Contrast-enhanced CT parameters predict short-term tumor response in patients with hepatocellular carcinoma who received sequential combined anti-angiogenesis and immune checkpoint inhibitor treatment.

PURPOSE: To evaluate whether relative Hounsfield unit attenuation index (rHUAI) on contrast-enhanced computed tomography (CECT) can predict tumor response in advanced hepatocellular carcinoma (HCC) patients who received sequential combined treatment of immune checkpoint inhibitor (ICI) and anti-angiogenesis therapy. METHOD: One hundred seventeen advanced HCC patients who underwent the sequential combined treatment in a tertiary hospital between March 2020 and December 2021 were allocated to prediction and validation cohorts (with a ratio of 2:1) based on the time of initial ICI treatment. rHUAI from the arterial to the portal-venous phase (rHU_ap) and from the portal-venous to the delayed phase (rHU_pd) was calculated. The optimal cut-off values (COVs) of rHU_ap and rHU_pd for predicting tumor response were identified using Youden's index. Univariate and multivariable analyses were performed to assess the relationship between the COVs and tumor response. The validity of COVs was verified in the validation cohort using the chi-square test and Cramer's V coefficient (V). RESULTS: The optimal COVs of the two observers were 0.5316 and 0.3265 for rHU_ap, and -0.0208 and -0.0048 for rHU_pd, respectively. Multivariable analysis suggested that the COVs were independently associated with tumor response in the prediction cohort (rHU_ap, Odds ratio: 7.727 and 7.808, 95 % CI: 2.516-23.728 and 2.399-25.410, p value < 0.001 and 0.001; rHU_pd, Odds ratio: 0.034 and 0.011, 95 % CI: 0.002-0.600 and 0.001-0.209, p value of 0.021 and 0.003). In the validation cohort, the optimal COVs of rHU_ap had a moderate to a strong association with tumor response (V = 0.362-0.545, p < 0.05). The association between COVs of rHU_pd and tumor response was slight to strong (V = 0.24-0.545, p = 0.001 to 0.134). CONCLUSION: rHUAI obtained from CECT has the potential as a non-invasive tool for predicting tumor response in advanced HCC patients who have received combined ICI and anti-angiogenesis treatment.

Pretreatment Computed Tomography-Based Machine Learning Models to Predict Outcomes in Hepatocellular Carcinoma Patients who Received Combined Treatment of Trans-Arterial Chemoembolization and Tyrosine Kinase Inhibitor.

Aim: Trans-arterial chemoembolization (TACE) in combination with tyrosine kinase inhibitor (TKI) has been evidenced to improve outcomes in a portion of patients with hepatocellular carcinoma (HCC). Developing biomarkers to identify patients who might benefit from the combined treatment is needed. This study aims to investigate the efficacy of radiomics/deep learning features-based models in predicting short-term disease control and overall survival (OS) in HCC patients who received the combined treatment. Materials and Methods: A total of 103 HCC patients who received the combined treatment from Sep. 2015 to Dec. 2019 were enrolled in the study. We exacted radiomics features and deep learning features of six pre-trained convolutional neural networks (CNNs) from pretreatment computed tomography (CT) images. The robustness of features was evaluated, and those with excellent stability were used to construct predictive models by combining each of the seven feature exactors, 13 feature selection methods and 12 classifiers. The models were evaluated for predicting short-term disease by using the area under the receiver operating characteristics curve (AUC) and relative standard deviation (RSD). The optimal models were further analyzed for predictive performance on overall survival. Results: A total of the 1,092 models (156 with radiomics features and 936 with deep learning features) were constructed. Radiomics_GINI_Nearest Neighbors (RGNN) and Resnet50_MIM_Nearest Neighbors (RMNN) were identified as optimal models, with the AUC of 0.87 and 0.94, accuracy of 0.89 and 0.92, sensitivity of 0.88 and 0.97, specificity of 0.90 and 0.90, precision of 0.87 and 0.83, F1 score of 0.89 and 0.92, and RSD of 1.30 and 0.26, respectively. Kaplan-Meier survival analysis showed that RGNN and RMNN were associated with better OS (p = 0.006 for RGNN and p = 0.033 for RMNN). Conclusion: Pretreatment CT-based radiomics/deep learning models could non-invasively and efficiently predict outcomes in HCC patients who received combined therapy of TACE and TKI.

Multiple Aneurysms of the Subclavian-Axillary, Abdominal Aortoiliac, Lower Extremity, and Coronary Arteries in A Pediatric Patient of Takayasu Arteritis.

Takayasu arteritis is a chronic inflammatory vasculitis affecting mainly the aorta and its branches. Stenosis and occlusion of the involved vessels usually develop; however, their dilation and aneurysmal formation are extremely rare. Although aneurysmal disease has been reported in adults with Takayasu arteritis, it is a rare entity in children. The present report described an 11-year-old male found to have the subclavian-axillary, abdominal aortoiliac, lower extremity artery aneurysms with mural thrombi. Aneurysms were also found at the proximal and middle segments of the right coronary artery. The patient was conservatively treated with corticosteroid in addition to antiplatelet and anticoagulation agents.

The role of natural polymers in bone tissue engineering.

Bone is a dynamic self-healing organ and a continuous remodeling ensures the restoration of the bone structure and function over time. However, bone remodeling is not able to repair large traumatic injuries. Therefore, surgical interventions and bone substitutes are required. The aim of bone tissue engineering is to repair and regenerate tissues and engineered a bone graft as a bone substitute. To met this goal, several natural or synthetic polymers have been used to develop a biocompatible and biodegradable polymeric construct. Among the polymers, natural polymers have higher biocompatibility, excellent biodegradability, and no toxicity. So far, collagen, chitosan, gelatin, silk fibroin, alginate, cellulose, and starch, alone or in combination, have been widely used in bone tissue engineering. These polymers have been used as scaffolds, hydrogels, and micro-nanospheres. The functionalization of the polymer with growth factors and bioactive glasses increases the potential use of polymers for bone regeneration. As bone is a dynamic highly vascularized tissue, the vascularization of the polymeric scaffolds is vital for successful bone regeneration. Several in vivo and in vitro strategies have been used to vascularize the polymeric scaffolds. In this review, the application of the most commonly used natural polymers is discussed.

Deep Learning for Detecting Cerebral Aneurysms with CT Angiography.

Background Cerebral aneurysm detection is a challenging task. Deep learning may become a supportive tool for more accurate interpretation. Purpose To develop a highly sensitive deep learning-based algorithm that assists in the detection of cerebral aneurysms on CT angiography images. Materials and Methods Head CT angiography images were retrospectively retrieved from two hospital databases acquired across four different scanners between January 2015 and June 2019. The data were divided into training and validation sets; 400 additional independent CT angiograms acquired between July and December 2019 were used for external validation. A deep learning-based algorithm was constructed and assessed. Both internal and external validation were performed. Jackknife alternative free-response receiver operating characteristic analysis was performed. Results A total of 1068 patients (mean age, 57 years ± 11 [standard deviation]; 660 women) were evaluated for a total of 1068 CT angiograms encompassing 1337 cerebral aneurysms. Of these, 534 CT angiograms (688 aneurysms) were assigned to the training set, and the remaining 534 CT angiograms (649 aneurysms) constituted the validation set. The sensitivity of the proposed algorithm for detecting cerebral aneurysms was 97.5% (633 of 649; 95% CI: 96.0, 98.6). Moreover, eight new aneurysms that had been overlooked in the initial reports were detected (1.2%, eight of 649). With the aid of the algorithm, the overall performance of radiologists in terms of area under the weighted alternative free-response receiver operating characteristic curve was higher by 0.01 (95% CI: 0.00, 0.03). Conclusion The proposed deep learning algorithm assisted radiologists in detecting cerebral aneurysms on CT angiography images, resulting in a higher detection rate. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Kallmes and Erickson in this issue.

[Experimental study on transplantation of embryonic stem cells in treating spinal cord injury].

OBJECTIVE: To observe the effect of transplantation of embryonic stem cell (ES) on neurological functional recovery of injured spinal cord in adult mouse. METHODS: The ES cells were cultured and induced in vitro. Fifty C57/BL6J mice were made animal model of semi-cut mice of T9,10. The ES cell-derived neural precursors cells were transplanted into the vertebral canal around injured spinal cord semi-cut mice. Twenty-eight C57/BL6J mice were randomly divided into three groups: sham operation group(group A, n=9), operation/cell group (group B, n = 10), and operation/DMEM group (group C, n = 9). RT-PCR analysis, X-gal staining and immunofluorescence were used to observe the cells survival and differentiation in the spinal crod. BBB test was performed to study functional improvement. RESULTS: ES cells induced and cultured in vitro displayed clonal growth with circle or ovoid shape and had one or more nucleoli. RT-PCR result showed that the induced ES cells expressed mRNA of Nestin and microtubule-associated protein, but did not express glial fibrillory acidic protein (GFAP). There was statistically significant difference in BBB scoring between group A and groups B, C after operation (P <0.01). There was statistically significant difference in BBB scoring at 1, 2 and 4 weeks of operation (P < 0.01), but no statistically significant difference at 6 and 8 weeks of operation between groups B and C (P>0.05). The X-gal staining results were positive in group B and negative in groups A and C. The immunoflurescence result showed neurofilament green fluor and no expression of GFAP in injured spinal cord region. CONCLUSION: After transplantation, ES cell-derived cells can survive, transfer into the injury position, and differentiate into neurons, but spinal cord function has no obvious improvement.