Severity Assessment of COVID-19 based on Clinical and Imaging Data

ObjectivesThis study aims to develop a machine learning approach for automated severity assessment of COVID-19 patients based on clinical and imaging data. Materials and MethodsClinical data--demographics, signs, symptoms, comorbidities and blood test results--and chest CT scans of 346 patients from two hospitals in the Hubei province, China, were used to develop machine learning models for automated severity assessment of diagnosed COVID-19 cases. We compared the predictive power of clinical and imaging data by testing multiple machine learning models, and further explored the use of four oversampling methods to address the imbalance distribution issue. Features with the highest predictive power were identified using the SHAP framework. ResultsTargeting differentiation between mild and severe cases, logistic regression models achieved the best performance on clinical features (AUC:0.848, sensitivity:0.455, specificity:0.906), imaging features (AUC:0.926, sensitivity:0.818, specificity:0.901) and the combined features (AUC:0.950, sensitivity:0.764, specificity:0.919). The SMOTE oversampling method further improved the performance of the combined features to AUC of 0.960 (sensitivity:0.845, specificity:0.929). DiscussionImaging features had the strongest impact on the model output, while a combination of clinical and imaging features yielded the best performance overall. The identified predictive features were consistent with findings from previous studies. Oversampling yielded mixed results, although it achieved the best performance in our study. ConclusionsThis study indicates that clinical and imaging features can be used for automated severity assessment of COVID-19 patients and have the potential to assist with triaging COVID-19 patients and prioritizing care for patients at higher risk of severe cases.

Severity Assessment and Progression Prediction of COVID-19 Patients based on the LesionEncoder Framework and Chest CT

Automatic severity assessment and progression prediction can facilitate admission, triage, and referral of COVID-19 patients. This study aims to explore the potential use of lung lesion features in the management of COVID-19, based on the assumption that lesion features may carry important diagnostic and prognostic information for quantifying infection severity and forecasting disease progression. A novel LesionEncoder framework is proposed to detect lesions in chest CT scans and to encode lesion features for automatic severity assessment and progression prediction. The LesionEncoder framework consists of a U-Net module for detecting lesions and extracting features from individual CT slices, and a recurrent neural network (RNN) module for learning the relationship between feature vectors and collectively classifying the sequence of feature vectors. Chest CT scans of two cohorts of COVID-19 patients from two hospitals in China were used for training and testing the proposed framework. When applied to assessing severity, this framework outperformed baseline methods achieving a sensitivity of 0.818, specificity of 0.952, accuracy of 0.940, and AUC of 0.903. It also outperformed the other tested methods in disease progression prediction with a sensitivity of 0.667, specificity of 0.838, accuracy of 0.829, and AUC of 0.736. The LesionEncoder framework demonstrates a strong potential for clinical application in current COVID-19 management, particularly in automatic severity assessment of COVID-19 patients. This framework also has a potential for other lesion-focused medical image analyses.

Clinical and genetic analysis of eleven pediatric patients with Alagille syndrome / 中华儿科杂志

Objective@#To explore the clinical and molecular genetic features of patients with Alagille syndrome (AS).@*Methods@#The clinical data of eleven pediatric patients, who were suspected to have AS at the Department of Pediatrics in the First Affiliated Hospital of Jinan University from August 2010 to March 2017, were collected and analyzed. Genomic DNA was extracted from peripheral blood leukocytes of the patients and their parents. For 5 patients collected before March 2006, all JAG1 exons and their flanking sequences were directly sequenced. For the remaining 6 patients, high-throughput gene capture technology, chromosomal microarray analysis (CMA) and whole-genome copy-number variant(CNV) analysis were utilized, when necessary, to explore the genetic causes.@*Results@#All patients had cholestasis. However, the γ-glutamyl transpeptidase (GGT) levels in one patient were normal. Nine patients had posterior embryotoxon and facial malformations. Eight patients displayed heart defects. Seven patients presented with vertebral anomalies and among them, 1 patient had sacralization of the cubitus and radius. The condition of nine patients tended to be stabilized on follow-up, but 1 patient died of liver failure in late infancy and 1 got worse. Seven JAG1 variants were detected in 9 out of the 11 AS patients, with c.1977G>A (p.Trp659*) and c.1106_1107delCC (p.Pro369fs) being two novel variants. Two heterozygous interstitial deletions of 3.0 Mb and 9.24 Mb in size, respectively, in chromosome 20 were discovered in the remaining 2 patients. Both deletions involved the entire JAG1 gene. De novo origin was unveiled for the detected variants in 7 patients and interstitial deletions in two. Although the mother of 2 patients carried the relevant variant, she did not demonstrate any clinical features of AS.@*Conclusions@#With cholestasis, posterior embryotoxon, facial malformations, heart defects and vertebral anomalies being the major manifestations, AS demonstrated variable clinical expressivities and incomplete penetrance. This study identified a total of 7 JAG1 variants as well as 2 interstitial deletions involving this gene, and among them, the variants c.1977G>A (p.Trp659* ) and c.1106_1107delCC (p.Pro369fs) as well as the 9.24 Mb chromosomal interstitial deletion had not been reported previously.

MRI in assessment of relationship between acromial morphology and rotator cuff tear / 中国医学影像技术

Objective To analyze the relationship between the morphological characteristics and rotator cuff tear (RCT) by MR.Methods The data of clinic and shoulder MRI of 37 patients with RCT (patients group) were analyzed retrospectively,and 19 healthy volunteers were collected in control group.The acromial shapes were classified into type Ⅰ (flat),type Ⅱ (curved),type Ⅲ (hooked) and type Ⅳ (convex).Additional measurements about pathogenesis of RCT including acromio-humeral distance (AHD),acromial index (AI),lateral acromial angle (LAA) and acromial thickness were performed for further assessment.Results Type Ⅱ was the most commonly encountered acromial shape in patients group (16/37,43.24 ％) and control group (11/19,57.89 ％).There was no statistically significant difference in the incidence of each acromial shape between two groups (P＞0.05).However,the AHD,AI,LAA and acromial thickness showed statistically significant difference between the patients group and control group (all P＜0.05).The type Ⅲ acromion was signifi candy different from the other types in patients group (P＜0.05).Conclusion MRI can clearly display RCT.The AHD and LAA are smaller,the AI and acromial thickness are bigger in RCT patients.Type Ⅲ acromion may increase risks for RCT.

Analysis of clinical features and AGL gene mutations in a family with glycogen storage disease type IIIa / 中华医学遗传学杂志

<p><b>OBJECTIVE</b>To investigate the clinical features and AGL gene mutations in a family with glycogen storage disease type IIIa (GSD IIIa).</p><p><b>METHODS</b>Clinical data for diagnosis, treatment and follow-up of a sick child with GSD III was collected and analyzed. Genomic DNA was extracted from the peripheral blood samples from the patient and his parents. Polymerase chain reaction and direct DNA sequencing were utilized to analyze all of the exons of the AGL gene.</p><p><b>RESULTS</b>The genotype of the child was found to be c.3710_3711delTA/IVS14+1G>T. The former was a maternally-inherited mutation, which has not been reported previously. The latter was an abnormal splice-site mutation inherited from the father.</p><p><b>CONCLUSION</b>Based on its clinical and molecular evidences, the patient was diagnosed as GSD IIIa in conjunction with retrobular optic neuritis.</p>

Value of dual-energy computed tomography in the diagnosis of gouty arthritis / 南方医科大学学报

<p><b>OBJECTIVE</b>To investigate the value of dual-energy computed tomography (DECT) in the diagnosis of gouty arthritis.</p><p><b>METHODS</b>Sixty-one patients with gout, 30 with ankylosing spondylitis and 30 with rheumatoid arthritis were included in the study. DECT scans of the hands, wrists, elbows, feet, ankles, knees, lumbar, pelvis and sacroiliac joint were performed. For post-processing, a color-coding gout software protocol was used. The demographic data and blood uric acid levels were recorded. For 3 gout patients, the findings of puncture biopsy and DECT were compared. Ten gout patients with urate crystal deposition upon recruitment underwent DECT scans again after a 6-month urate-lowering therapy.</p><p><b>RESULTS</b>The positivity rates of DECT scan differed significantly among the patients with gout, ankylosing spondylitis and rheumatoid arthritis [98.4% (60/61), 13.3% (4/30), and 6.7% (2/30), respectively; χ² =95.522, P<0.05). Of the 21 patients with acute gouty arthritis, 20 (95.2%) showed positive DECT finding, and all the 40 patients with chronic gouty arthritis showed positive findings. In the patients with patients with gout, ankylosing spondylitis and rheumatoid arthritis, the positivity rates of hyperuricemia were 97.3% (36/37), 44.4% (4/9), and 28.6% (2/7), respectively (χ² =24.197, P<0.05). A total of 344 urate deposition sites were detected in the gout patients, involving most commonly the first metatarsophalangeal joint (22.1%), the middle and distal end of the first phalanges of the toes (19.8%), the calcaneus (17.4%), and the inferior extremity of the tibia (13.4%). Seventeen and 5 urate deposition sites were found in ankylosing spondylitis patients and rheumatoid arthritis patients, respecitvely. The 10 gout patients receiving a 6-month urate-lowering therapy showed decreased urate deposition on DECT scan.</p><p><b>CONCLUSIONS</b>DECT scan can detect urate deposition to allow differentiation diagnosis and follow-up in gout patients.</p>

Colorectal carcinoma: a preliminary study on magnetic resonance imaging / 中华放射学杂志

Objective To explore the feasibility and value of MRI examination in colorectal carcinoma. Methods Thirty-eight patients with clinically suspected colorectal carcinoma were included in this study and all of them underwent MRI examination. About 300- 1 000 ml water was administered transrectally before scanning. Axial plain scan, three-planar enhanced scan and hydrography were performed, respectively. Results In 38 patients, 31 cases were pathologically proved as colorectal carcinoma. The main MRI findings were: intraluminal soft tissue mass (n=5), irregular thickening of colorectal wall and partial or circumferential stenosis of colorectal lumen (n=26). The tumors were all demonstrated as medium signal intensity on T 1-weighted images. Thirty of 31 cases showed slightly hyperintensity and one was conspicuous high signal intensity on T 2-weighted images. Marked enhancement was demonstrated in all foci. Focal low signal structures were presented in the pericolorectal fat on precontrast T 1-weighted SE sequence in 5 cases. Smooth margins were revealed in 12 cases and irregular and/or nodular margins of peripheral wall in 19 cases. Colorectal hydrography showed that there were intraluminal irregular filling defects (n=14), and that "cuff sign" or "sawn-off sign" was demonstrated in the distal end of tumor and its proximal part couldn't be seen (n=10). Thirty of 31 patients with pathologically proved colorectal carcinoma and 4 of 7 patients with non-colorectal carcinoma were correctly identified by MRI. The sensitivity, specificity, and accuracy were 96.8%, 57.1%, and 89.5%, respectively. The accuracy of MRI in T staging was 83.9% (26/31). The accuracy in T1-2, T3, and T4 staging was 75.0% (9/12), 88.2% (15/17), and 100% (2/2), respectively. Conclusion MRI could clearly show the longitudinal and horizontal invasion of colorectal carcinoma as a whole. It could accurately determine the invasive depth of local lesions. Barium enema examination may be partially replaced by colorectal hydrography with MIP reconstruction. It is of directorial value in clinical treatment.