Compare deep learning model and conventional logistic regression model for the identification of unstable saccular intracranial aneurysms in computed tomography angiography.

Background: It is crucial to distinguish unstable from stable intracranial aneurysms (IAs) as early as possible to derive optimal clinical decision-making for further treatment or follow-up. The aim of this study was to investigate the value of a deep learning model (DLM) in identifying unstable IAs from computed tomography angiography (CTA) images and to compare its discriminatory ability with that of a conventional logistic regression model (LRM). Methods: From August 2011 to May 2021, a total of 1,049 patients with 681 unstable IAs and 556 stable IAs were retrospectively analyzed. IAs were randomly divided into training (64%), internal validation (16%), and test sets (20%). Convolutional neural network (CNN) analysis and conventional logistic regression (LR) were used to predict which IAs were unstable. The area under the curve (AUC), sensitivity, specificity and accuracy were calculated to evaluate the discriminating ability of the models. One hundred and ninety-seven patients with 229 IAs from Banan Hospital were used for external validation sets. Results: The conventional LRM showed 11 unstable risk factors, including clinical and IA characteristics. The LRM had an AUC of 0.963 [95% confidence interval (CI): 0.941-0.986], a sensitivity, specificity and accuracy on the external validation set of 0.922, 0.906, and 0.913, respectively, in predicting unstable IAs. In predicting unstable IAs, the DLM had an AUC of 0.771 (95% CI: 0.582-0.960), a sensitivity, specificity and accuracy on the external validation set of 0.694, 0.929, and 0.782, respectively. Conclusions: The CNN-based DLM applied to CTA images did not outperform the conventional LRM in predicting unstable IAs. The patient clinical and IA morphological parameters remain critical factors for ensuring IA stability. Further studies are needed to enhance the diagnostic accuracy.

MXene/Hydrogel-based bioelectronic nose for the direct evaluation of food spoilage in both liquid and gas-phase environments.

Various bioelectronic noses have been recently developed for mimicking human olfactory systems. However, achieving direct monitoring of gas-phase molecules remains a challenge for the development of bioelectronic noses due to the instability of receptor and the limitations of its surrounding microenvironment. Here, we report a MXene/hydrogel-based bioelectronic nose for the sensitive detection of liquid and gaseous hexanal, a signature odorant from spoiled food. In this study, a conducting MXene/hydrogel structure was formed on a sensor via physical adsorption. Then, canine olfactory receptor 5269-embedded nanodiscs (cfOR5269NDs) which could selectively recognize hexanal molecules were embedded in the three-dimensional (3D) MXene/hydrogel structures using glutaraldehyde as a linker. Our MXene/hydrogel-based bioelectronic nose exhibited a high selectivity and sensitivity for monitoring hexanal in both liquid and gas phases. The bioelectronic noses could sensitively detect liquid and gaseous hexanal down to 10-18 M and 6.9 ppm, and they had wide detection ranges of 10-18 - 10-6 M and 6.9-32.9 ppm, respectively. Moreover, our bioelectronic nose allowed us to monitor hexanal levels in fish and milk. In this respect, our MXene/hydrogel-based bioelectronic nose could be a practical strategy for versatile applications such as food spoilage assessments in both liquid and gaseous systems.

A deep-learning-based framework for identifying and localizing multiple abnormalities and assessing cardiomegaly in chest X-ray.

Accurate identification and localization of multiple abnormalities are crucial steps in the interpretation of chest X-rays (CXRs); however, the lack of a large CXR dataset with bounding boxes severely constrains accurate localization research based on deep learning. We created a large CXR dataset named CXR-AL14, containing 165,988 CXRs and 253,844 bounding boxes. On the basis of this dataset, a deep-learning-based framework was developed to identify and localize 14 common abnormalities and calculate the cardiothoracic ratio (CTR) simultaneously. The mean average precision values obtained by the model for 14 abnormalities reached 0.572-0.631 with an intersection-over-union threshold of 0.5, and the intraclass correlation coefficient of the CTR algorithm exceeded 0.95 on the held-out, multicentre and prospective test datasets. This framework shows an excellent performance, good generalization ability and strong clinical applicability, which is superior to senior radiologists and suitable for routine clinical settings.

Research progress on preparation methods and sensing applications of molecularly imprinted polymer-aptamer dual recognition elements.

The aptamer (Apt) and the molecularly imprinted polymer (MIP), as effective substitutes for antibodies, have received widespread attention from researchers because of their creation. However, the low stability of Apt in harsh detection environment and the poor specificity of MIP have hindered their development. Therefore, some researchers have attempted to combine MIP with Apt to explore whether the effect of "1 + 1 > 2" can be achieved. Since its first report in 2013, MIP-Apt dual recognition elements have become a highly focused research direction in the fields of biology and chemistry. MIP-Apt dual recognition elements not only possess the high specificity of Apt and the high stability of MIP in harsh detection environment, but also have high sensitivity and affinity. They have been successfully applied in medical diagnosis, food safety, and environmental monitoring fields. This article provides a systematic overview of three preparation methods for MIP-Apt dual recognition elements and their application in eight different types of sensors. It also provides effective insights into the problems and development directions faced by MIP-Apt dual recognition elements.

Hydrogel-based Bioelectronic Tongue for the Evaluation of Umami Taste in Fermented Fish.

Bioelectronic tongues based on umami taste receptors have recently been reported for versatile applications such as food analyses. However, their practical applications are still limited, partly due to their limited stability and non-specific responses in real sample environments. Herein, we have developed a hydrogel-based bioelectronic tongue for the sensitive assessment of umami intensity in fish extract samples. In this study, the T1R1 venus flytrap of an umami taste receptor was immobilized on the gold floating electrodes of a carbon nanotube-based field-effect transistor. A polyacrylamide conducting hydrogel film was further hybridized on the sensor surface via physical adsorption, which could provide a good physiological environment to maintain the activity of receptors due to its excellent hydrophilicity and biocompatibility. The bioelectronic tongue with a receptor-embedded hydrogel structure showed a sensitive detection of umami substances down to 1 fM, and it also had a wide detection range of 10-15-10-2 M for monosodium glutamate and disodium inosinate, which covers the human taste threshold. More importantly, the proposed sensor could significantly reduce the non-specific binding of non-target molecules to a carbon nanotube channel as well as exhibit long-term stability, enabling sensitive detection of umami substances even in fish extract samples. Our hydrogel-based bioelectronic tongue provides a promising platform for future applications such as the flavor evaluation of foods and beverages.

Novel Broad-Spectrum Aptamer Used to Build Fe-Co Nanoporous Carbon Highly Sensitive Sensing Platform for the Detection of Multiple Organophosphorus Pesticides.

In this work, broad-spectrum aptamers for organophosphorus pesticides (OPs) were obtained by alternate target systematic evolution of ligands by exponential enrichment screening. The secondary and tertiary structure analyses of the aptamer inferred that the neck-loop structure formed a G-triplex structure with the target. In addition, optimization of the sheared aptamer resulted in a stronger affinity (Kd = 86.74 nM), which was increased by 2 orders of magnitude compared to similar aptamers. A novel electrochemical biosensor was prepared by modifying an aptamer labeled with an electroactive substance (methylene blue) on the surface of nanoporous carbon containing Fe-Co (Fe-Co/NPC). When a target bound to the aptamer, a G-triplex structure was formed close to the electrode surface. The aptamer phosphate backbone labeled with methylene blue enhanced the electron-transfer efficiency and resulted in signal changes. The biosensor exhibited an excellent sensitivity (7.32 fM) and a wide detection range (1 × 10-13 to 1 × 10-3 M) for OPs under optimal conditions, enabling simultaneous detection of multiple OPs in vegetables.

A Prospective Follow-up Study on Risk Factors to Predict the Progression of Unruptured Intracranial Aneurysms on Enhanced HR-MRI.

RATIONALE AND OBJECTIVES: To prospectively investigate the potential correlation between qualitative and quantitative assessment of aneurysm wall enhancement (AWE) on initial enhanced high-resolution magnetic resonance imaging (HR-MRI) and aneurysm progression during follow-up. MATERIALS AND METHODS: From June 2016 to January 2021, we prospectively recruited patients with unruptured intracranial aneurysms (UIAs) for enhanced HR-MRI examination. The patients' demographic and clinical data and aneurysm characteristics, including AWE features, were collected and analyzed. Follow-up images were compared to evaluate IA progression. Univariate and multivariate Cox proportional hazards regression analyses were performed to identify the risk factors associated with aneurysm progression. RESULTS: Seventy-seven patients with 95 UIAs met our research criteria, and the median follow-up time was 15.7 months. Progression was observed in 18 aneurysms; the remaining 77 remained stable. Progressive UIAs were larger in size, more frequently displayed obvious AWE and showed a higher enhancement ratio (ER) than nonprogressive UIAs. Multivariate Cox regression analysis showed that both ER (hazard ratio, 6.304, p < 0.001) and aneurysm size (hazard ratio, 1.343, p = 0.014) were independent risk factors for aneurysm progression. The combination of ER and aneurysm size had an area under the curve of 0.920 for the prediction of aneurysm progression, with a sensitivity of 88.9% and specificity of 87.0%. CONCLUSION: A higher ER value of the aneurysm wall and a larger aneurysm size on initial HR-MRI may predict an increased risk of aneurysm progression, which suggests that closer monitoring by imaging or preventive intervention may be required for the clinical management of these aneurysms.

Portable electrochemiluminescence detection system based on silicon photomultiplier single photon detector and aptasensor for the detection of tetracycline in milk.

In this work, a portable electrochemiluminescence (ECL) detection system based on silicon photomultiplier (SiPM) single photon detector was proposed for the detection of ECL signals on a screen-printed electrode (SPE). This instrument innovatively used SiPM single photon detector to detect the ECL signal, which solved friability and bloat caused by the high operating voltage and the limitation of detection components in the traditional ECL detection instrument. This detection instrument showed excellent electrochemical and ECL detection performance. On this basis, an aptasensor based on silver (core)-gold (shell) bimetallic nanoparticles (Ag@AuNPs) was constructed for the detection of tetracycline (TET) in milk on SPE. Here, Ag@AuNPs had a significant effect in enhancing luminol ECL signal and immobilizing aptamer. The concentration of TET was detected according to the changes of the ECL signal intensity of the detection instrument. This instrument exhibited an excellent linearity ranging from 0.01 ng/mL to 1,000 ng/mL for the detection of TET, and a limit of detection (LOD) was 0.0053 ng/mL. The developed portable ECL detection instrument provides a new platform for the detection of small molecule contaminants.

Effects of genes required for exopolysaccharides biosynthesis in Lacticaseibacillus paracasei S-NB on cell surface characteristics and probiotic properties.

The cell surface of lactic acid bacteria (LAB) plays an essential role in cell-cell and cell-host interactions. Exopolysaccharides (EPSs) are produced on the cell surface of LAB or in the surrounding medium and are considered to be in favor of the strain- specific probiotic surface characteristics. In this work, the structure features of EPS from Lacticaseibacillus paracasei S-NB were analyzed preliminarily, and the genes involved in EPS biosynthesis of S-NB strain were hypothesized and annotated, and their role in phenotypic characteristics were demonstrated by gene deletion analysis. Four mutant strains with deletion of crucial genes involved in EPS synthesis were analyzed for strain characteristics that are closely related to their ability to interact with the host intestinal epithelium cells, including strain surface characteristics and viability under the gastrointestinal stress conditions (both acid and bile stress). Furthermore, the adherence and immunomodulatory properties of wild-type S-NB and its mutant strains were compared using Caco-2 and RAW 264.7 cell lines, respectively. Taken together, the results indicated the importance of genes associated with EPS biosynthesis in L. paracasei S-NB as a determinant in strain surface characteristics and cell-host interaction, especially for S-NB_2176 (responsible for EPS polymerization) and S-NB_2175 (responsible for CpsD/CapB family tyrosine-protein kinase).

CT findings of Talaromyces marneffei infection among HIV patients with lymphadenopathy.

Background: Talaromyces marneffei (T. marneffei) is an opportunistic fungal pathogen commonly found in human immunodeficiency virus (HIV) patients that often infects lymph nodes. Knowledge about the computed tomography (CT) characteristics of T. marneffei lymphadenopathy in HIV patients is limited. The aim of this study was to investigate the clinical and CT characteristics of T. marneffei lymphadenopathy to improve its diagnosis and promote recognition of this type of infection in radiology. Methods: Between February 2019 and June 2021, we retrospectively reviewed the clinical features and CT characteristics of T. marneffei lymphadenopathy in 21 HIV patients. Results: The clinical symptoms of T. marneffei infection are non-specific. Anemia (100%), fever (85.7%) and cough and sputum production (76.2%) were the most frequent symptoms. Multiple lymphadenopathies, mainly in the mediastinum (76.2%) and mesentery (82.4%), can be fused (14.3%) and necrotic (52.4%), with slight (41.7%) and moderate enhancement (58.3%) that is heterogeneous. In addition to involving the lymph nodes, the lesions involved the lungs (81.0%), liver and spleen (42.9%), and small intestine (14.3%). Conclusions: T. marneffei is prone to affecting lymph nodes and extranodal organs in HIV patients. Although the clinical manifestations of T. marneffei infection are not specific, the possibility of T. marneffei infection should be considered if CT findings indicate multiple lesion sites.

Structural and prebiotic activity analysis of the polysaccharide produced by Lactobacillus helveticus SNA12.

Exopolysaccharide (EPS) isolated from Lactobacillus helveticus SNA12 were purified, and one fraction (SNA12-EPS) was obtained. The structure of SNA12-EPS was proposed using Fourier-transform infrared spectroscopy, gas chromatography-mass spectrometry, and nuclear magnetic resonance. Results showed that SNA12-EPS was rich in galactose and glucose with the molar ratios of 2.1:1.0, and SNA12-EPS possessed the repeat units of →3)-ß-D-Glcp-(1→3)-ß-D-Glcp-(1→4)-ß-D-Galp-(1→ with an average molecular weight of 3.81 × 105 Da. The scanning electron microscope results showed that SNA12-EPS had a tight structure with a smooth and uneven surface. Furthermore, the prebiotic potential of SNA12-EPS was performed using in vitro simulated digestion with human faecal fermentation. SNA12-EPS was not digested by digestive juice, and it could markedly regulate the gut microbiota composition by increasing the relative abundances of Parabacteroides and Blautia and decreasing the abundance of pathogenic bacteria of Fusobacterium. Additionally, SNA12-EPS improved the ability of gut microbiota to produce short-chain fatty acids.

Novel colorimetric aptasensor based on MOF-derived materials and its applications for organophosphorus pesticides determination.

For the visual detection of four organophosphorus pesticides (OPs), a colorimetric aptasensor was developed based on aptamer-mediated bimetallic metal-organic frameworks (MOFs) nano-polymers. Fe-Co magnetic nanoparticles (MNPs) and Fe-N-C nanozymes were prepared based on pyrolytic reaction, and were labeled with broad spectrum aptamers and complementary chains of organophosphorus pesticides respectively. The hybridization of aptamers and complementary chains led to the formation of nano-polymers. In the presence of target pesticides, they competed with complementary chains for aptamers on Fe-Co MNPs, resulting in a large number of Fe-N-C nanozymes signal labels being released into the supernatant. Fe-N-C nanozymes showed similar activity to peroxidase and catalyzed the 3,3',5,5'-tetramethylbenzidine-hydrogen peroxide (TMB-H2O2) color system to turn the solution blue-green under mild conditions. The magnetic probes had good selectivity and sensitivity, and were easily separated by magnetic absorption. The sensor functioned well under optimal conditions, demonstrating good stability and specificity for four pesticides: phorate, profenofos, isocarbophos and omethoate, and the detection limits of four pesticides were as low as 0.16 ng/mL, 0.16 ng/mL, 0.03 ng/mL and 1.6 ng/mL respectively, and the recovery rate of OPs residue in vegetable samples was satisfactory. The work described here provided a simple, rapid and sensitive way to construct a biosensor.

A novel umami electrochemical biosensor based on AuNPs@ZIF-8/Ti3C2 MXene immobilized T1R1-VFT.

The bioelectronic tongues based on taste receptors have been emerging with human-like taste perception. However, the practical applications of the receptor-based biosensors were restricted by their narrow and low dynamic ranges. Here, a novel immobilization strategy based on AuNPs@ZIF-8/Ti3C2 MXene was developed to immobilize the umami ligand binding domain (T1R1-VFT), to fabricate an umami biosensor for umami substances detection. Through the synergic effect of AuNPs@ZIF-8 and Ti3C2 MXene, the capacity to load T1R1-VFT was effectively increased, and the response signal was also amplified by approximately 3 times. The proposed biosensor showed an ultrawide dynamic range of 10-11-10-3 M, and a high upper limit of detection, which was closer to the human taste threshold and suitable for detecting foods rich in umami substances. Additionally, the biosensor was successfully applied to detect real samples and analyze the synergistic effects of binary umami substances.

Identification of ruptured intracranial aneurysms using the aneurysm-specific prediction score in patients with multiple aneurysms with subarachnoid hemorrhages- a Chinese population based external validation study.

BACKGROUND: For patients with aneurysmal subarachnoid hemorrhages (SAHs) and multiple intracranial aneurysms (MIAs), a simple and fast imaging method that can identify ruptured intracranial aneurysms (RIAs) may have great clinical value. We sought to use the aneurysm-specific prediction score to identify RIAs in patients with MIAs and evaluate the aneurysm-specific prediction score. METHODS: Between May 2018 and May 2021, 134 patients with 290 MIAs were retrospectively analyzed. All patients had an SAH due to IA rupture. CT angiography (CTA) was used to assess the maximum diameter, shape, and location of IAs to calculate the aneurysm-specific prediction score. Then, the aneurysm-specific prediction score was applied to RIAs in patients with MIAs. RESULTS: The IAs with the highest aneurysm-specific prediction scores had not ruptured in 17 (12.7%) of the 134 patients with 290 MIAs. The sensitivity, specificity, false omission rate, diagnostic error rate, and diagnostic accuracy of the aneurysm-specific prediction score were higher than those of the maximum diameter, shape, and location of IAs. CONCLUSIONS: The present study suggests that the aneurysm-specific prediction score has high diagnostic accuracy in identifying RIAs in patients with MIAs and SAH, but that it needs further evaluation.

JNK Signaling Promotes Bladder Cancer Immune Escape by Regulating METTL3-Mediated m6A Modification of PD-L1 mRNA.

The RNA N6-methyladenosine (m6A) writer methyltransferase-like 3 (METTL3) is upregulated in many types of cancer and promotes cancer progression by increasing expression of several oncogenes. Therefore, a better understanding of the mechanisms regulating METTL3 expression and the key targets of METTL3 in cancer cells could provide new therapeutic targets. In this study, we found that activated JNK signaling is associated with increased METTL3 expression in bladder cancer. Knockdown of JNK1 or administration of a JNK inhibitor impaired the binding of c-Jun with the METTL3 promoter, thereby decreasing the expression of METTL3 and global RNA m6A levels. Moreover, RNA m6A sequencing indicated enrichment of m6A in the 3'-UTR of immune checkpoint PD-L1 mRNA, which could be recognized by the m6A reader IGF2BP1 to mediate RNA stability and expression levels of PD-L1. Inhibition of JNK signaling suppressed m6A abundance in PD-L1 mRNA, leading to decreased PD-L1 expression. Functionally, METTL3 was essential for bladder cancer cells to resist the cytotoxicity of CD8+ T cells by regulating PD-L1 expression. Additionally, JNK signaling contributed to tumor immune escape in a METTL3-dependent manner both in vitro and in vivo. These data reveal the JNK/METTL3 axis as a mechanism of aberrant m6A modification and immune regulation in bladder cancer. SIGNIFICANCE: The identification of a novel m6A-dependent mechanism underlying immune system evasion by bladder cancer cells reveals JNK signaling as a potential target for bladder cancer immunotherapy.

Lactobacillus plantarum Lp3a improves functional constipation: evidence from a human randomized clinical trial and animal model.

Background: Functional constipation (FC) is a common gastrointestinal (GI) disorder characterized by symptoms of constipation without a clear physiologic or anatomic cause. Gut microbiome dysbiosis has been postulated to be a factor in the development of FC, and treatment with probiotic regimens, including strains of Lactobacillus plantarum (L. plantarum), has demonstrated efficacy in managing symptoms. To further understand the role of L. plantarum in GI health, we conducted an animal study and a randomized, double-blind, placebo-controlled clinical trial to evaluate the effect of a specific sub-strain, Lp3a, on FC. Methods: For the animal study, male Kunming mice were treated with doses of L. plantarum Lp3a ranging from 0.67 to 2.00 g/kg or an equivalent amount of placebo for 15 days prior to the induction of constipation via 20 mL/kg of 25% diphenoxylate solution. GI motility parameters including intestinal motion and stool amount were then assessed. In the human study, 120 patients with FC were randomized to treatment [L. plantarum Lp3a; 2×1.0×1010 (colony forming units; CFU) ×7 days] or control groups (n=60 each). The primary endpoint was survey information on FC signs/symptoms. Participants and observers were blinded to group allocation. A subset of 20 Lp3a treated patients underwent pre- and post-treatment 16 s ribosomal ribonucleic acid (rRNA) gene sequencing. Whole genome sequencing (WGS) of L. plantarum Lp3a was also performed. Results: Lp3a-treated mice showed significantly improved intestinal motion, reduced time to first defecation, and increased stool amounts. Similarly, patients in the treatment group (n=59) reported significant improvements in FC signs/symptoms compared to controls (n=58; all P<0.05). Although 16 s rRNA sequencing revealed no significant variations between pre- and post-treatment samples, WGS of Lp3a itself revealed several biological pathways that may underlie the relief of FC symptoms in animals and humans, including methane and fatty acid metabolism and bile acid biosynthesis. Conclusions: We found that the use of the novel probiotic sub-strain, L. plantarum Lp3a, led to clinically significant improvements in FC in both mice and humans, and identified the potential biological mechanisms underlying this activity.

Low attenuation plaque volume on coronary computed tomography angiography is associated with plaque progression.

BACKGROUND: Patient-related clinical factors, laboratory factors, and some imaging factors may lead to statistical bias when investigating coronary plaque progression. In this study, we avoided patient characteristics by comparing morphological characteristics of plaque progression and nonprogression within the same patient with multiple plaques. METHODS: From August 2011 to December 2018, 177 consecutive patients with 424 plaques who were followed with coronary computed tomography angiography (CTA) were reviewed retrospectively. Follow-up images of the plaques were used to determine whether the plaque volume or stenosis grade increased. The plaques were divided into progressive and nonprogressive groups. Logistic regression analysis was used to identify the factors associated with plaque progression. Through clinical follow-up, we analyzed whether the factors associated with plaque progression were related to major adverse cardiac events (MACEs). RESULTS: There were 223 plaques that progressed during a mean follow-up period of 27.6 ± 15.9 months. The univariate logistic regression model revealed that only low attenuation plaque (LAP) volume (P = 0.02) was associated with plaque progression. After a mean post-CTA follow-up period of 36.7 ± 18.4 months, 37 patients experienced MACEs, and LAP volume was significantly related to future MACEs. CONCLUSION: Only a high baseline LAP volume was associated with plaque progression, and patients with progressive plaques and a high LAP volume were more likely to have future MACEs. More attention should be given to plaques with LAP volumes larger than 2.4 mm3.

MRCP Combined With CT Promotes the Differentiation of Benign and Malignant Distal Bile Duct Strictures.

OBJECTIVE: To determine whether contrast-enhanced computed tomography (CT) can promote the identification of malignant and benign distal biliary strictures (DBSs) compared to the use of magnetic resonance cholangiopancreatography (MRCP) alone and to identify imaging findings of malignant DBSs. MATERIALS AND METHODS: A total of 168 consecutive patients with confirmed DBSs were reviewed. MRCP alone and MRCP combined with CT images were blindly analyzed by two radiologists (e.g., stricture pattern, margins), and malignant or benign DBSs were identified based on surgical findings, endoscopy findings, or follow-up. The diagnostic accuracy of the two reviewers using MRCP alone and MRCP combined with CT were evaluated. MRCP and CT features of malignant and benign DBSs were compared using multiple logistic regression analysis to identify independent malignant risk factors. RESULTS: MRCP combined with CT examination could improve the diagnostic accuracy, which increased from 70.2% to 81.5% in Doctor A and from 85.1% to 89.3% in Doctor B. The multiple logistic regression model revealed that stricture length [odds ratio (OR) 1.070, P=0.016], angle of the DBS (OR 1.061, P<0.001), double duct sign (OR 4.312, P=0.003) and low density in the arterial phase (OR 0.319, P=0.018) were associated with malignant DBS. A scoring model incorporating these four factors was established; at a threshold value of 1.75, and the sensitivity and specificity for the detection of malignant DBSs were 73.5 and 85.9%, respectively. CONCLUSIONS: Compared to the use of MRCP alone, MRCP combined with contrast-enhanced CT can improve the accuracy of DBS diagnosis. The scoring model accurately predicts malignant DBSs and helps make treatment decisions.

Radiomic Features of Multi-ROI and Multi-Phase MRI for the Prediction of Microvascular Invasion in Solitary Hepatocellular Carcinoma.

OBJECTIVES: To develop and validate an MR radiomics-based nomogram to predict the presence of MVI in patients with solitary HCC and further evaluate the performance of predictors for MVI in subgroups (HCC ≤ 3 cm and > 3 cm). MATERIALS AND METHODS: Between May 2015 and October 2020, 201 patients with solitary HCC were analysed. Radiomic features were extracted from precontrast T1WI, arterial phase, portal venous phase, delayed phase and hepatobiliary phase images in regions of the intratumoral, peritumoral and their combining areas. The mRMR and LASSO algorithms were used to select radiomic features related to MVI. Clinicoradiological factors were selected by using backward stepwise regression with AIC. A nomogram was developed by incorporating the clinicoradiological factors and radiomics signature. In addition, the radiomic features and clinicoradiological factors related to MVI were separately evaluated in the subgroups (HCC ≤ 3 cm and > 3 cm). RESULTS: Histopathological examinations confirmed MVI in 111 of the 201 patients (55.22%). The radiomics signature showed a favourable discriminatory ability for MVI in the training set (AUC, 0.896) and validation set (AUC, 0.788). The nomogram incorporating peritumoral enhancement, tumour growth type and radiomics signature showed good discrimination in the training (AUC, 0.932) and validation sets (AUC, 0.917) and achieved well-fitted calibration curves. Subgroup analysis showed that tumour growth type was a predictor for MVI in the HCC ≤ 3 cm cohort and peritumoral enhancement in the HCC > 3 cm cohort; radiomic features related to MVI varied between the HCC ≤ 3 cm and HCC > 3 cm cohort. The performance of the radiomics signature improved noticeably in both the HCC ≤ 3 cm (AUC, 0.953) and HCC > 3 cm cohorts (AUC, 0.993) compared to the original training set. CONCLUSIONS: The preoperative nomogram integrating clinicoradiological risk factors and the MR radiomics signature showed favourable predictive efficiency for predicting MVI in patients with solitary HCC. The clinicoradiological factors and radiomic features related to MVI varied between subgroups (HCC ≤ 3 cm and > 3 cm). The performance of radiomics signature for MVI prediction was improved in both the subgroups.