combined with iDose(4) iterative reconstruction technique. METHODS: A total of 52 continuous patients with clinically suspected lesions of lower extremity arteries underwent CT angiography of lower extremities, divided into 2 groups, and different scan protocols were adopted. "double low" group included 26 patients using 100 kVp, Iodixanol (270 mg I/ml) and iDose(4) -4 iterative reconstruction algorithm; routine group included 26 patients using 120 kVp, Iopromide(370 mg I/ml) and filtered back projection reconstruction algorithm. The total amount of contrast medium in both groups was 95 ml. Artery CT value and background noise at the level of L4 vertebral, hip, knee and ankle were measured, signal to noise ratio (SNR) and contrast to noise ratio (CNR) were calculated, and the quality of images was evaluated subjectively. Scan length (L), volume CT dose index (CTDI(VOL)) and dose length product(DLP) were recorded, and the effective dose (ED) was calculated. The measurement results and subjective evaluation were analyzed statistically. CTA results were analyzed with the digital subtraction angiography (DSA) as the "gold standard". RESULTS: (1) No significant difference was existed in gender, age, height, weight and body mass index (BMI) of these two groups (P > 0.05). (2) No significant difference was existed in artery CT value, SNR and CNR at the level of L4 vertebral, hip, knee and ankle of the two groups (P > 0.05). Compared with routine group, background noise of "double low" group at the level of hip and knee increased by 16.6% and 13.8%, respectively (P < 0.05). (3) The image quality of the two groups met the requirement of diagnosis, no significant statistical difference was existed in subjective evaluation (P > 0.05). (4) The CTDI(VOL), DLP, ED of "double low" group were lower than that of routine group, with significant statistical difference (P < 0.05). The total amount of iodine in "double group" was lower than that of routine group (25.6 g vs 35.2 g). (5) A total of 7 cases from "double low" group underwent DSA examination, and 18 pathological changes (stenosis and occlusion) were found totally, in which 16 matched CTA; 6 cases from routine group underwent DSA examination, and 15 pathological changes (stenosis and occlusion) were found totally, in which 13 matched. CTA. No significant statistical difference was existed between the two groups in diagnostic efficacy (P = 0.626). CONCLUSIONS: Using 100 kVp and Iodixanol (270 mg I/ml) combined with iDose(4) -4 iterative reconstruction technique for CT angiography of lower extremities, the image quality could meet the requirement for clinical diagnosis, the radiation dose and the volume of contrast medium could be lowered.

[Reliability research of multi-planar reformation of multi-spiral computerized tomography in measuring kyphosis angle after thoracolumbar fracture].

OBJECTIVE: To study the feasibility and reliability of the multi-planar reformation (MPR) of multi-spiral CT (MSCT) in measuring the kyphosis angle (KA) after thoracolumbar fracture. Methods From December 2007 to December 2009, 45 thoracolumbar fracture patients who underwent computed radiology (CR) and MSCT were recruited. There were 32 males and 13 females with a mean age of 48 years (range, 24-63 years), including 36 simple compression fractures and 9 burst fractures. The fracture locations were T11 in 6 cases, T12 in 11 cases, L1 in 20 cases, and L2 in 8 cases. Fracture was caused by traffic accident in 25 cases, by falling from height in 12 cases, and by others in 8 cases. The imaging examination was performed after 2 hours to 7 days of injury in 22 cases and after more than 7 days in 23 cases. The KA was measured on the lateral X-ray films of CR and MPR by two observers, then the measurements were done again after three weeks. The data were statistically analyzed. RESULTS: The average KA values on CR by two observers were (20.75 +/- 8.31)degrees and (22.49 +/- 9.07)degrees, respectively; showing significant difference (P < 0.05), and the correlation was good (r = 0.882, P < 0.05). The average KA values on MPR by two observers were (16.65 +/- 8.62)degrees and (17.08 +/- 7.88)degrees, respectively, showing no significant difference (P > 0.05), the correlation was excellent (r = 0.976, P < 0.05). The average KA values on CR and MPR were (21.61 +/- 8.43)degrees and (16.87 +/- 8.20)degrees, respectively; showing significant difference (P < 0.05), the correlation was good (r = 0.852, P < 0.05). CONCLUSION: It is more feasible and reliable in measuring the KA on MRP of MSCT than CR, but the value is larger on CR.

[Clinical analysis of 234 cases with congenital malformations of respiratory system].

OBJECTIVE: To explore clinical characteristics, radiographic findings and diagnostic methods of patients with congenital malformations of respiratory system for enhancing the diagnosis of congenital malformations of respiratory system in children. METHOD: Totally 234 patients with congenital malformations of respiratory system were chosen from the inpatient department of Yuying Children's Hospital Affiliated to Wenzhou Medical College from July 2003 to June 2008. The clinical presentations and radiographic findings of these children were analyzed. RESULT: Of the 234 patients with congenital malformations of respiratory system, the age at diagnosis was between the first day and 14 years of age, mean age was 1.12 years. The main symptoms were persistent laryngeal stridor, recurrent wheezing, recurrent respiratory tract infections and dyspnea. Through the use of chest X-ray, spiral CT 3D reconstructions, fiberoptic bronchoscopy and other laboratory techniques, 213 cases were diagnosed as having single malformation and 21 cases were found to have multiple malformations. Of the 213 cases with single malformation, 97 cases had laryngeal malformation (congenital laryngeal stridor in 90 cases, congenital laryngeal webs in 5 cases and congenital laryngeal cyst in 2 cases), 35 cases had tracheal-bronchial malformation (congenital tracheobronchial stenosis in 17 cases, congenital abnormal bronchial origin in 7 cases, tracheobronchomalacia in 10 cases and tracheoesophageal fistula in 1 case), 43 cases had lung malformation (pulmonary sequestration in 5 cases, congenital lung cysts in 22 cases, congenital lobar emphysema in 1 case, agenesis of lung and hypoplasia of lung in 8 cases and congenital cystic adenomatoid malformation in 7 cases), 38 cases had diaphragm malformation, 28 cases had congenital tracheal-bronchial stenosis as confirmed by spiral CT 3D reconstructions and fiberoptic bronchoscopy. Ten cases with congenital abnormal bronchial origin were diagnosed with spiral CT 3D reconstructions. Laryngeal stridor and tracheobronchomalacia were diagnosed by fiberoptic laryngoscope and fiberoptic bronchoscopy. The accuracy rates of preoperative diagnosis through clinical and radiographic examinations of 37 cases with lung malformation and 36 cases with diaphragm malformation were 83.78% and 91.67%. CONCLUSION: Congenital malformations of respiratory system are a group of diseases that are important for pediatric respiratory clinicians. Congenital malformations of respiratory system should be considered in children with persistent laryngeal stridor, recurrent wheezing, recurrent respiratory tract infections and dyspnea. The radiographic examination and respiratory endoscope play important roles in the diagnosis of congenital malformations of respiratory system.

Localized Castleman disease in retroperitoneum: newly discovered features by multi-detector helical CT.

OBJECTIVE: We describe CT features of our three cases with localized Castleman disease in the retroperitoneum and review literature. Besides those CT features, which have been reported before, we mainly present some newly discovered CT findings of the disease. These new CT findings include the sign of peripheral 'rim-like' enhancement at the early phase of an enhanced CT scan, a higher ratio of the left sided retroperitoneal location to the right side and the presence of local peritoneal thickening around the lesion. In addition, the feeding artery of the lesion is more visually pronounced than ever before by a 16-detector CT scanner. CONCLUSION: After reviewing the literature and comparing with their histological findings, we suggest these newly discovered findings are relatively characteristic CT features of the disease. Moreover, multi-detector helical CT can now show more details of the disease than ever before.

Eosinophilic enteritis: CT features.

We report CT features of four cases of Eosinophilic enteritis. The disease involves the jejunum in one case, the ileum in two cases, and the colon in the remaining case. Two cases demonstrate a predominantly mucosal type of eosinophilic enteritis, while the other two cases demonstrate a predominantly subserosal type. CT findings include bowel wall thickening in four cases, bowel fold thickening in two cases, layering of the bowel wall in two cases, luminal narrowing without obstruction in three cases, an intra-luminal granuloma mimicking a huge polyp in one case, an extra-luminal irregular granuloma markedly enhancing and slightly necrosing in one case, mesenteric lymphadenopathy with peripheral rim-like enhancement and marked necrosis in one case and ascites in one case. CT findings are more characteristic of an inflammatory disease rather than of a tumor, and these findings are helpful for assessing the extent and location of the disease. Moreover, combined with its typical clinical manifestations, CT findings may lead to the correct diagnosis.

An emission inventory of marine vessels in Shanghai in 2003.

We developed an air pollutant emission inventory for marine vessels in the Shanghai Port in 2003. We estimated emissions under cruising and maneuvering conditions based on two categories of vessels: (1) vessels in the Outer Port, which enter the area following notification of the Shanghai Maritime Safety Administration, a division of the Ministry of Communications of P.R. China; and (2) vessels in internal waterways, which enter those waters following notification of the local Port Administration Centers. Vessels in the Outer Port consist of three subcategories: (1) international vessels that are engaged in foreign commerce; (2) domestic vessels traveling along the downstream portion of the Huangpu River; and (3) domestic vessels traveling along the coast. We also estimate emissions from vessels over 1000 DWT operating under hotelling conditions in the Outer Port. In 2003, the total number of calls was approximately 1.3 million, of which 57% is attributed to vessels in internal waterways and 43% to vessels in the Outer Port. Total marine emissions for NOx, SO2, PM, HC, and CO2 in 2003 are estimated to be 58,160, 51,180, 6960, 4560, and 3,012,800 tons, respectively. Emissions are allocated to 1 km x 1 km grid cells for the 129 km x 102 km Shanghai Port study domain.