Artificial Intelligence-Driven Morphology-Based Enrichment of Malignant Cells from Body Fluid.

Cell morphology is a fundamental feature used to evaluate patient specimens in pathologic analysis. However, traditional cytopathology analysis of patient effusion samples is limited by low tumor cell abundance coupled with the high background of nonmalignant cells, restricting the ability of downstream molecular and functional analyses to identify actionable therapeutic targets. We applied the Deepcell platform that combines microfluidic sorting, brightfield imaging, and real-time deep learning interpretations based on multidimensional morphology to enrich carcinoma cells from malignant effusions without cell staining or labels. Carcinoma cell enrichment was validated with whole genome sequencing and targeted mutation analysis, which showed a higher sensitivity for detection of tumor fractions and critical somatic variant mutations that were initially at low levels or undetectable in presort patient samples. Our study demonstrates the feasibility and added value of supplementing traditional morphology-based cytology with deep learning, multidimensional morphology analysis, and microfluidic sorting.

Datanator: an integrated database of molecular data for quantitatively modeling cellular behavior.

Integrative research about multiple biochemical subsystems has significant potential to help advance biology, bioengineering and medicine. However, it is difficult to obtain the diverse data needed for integrative research. To facilitate biochemical research, we developed Datanator (https://datanator.info), an integrated database and set of tools for finding clouds of multiple types of molecular data about specific molecules and reactions in specific organisms and environments, as well as data about chemically-similar molecules and reactions in phylogenetically-similar organisms in similar environments. Currently, Datanator includes metabolite concentrations, RNA modifications and half-lives, protein abundances and modifications, and reaction rate constants about a broad range of organisms. Going forward, we aim to launch a community initiative to curate additional data. Datanator also provides tools for filtering, visualizing and exporting these data clouds. We believe that Datanator can facilitate a wide range of research from integrative mechanistic models, such as whole-cell models, to comparative data-driven analyses of multiple organisms.

Safety and effectiveness of mucosal traction using a snare combined with endoclips to assist the resection of esophageal intraepithelial neoplasia: a propensity score matching analysis.

Currently, the reports on esophageal endoscopic submucosal dissection (ESD) assisted by traction with a snare are rare. Because a snare is a commonly used endoscopic accessory and is easily available, its application in mucosal traction is worth exploring. The present study aims to evaluate the safety and effectiveness of snare-endoclip traction-assisted ESD for esophageal intraepithelial neoplasia. Cases of esophageal intraepithelial neoplasia resected using ESD in the Digestive Endoscopy Center of Guangzhou Nanfang Hospital, China from June 2013 to March 2019 were retrospectively analyzed. The procedure of snare-endoclip traction-assisted ESD was compared with nontraction-assisted ESD by using a propensity score matching analysis. Operation time, en bloc and R0 resection, intra- and postoperative complications, and surgery-related costs were mainly evaluated. Overall, 99 cases of esophageal intraepithelial neoplasia under tissue biopsy were included in the present study. Further, 22 exact matched pairs were obtained. There were no differences in en bloc and R0 resection rates, intra- and postoperative complications, and costs of disposable surgical accessories between the traction group and the nontraction group. However, median operation time showed a significant difference: traction group, 50.0 min (range, 20-100 min); nontraction group, 70.0 min (range, 35-133 min), P=0.012. In conclusion, snare-endoclip traction-assisted ESD for esophageal intraepithelial neoplasia was safe and shortened operation time in the study, thereby improving the efficiency of ESD. Despite the additional use of a snare and endoclips for traction, the total costs of endoscopic accessories seemed not to be increased.

Machine-learning based MRI radiomics models for early detection of radiation-induced brain injury in nasopharyngeal carcinoma.

BACKGROUND: Early radiation-induced temporal lobe injury (RTLI) diagnosis in nasopharyngeal carcinoma (NPC) is clinically challenging, and prediction models of RTLI are lacking. Hence, we aimed to develop radiomic models for early detection of RTLI. METHODS: We retrospectively included a total of 242 NPC patients who underwent regular follow-up magnetic resonance imaging (MRI) examinations, including contrast-enhanced T1-weighted and T2-weighted imaging. For each MRI sequence, four non-texture and 10,320 texture features were extracted from medial temporal lobe, gray matter, and white matter, respectively. The relief and 0.632 + bootstrap algorithms were applied for initial and subsequent feature selection, respectively. Random forest method was used to construct the prediction model. Three models, 1, 2 and 3, were developed for predicting the results of the last three follow-up MRI scans at different times before RTLI onset, respectively. The area under the curve (AUC) was used to evaluate the performance of models. RESULTS: Of the 242 patients, 171 (70.7%) were men, and the mean age of all the patients was 48.5 ± 10.4 years. The median follow-up and latency from radiotherapy until RTLI were 46 and 41 months, respectively. In the testing cohort, models 1, 2, and 3, with 20 texture features derived from the medial temporal lobe, yielded mean AUCs of 0.830 (95% CI: 0.823-0.837), 0.773 (95% CI: 0.763-0.782), and 0.716 (95% CI: 0.699-0.733), respectively. CONCLUSION: The three developed radiomic models can dynamically predict RTLI in advance, enabling early detection and allowing clinicians to take preventive measures to stop or slow down the deterioration of RTLI.

Prognostic value of the radiomics-based model in progression-free survival of hypopharyngeal cancer treated with chemoradiation.

PURPOSE: To develop a radiomics-based model to stratify the risk of early progression (local/regional recurrence or metastasis) among patients with hypopharyngeal cancer undergoing chemoradiotherapy and modify their pretreatment plans. MATERIALS AND METHODS: We randomly assigned 113 patients into two cohorts: training (n = 80) and validation (n = 33). The radiomic significant features were selected in the training cohort using least absolute shrinkage and selection operator and Akaike information criterion methods, and they were used to build the radiomic model. The concordance index (C-index) was applied to evaluate the model's prognostic performance. A Kaplan-Meier analysis and the log-rank test were used to assess risk stratification ability of models in predicting progression. A nomogram was plotted to predict individual risk of progression. RESULTS: Composed of four significant features, the radiomic model showed good performance in stratifying patients into high- and low-risk groups of progression in both the training and validation cohorts (log-rank test, p = 0.00016, p = 0.0063, respectively). Peripheral invasion and metastasis were selected as significant clinical variables. The combined radiomic-clinical model showed good discriminative performance, with C-indices 0.804 (95% confidence interval (CI), 0.688-0.920) and 0.756 (95% CI, 0.605-0.907) in the training and validation cohorts, respectively. The median progression-free survival (PFS) in the high-risk group was significantly shorter than that in the low-risk group in the training (median PFS, 9.5 m and 19.0 m, respectively; p [log-rank] < 0.0001) and validation (median PFS, 11.3 m and 22.5 m, respectively; p [log-rank] = 0.0063) cohorts. CONCLUSIONS: A radiomics-based model was established to predict the risk of progression in hypopharyngeal cancer with chemoradiotherapy. KEY POINTS: • Clinical information showed limited performance in stratifying the risk of progression among patients with hypopharyngeal cancer. • Imaging features extracted from CECT and NCCT images were independent predictors of PFS. • We combined significant features and valuable clinical variables to establish a nomogram to predict individual risk of progression.

Preoperative prediction of sentinel lymph node metastasis in breast cancer based on radiomics of T2-weighted fat-suppression and diffusion-weighted MRI.

OBJECTIVES: To predict sentinel lymph node (SLN) metastasis in breast cancer patients using radiomics based on T2-weighted fat suppression (T2-FS) and diffusion-weighted MRI (DWI). METHODS: We enrolled 146 patients with histologically proven breast cancer. All underwent pretreatment T2-FS and DWI MRI scan. In all, 10,962 texture and four non-texture features were extracted for each patient. The 0.623 + bootstrap method and the area under the curve (AUC) were used to select the features. We constructed ten logistic regression models (orders of 1-10) based on different combination of image features using stepwise forward method. RESULTS: For T2-FS, model 10 with ten features yielded the highest AUC of 0.847 in the training set and 0.770 in the validation set. For DWI, model 8 with eight features reached the highest AUC of 0.847 in the training set and 0.787 in the validation set. For joint T2-FS and DWI, model 10 with ten features yielded an AUC of 0.863 in the training set and 0.805 in the validation set. CONCLUSIONS: Full utilisation of breast cancer-specific textural features extracted from anatomical and functional MRI images improves the performance of radiomics in predicting SLN metastasis, providing a non-invasive approach in clinical practice. KEY POINTS: • SLN biopsy to access breast cancer metastasis has multiple complications. • Radiomics uses features extracted from medical images to characterise intratumour heterogeneity. • We combined T 2 -FS and DWI textural features to predict SLN metastasis non-invasively.

A modified endoscopic method for resection of gastric submucosal tumor.

BACKGROUND AND OBJECTIVE: Endoscopic mucosa-sparing lateral dissection (EMSLD) was developed by our group, and is used to remove gastric submucosal tumor (SMT). This study aims to evaluate the feasibility and safety of this method. METHODS: This retrospective study included 25 patients who underwent EMSLDs at an endoscopy center as a national key unit in china from October 2015 to July 2016. The main data collected were the size of the gastric SMT, its location and origin, en bloc resection rate, operating time, intraoperative and postoperative complications, hospitalization expense, hospital days, and follow-up after hospital discharge. RESULTS: The mean (SD) size of the gastric SMTs was 18.3 (5.9) mm; 96% (24/25) of the tumors originated in the muscularis propria; and 64% (16/25) and 28% (7/25) were located in the gastric fundus and gastric body, respectively. The rate of en bloc resection was 96% (24/25), and the rate of intraoperative perforations due to endoscopic full-thickness resection was 48% (12/25). All wounds and perforations were effectively closed using endoscopic clips combined with the retained mucosa. The mean operative time was 74.2 (38.0) min. Delayed bleeding and perforation were not observed. CONCLUSION: Endoscopic mucosa-sparing lateral dissection is safe and feasible for the removal of gastric SMTs. The wound can be effectively closed using the retained mucosa and endoscopic clips, even if perforation has occurred. EMSLD provides an alternative to the resection of gastric SMTs, especially for tumors with a risk of intraoperative perforation.

Ultrasonic-Assisted Dispersion of ZnO Nanoparticles and Its Inhibition Activity to Trichoderma viride.

The activity inhibition of fungi by ZnO nanoparticles (NPs) has shown huge potential applications in the area of hygienic coatings. However, the inhibition efficiency was limited due to the agglomeration of NPs. To obtain well-dispersed and highly stabilized ZnO nanofluids, ZnO NPs were capped with four kinds of surfactants under ultrasonication. The capping procedure was optimized by varying the dosage of surfactants, the ultrasonic duration, ultrasonic power and temperature. Capped ZnO nanofluids were then used for the inhibition of Trichoderma viride. The influence on the activity of the capping conditions, illumination, ZnO NPs content, humidity and temperature were investigated in details. Results suggest that well-dispersed ZnO NPs were obtained through ultrasonic-assisted functionalization using sodium polyacrylate as a dispersant. Moreover, capped ZnO nanofluids revealed long-term stability at pH above 6. The optimal capping procedure was obtained for a sonication power of 250 W, treatment duration of 40 min, dosage of 0.4% and temperature of 60 °C. Antifungal tests indicated that capped ZnO NPs showed an inhibition ability versus T. viride even in the dark. The antifungal ability of ZnO NPs increased with the increasing ZnO content, and humidity and temperature only affected the growth of fungi. Capped ZnO NPs showed an excellent antifungal performance even in the circumstance that was beneficial for the fungi growth (temperature of 30 °C, humidity of 95%), demonstrating the antimicrobial capability in practical applications.

Extrinsic warming of low-osmolality iodinated contrast media to 37°C reduced the rate of allergic-like reaction.

Background: Although there is good evidence that warming of contrast media changes the bolus kinetics and injection pressure of iodinated contrast media, there has been little evidence that it affects clinical adverse event rates in a meaningful way. Objective: To determine whether the extrinsic warming of low-osmolality iodinated contrast media to 37°C reduced adverse reactions. Methods: Data on adverse reactions were collected from two cohorts, one of which used contrast media at room temperature and the other in which contrast media were warmed to 37°C before administration. Adverse reactions, including allergic-like and physiological reactions, were reviewed. We compared the incidence rates of adverse reactions between the two cohorts by using the χ2 test. Results: A total of 70,446 injections in cohort 1 and 203,873 injections in cohort 2 were included. Extrinsic warming reduced the rate of allergic-like reactions to iopromide 370, iopamidol 370, and iohexol 350 (0.32% in cohort 1 versus 0.21% in cohort 2, p = 0.003; 0.14% versus 0.10%, p = 0.046; and 0.32% versus 0.13%, p = .003, respectively). However, the physiological reaction rates could not be reduced (p = 0.057, p = 0.107, and p = 0.962, respectively). The extrinsic warming of iopromide 300 could not reduce adverse reaction rates (allergic-like reaction rates: 0.21% versus 0.16%, p = 0.407; physiological reaction rates: 0.17% versus 0.13%, p = 0.504). Conclusion: Extrinsic warming to 37°C before intravenous administration was associated with a reduction in the rate of allergic-like reactions to iopromide 370, iopamidol 370, and iohexol 350.

Neuro-Magnetic Resonance Imaging in Hand, Foot, and Mouth Disease: Finding in 412 Patients and Prognostic Features.

PURPOSE: The aims of this study were to describe the neuroimaging findings in hand, foot, and mouth disease and determine those who may provide prognosis. MATERIAL AND METHODS: Magnetic resonance imaging scans in 412 severe hand, foot, and mouth disease between 2009 and 2014 were retrospectively evaluated. The patients who had the neurological signs were followed for 6 months to 1 year. According to the good or poor prognosis, 2 groups were categorized. The incidence of lesions in different sites between the 2 groups was compared, and multivariate analysis was used to look for risk factors. RESULTS: The major sites of involvement for all patients with percentages were the medulla oblongata (16.1%), spinal anterior nerve roots (12.4%), thoracic segments (11.1%), brain or spinal meninges (8.3%), and so on. There were 347 patients (84.2%) with good prognosis and 65 (15.8%) with poor prognosis in the follow-up. There was a significantly higher rate of lesions involving the cerebral white substance, thalamus, medulla oblongata, pons, midbrain, and spinal cord in the group with poor prognosis. Multivariate analysis showed 2 independent risk factors associated with poor prognosis: lesions located in the medulla oblongata (P < 0.015) and spinal cord (P < 0.001) on magnetic resonance imaging; the latter was the most significant prognostic factor (odds ratio, 29.11; P < 0.001). CONCLUSIONS: We found that the distribution patterns for all patients mainly involved the medulla oblongata, spinal anterior nerve roots, thoracic segments, and brain or spinal meninges. Our findings suggested that patients with lesions located in the medulla oblongata and spinal cord may be closely monitored for early intervention and meticulous management. For children with the symptom of nervous system, they are strongly recommended for magnetic resonance examination.

Machine learning-based multimodal MRI texture analysis for assessing renal function and fibrosis in diabetic nephropathy: a retrospective study.

Introduction: Diabetic nephropathy (DN) has become a major public health burden in China. A more stable method is needed to reflect the different stages of renal function impairment. We aimed to determine the possible practicability of machine learning (ML)-based multimodal MRI texture analysis (mMRI-TA) for assessing renal function in DN. Methods: For this retrospective study, 70 patients (between 1 January 2013 and 1 January 2020) were included and randomly assigned to the training cohort (n1 = 49) and the testing cohort (n2 = 21). According to the estimated glomerular filtration rate (eGFR), patients were assigned into the normal renal function (normal-RF) group, the non-severe renal function impairment (non-sRI) group, and the severe renal function impairment (sRI) group. Based on the largest coronal image of T2WI, the speeded up robust features (SURF) algorithm was used for texture feature extraction. Analysis of variance (ANOVA) and relief and recursive feature elimination (RFE) were applied to select the important features and then support vector machine (SVM), logistic regression (LR), and random forest (RF) algorithms were used for the model construction. The values of area under the curve (AUC) on receiver operating characteristic (ROC) curve analysis were used to assess their performance. The robust T2WI model was selected to construct a multimodal MRI model by combining the measured BOLD (blood oxygenation level-dependent) and diffusion-weighted imaging (DWI) values. Results: The mMRI-TA model achieved robust and excellent performance in classifying the sRI group, non-sRI group, and normal-RF group, with an AUC of 0.978 (95% confidence interval [CI]: 0.963, 0.993), 0.852 (95% CI: 0.798, 0.902), and 0.972 (95% CI: 0.995, 1.000), respectively, in the training cohort and 0.961 (95% CI: 0.853, 1.000), 0.809 (95% CI: 0.600, 0.980), and 0.850 (95% CI: 0.638, 0.988), respectively, in the testing cohort. Discussion: The model built from multimodal MRI on DN outperformed other models in assessing renal function and fibrosis. Compared to the single T2WI sequence, mMRI-TA can improve the performance in assessing renal function.

Diffusion-weighted imaging in the diagnosis of enterovirus 71 encephalitis.

BACKGROUND: In the early phase of viral encephalitis, conventional MRI may appear normal. Diffusion-weighted imaging (DWI) is a sensitive tool for detecting early changes in cellular function in the central nervous system. PURPOSE: To investigate the usefulness of DWI in the diagnosis of enterovirus 71 (EV71) encephalitis, and to determine whether DWI is superior to conventional MR sequences. MATERIAL AND METHODS: MRI scans in 26 patients were retrospectively evaluated for distribution of lesions on T1-weighted images (T1WI), T2-weighted images (T2WI), fluid-attenuated inversion recovery (FLAIR), and DWI. Contrast-to-noise ratios (CNRs) were calculated for all regions on each sequence and differences in the four MRI sequences were assessed using CNRs. Apparent diffusion coefficient (ADC) values were measured for all regions to look for true restriction of diffusion. RESULTS: Fifteen out of 26 cases showed positive findings on MR imaging. The brain stem was involved in 11 patients, cortex and subcortical white matter in four patients. DWI was more sensitive in detecting the abnormalities (89.7%) compared to T2WI (48.7%), FLAIR (41.0%), and T1WI (35.9%), and the positive ratio of DWI was significantly higher compared to other sequences. Furthermore, no significant difference was found between T2WI and FLAIR (P = 0.649). The corresponding mean CNRs were 8.73 ± 2.57, 83.59 ± 29.28, 24.22 ± 6.22, and 132.27 ± 78.32 on T1WI, T2WI, FLAIR, and DWI, respectively. The absolute values of CNRs of lesions on DWI were significantly greater than those on other sequences. CONCLUSION: DWI appears to be more sensitive in detecting EV71 encephalitis than conventional MRI sequences. This capability may improve the accuracy in diagnosing EV71 encephalitis, especially at the early stage.

Analysis of alarming signals for the progression of atrophic gastritis to dysplasia.

BACKGROUND AND AIMS: atrophic gastritis (AG) and dysplasia are precancerous lesions, with which early surveillance of disease progression is particularly important for patients. The present study aimed to explore potential alarming signals for atrophic gastritis progression towards dysplasia. METHODS: clinical data for patients with AG in the Digestive Endoscopy Center of Guangzhou Nanfang Hospital between 2001 and 2011 were retrospectively reviewed. Survival analysis, dichotomous logistic regression analysis and rank correlation analysis were carried out. RESULTS: in 234 patients with atrophic gastritis, after follow up of 0.5, 1, 2, 5 and 10 years, the occurrence rates of dysplasia were respectively 2.3, 4.4, 9.6, 19.3, and 42.4%. Patients with AG combined with antral ulcer or gastric angle ulcer, had a higher risk for dysplasia than patients with simple AG (OR = 2.427, 95%Cl 1.069 ~ 5.511, p = 0.034; OR = 2.961, 95%Cl 1.336 ~ 6.564, p = 0.008). The constituent ratios of moderate to severe dysplasia were respectively 8.6, 2.7 and 1.2% (p = 0.000) in three patient groups: AG combined with antral ulcer/gastric angle ulcer (n = 255), antral ulcer/gastric angle ulcer alone (n = 1,389), and AG alone (n = 3,106). The Spearman correlation coefficients between a) Hp infection; and b) atrophy, intestinal metaplasia and dysplasia were -0.114 (p = 0.078), -0.169 (p = 0.009) and 0.064 (p > 0.05), respectively. CONCLUSIONS: long follow-up interval and gastric ulcer may be alarming signals for the progression of AG to dysplasia. But in patients with atrophy, intestinal metaplasia or dysplasia, Hp infection may not be a risk factor for these lesions aggravated, further studies are required to confirm it.

Preparation of PP-g-(AA-MAH) Fibers Using Suspension Grafting and Melt-Blown Spinning and its Adsorption for Aniline.

This paper uses polypropylene (PP) as the matrix and acrylic acid (AA) and maleic anhydride (MAH) as functional monomers to prepare PP-g-(AA-MAH) fibers by suspension grafting and melt-blown spinning technology that are easy to industrially scale-up. The fibers can be used to adsorb aniline. Results showed that the grafting ratio reached the maximum of 12.47%. The corresponding optimal conditions were grafting time of 3 h, AA: MAH = 0.75, total monomer content of 55%, benzoyl peroxide 1.4%, xylene concentration of 6 mL/g PP, and deionized water content of 8 mL/g PP. Owing to its good fluidity and thermal stability, the product of suspension grafting can be used for melt-blown spinning. Infrared spectroscopic and nuclear magnetic resonance spectroscopic analyses indicated that AA and MAH were successfully grafted onto PP fibers. After grafting, the hydrophilicity of PP-g-(AA-MAH) fiber increased. Therefore, it had higher absorptivity for aniline and the adsorption capacity could reach 42.2 mg/g at 45 min and pH = 7. Moreover, the PP-g-(AA-MAH) fibers showed good regeneration performance.

Combined PET/CT with thoracic contrast-enhanced CT in assessment of primary cardiac tumors in adult patients.

BACKGROUND: 18F-FDG PET/CT is a key molecular imaging modality to noninvasively assess and differentiate benign and malignant cardiac tumors. However, few benign cardiac tumors can be characterized by increased 18F-FDG uptake, which makes differential diagnosis difficult. This study sought to retrospectively evaluate whether combined 18F-FDG PET/CT with thoracic contrast-enhanced CT (CECT) helps in assessing primary cardiac tumors in adult patients, compared with CECT or PET/CT alone. METHODS: Forty-six consecutive patients who were diagnosed as primary cardiac tumors were enrolled. All patients underwent 18F-FDG PET/CT followed by thoracic CECT before biopsy or surgery. Visual qualitative interpretation and quantitative analysis were performed, and diagnostic performance was evaluated. RESULTS: More than half (16/29) of benign tumors exhibited with mild 18F-FDG uptake. There were significant differences in 18F-FDG uptake and the degree of absolute enhancement between benign and malignant tumors (P < 0.001). The combination of two modalities improved the specificity from 79 to 93%, the positive predictive value from 73 to 89%, and the accuracy of diagnosis from 85 to 93%. There were significant differences between PET/CT alone or thoracic CECT alone and combined modalities (P = 0.034 and P = 0.026, respectively). The combination with the optimal SUVmax cutoff value generated 94% sensitivity, 100% specificity, 97% negative predictive values, 100% positive predictive values, and 98% accuracy rates. CONCLUSIONS: Combining 18F-FDG PET/C with thoracic CECT significantly improved specificity and accuracy compared to CECT or PET/CT alone in detecting tumors. This combination of diagnostic imaging is effective in differentiating malignant from benign masses.

Ion-Imprinted Polypropylene Fibers Fabricated by the Plasma-Mediated Grafting Strategy for Efficient and Selective Adsorption of Cr(VI).

To improve the adsorption selectivity towards hexavalent chromium anion (Cr(VI)), surface Cr(VI)-imprinted polypropylene (PP) fibers were fabricated by the plasma-mediated grafting strategy. Hence, a non-thermal Rradio frequency discharge plasma irradiation followed by a gaseous phase grafting was used to load acrylic acid (AA) onto PP fibers, which was afterwards amidated with triethylenetetramine and subjected to imprinting with a Cr(VI) template. The plasma irradiation conditions, i.e., gas species, output power, pressure, and time, were optimized and then the influence of grafting time, pressure, and temperature on the grafting degree of AA was investigated. Scanning electron microscopy and Fourier transform infrared spectroscopy were used for the characterization of pristine and modified fibers and to confirm the synthesis success. The hydrophilicity of modified fibers was greatly improved compared with pristine PP fibers. The adsorption thermodynamics and kinetics of Cr(VI) were investigated, as well as the elution efficiency and reusability. The prepared imprinted fibers showed superior adsorption selectivity to Cr(VI) compared with non-imprinted fibers. Finally, the stability of the imprinted fibers against the oxidation ability of Cr(VI) is discussed.

Development and validation of a magnetic resonance imaging-based model for the prediction of distant metastasis before initial treatment of nasopharyngeal carcinoma: A retrospective cohort study.

BACKGROUND: We aimed to identify a magnetic resonance imaging (MRI)-based model for assessment of the risk of individual distant metastasis (DM) before initial treatment of nasopharyngeal carcinoma (NPC). METHODS: This retrospective cohort analysis included 176 patients with NPC. Using the PyRadiomics platform, we extracted the imaging features of primary tumors in all patients who did not exhibit DM before treatment. Subsequently, we used minimum redundancy-maximum relevance and least absolute shrinkage and selection operator algorithms to select the strongest features and build a logistic model for DM prediction. The independent statistical significance of multiple clinical variables was tested using multivariate logistic regression analysis. FINDINGS: In total, 2780 radiomic features were extracted. A DM MRI-based model (DMMM) comprising seven features was constructed for the classification of patients into high- and low-risk groups in a training cohort and validated in an independent cohort. Overall survival was significantly shorter in the high-risk group than in the low-risk group (P < 0·001). A radiomics nomogram based on radiomic features and clinical variables was developed for DM risk assessment in each patient, and it showed a significant predictive ability in the training [area under the curve (AUC), 0·827; 95% confidence interval (CI), 0.754-0.900] and validation (AUC, 0.792; 95% CI, 0.633-0.952) cohorts. INTERPRETATION: DMMM can serve as a visual prognostic tool for DM prediction in NPC, and it can improve treatment decisions by aiding in the differentiation of patients with high and low risks of DM. FUND: This research received financial support from the National Natural Science Foundation of China (81571664, 81871323, 81801665, 81771924, 81501616, 81671851, and 81527805); the National Natural Science Foundation of Guangdong Province (2018B030311024); the Science and Technology Planning Project of Guangdong Province (2016A020216020); the Scientific Research General Project of Guangzhou Science Technology and Innovation Commission (201707010328); the China Postdoctoral Science Foundation (2016M600145); and the National Key R&D Program of China (2017YFA0205200, 2017YFC1308700, and 2017YFC1309100).

Radiomic Nomogram: Pretreatment Evaluation of Local Recurrence in Nasopharyngeal Carcinoma based on MR Imaging.

Background: To develop and validate a radiomic nomogram incorporating radiomic features with clinical variables for individual local recurrence risk assessment in nasopharyngeal carcinoma (NPC) patients before initial treatment. Methods: One hundred and forty patients were randomly divided into a training cohort (n = 80) and a validation cohort (n = 60). A total of 970 radiomic features were extracted from pretreatment magnetic resonance (MR) images of NPC patients from May 2007 to December 2013. Univariate and multivariate analyses were used for selecting radiomic features associated with local recurrence, and multivariate analyses was used for building radiomic nomogram. Results: Eight contrast-enhanced T1-weighted (CET1-w) image features and seven T2-weighted (T2-w) image features were selected to build a Cox proportional hazard model in the training cohort, respectively. The radiomic nomogram, which combined radiomic features and multiple clinical variables, had a good evaluation ability (C-index: 0.74 [95% CI: 0.58, 0.85]) in the validation cohort. The radiomic nomogram successfully categorized those patients into low- and high-risk groups with significant differences in the rate of local recurrence-free survival (P <0.05). Conclusions: This study demonstrates that MR imaging-based radiomics can be used as an aid tool for the evaluation of local recurrence, in order to develop tailored treatment targeting specific characteristics of individual patients.