Radiomics using non-contrast CT to predict hemorrhagic transformation risk in stroke patients undergoing revascularization

OBJECTIVES

This study explores whether textural features from initial non-contrast CT scans of infarcted brain tissue are linked to hemorrhagic transformation susceptibility.

MATERIALS AND METHODS

Stroke patients undergoing thrombolysis or thrombectomy from Jan 2012 to Jan 2022 were analyzed retrospectively. Hemorrhagic transformation was defined using follow-up magnetic resonance imaging. A total of 94 radiomic features were extracted from the infarcted tissue on initial NCCT scans. Patients were divided into training and test sets (7:3 ratio). Two models were developed with fivefold cross-validation: one incorporating first-order and textural radiomic features, and another using only textural radiomic features. A clinical model was also constructed using logistic regression with clinical variables, and test set validation was performed.

RESULTS

Among 362 patients, 218 had hemorrhagic transformations. The LightGBM model with all radiomics features had the best performance, with an area under the receiver operating characteristic curve (AUROC) of 0.986 (95% confidence interval [CI], 0.971-1.000) on the test dataset. The ExtraTrees model performed best when textural features were employed, with an AUROC of 0.845 (95% CI, 0.774-0.916). Minimum, maximum, and ten percentile values were significant predictors of hemorrhagic transformation. The clinical model showed an AUROC of 0.544 (95% CI, 0.431-0.658). The performance of the radiomics models was significantly better than that of the clinical model on the test dataset (p < 0.001).

CONCLUSIONS

The radiomics model can predict hemorrhagic transformation using NCCT in stroke patients. Low Hounsfield unit was a strong predictor of hemorrhagic transformation, while textural features alone can predict hemorrhagic transformation.

CLINICAL RELEVANCE STATEMENT

Using radiomic features extracted from initial non-contrast computed tomography, early prediction of hemorrhagic transformation has the potential to improve patient care and outcomes by aiding in personalized treatment decision-making and early identification of at-risk patients.

KEY POINTS

• Predicting hemorrhagic transformation following thrombolysis in stroke is challenging since multiple factors are associated. • Radiomics features of infarcted tissue on initial non-contrast CT are associated with hemorrhagic transformation. • Textural features on non-contrast CT are associated with the frailty of the infarcted tissue.

Kondo screening in a Majorana metal

Kondo impurities provide a nontrivial probe to unravel the character of the excitations of a quantum spin liquid. In the S = 1/2 Kitaev model on the honeycomb lattice, Kondo impurities embedded in the spin-liquid host can be screened by itinerant Majorana fermions via gauge-flux binding. Here, we report experimental signatures of metallic-like Kondo screening at intermediate temperatures in the Kitaev honeycomb material α-RuCl3 with dilute Cr3+ (S = 3/2) impurities. The static magnetic susceptibility, the muon Knight shift, and the muon spin-relaxation rate all feature logarithmic divergences, a hallmark of a metallic Kondo effect. Concurrently, the linear coefficient of the magnetic specific heat is large in the same temperature regime, indicating the presence of a host Majorana metal. This observation opens new avenues for exploring uncharted Kondo physics in insulating quantum magnets.

Natural language processing to predict isocitrate dehydrogenase genotype in diffuse glioma using MR radiology reports

OBJECTIVES

To evaluate the performance of natural language processing (NLP) models to predict isocitrate dehydrogenase (IDH) mutation status in diffuse glioma using routine MR radiology reports.

MATERIALS AND METHODS

This retrospective, multi-center study included consecutive patients with diffuse glioma with known IDH mutation status from May 2009 to November 2021 whose initial MR radiology report was available prior to pathologic diagnosis. Five NLP models (long short-term memory [LSTM], bidirectional LSTM, bidirectional encoder representations from transformers [BERT], BERT graph convolutional network [GCN], BioBERT) were trained, and area under the receiver operating characteristic curve (AUC) was assessed to validate prediction of IDH mutation status in the internal and external validation sets. The performance of the best performing NLP model was compared with that of the human readers.

RESULTS

A total of 1427 patients (mean age ± standard deviation, 54 ± 15; 779 men, 54.6%) with 720 patients in the training set, 180 patients in the internal validation set, and 527 patients in the external validation set were included. In the external validation set, BERT GCN showed the highest performance (AUC 0.85, 95% CI 0.81-0.89) in predicting IDH mutation status, which was higher than LSTM (AUC 0.77, 95% CI 0.72-0.81; p = .003) and BioBERT (AUC 0.81, 95% CI 0.76-0.85; p = .03). This was higher than that of a neuroradiologist (AUC 0.80, 95% CI 0.76-0.84; p = .005) and a neurosurgeon (AUC 0.79, 95% CI 0.76-0.84; p = .04).

CONCLUSION

BERT GCN was externally validated to predict IDH mutation status in patients with diffuse glioma using routine MR radiology reports with superior or at least comparable performance to human reader.

CLINICAL RELEVANCE STATEMENT

Natural language processing may be used to extract relevant information from routine radiology reports to predict cancer genotype and provide prognostic information that may aid in guiding treatment strategy and enabling personalized medicine.

KEY POINTS

• A transformer-based natural language processing (NLP) model predicted isocitrate dehydrogenase mutation status in diffuse glioma with an AUC of 0.85 in the external validation set. • The best NLP models were superior or at least comparable to human readers in both internal and external validation sets. • Transformer-based models showed higher performance than conventional NLP model such as long short-term memory.

Association of partial T2-FLAIR mismatch sign and isocitrate dehydrogenase mutation in WHO grade 4 gliomas: results from the ReSPOND consortium

PURPOSE

While the T2-FLAIR mismatch sign is highly specific for isocitrate dehydrogenase (IDH)-mutant, 1p/19q-noncodeleted astrocytomas among lower-grade gliomas, its utility in WHO grade 4 gliomas is not well-studied. We derived the partial T2-FLAIR mismatch sign as an imaging biomarker for IDH mutation in WHO grade 4 gliomas.

METHODS

Preoperative MRI scans of adult WHO grade 4 glioma patients (n = 2165) from the multi-institutional ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium were analyzed. Diagnostic performance of the partial T2-FLAIR mismatch sign was evaluated. Subset analyses were performed to assess associations of imaging markers with overall survival (OS).

RESULTS

One hundred twenty-one (5.6%) of 2165 grade 4 gliomas were IDH-mutant. Partial T2-FLAIR mismatch was present in 40 (1.8%) cases, 32 of which were IDH-mutant, yielding 26.4% sensitivity, 99.6% specificity, 80.0% positive predictive value, and 95.8% negative predictive value. Multivariate logistic regression demonstrated IDH mutation was significantly associated with partial T2-FLAIR mismatch (odds ratio [OR] 5.715, 95% CI [1.896, 17.221], p = 0.002), younger age (OR 0.911 [0.895, 0.927], p < 0.001), tumor centered in frontal lobe (OR 3.842, [2.361, 6.251], p < 0.001), absence of multicentricity (OR 0.173, [0.049, 0.612], p = 0.007), and presence of cystic (OR 6.596, [3.023, 14.391], p < 0.001) or non-enhancing solid components (OR 6.069, [3.371, 10.928], p < 0.001). Multivariate Cox analysis demonstrated cystic components (p = 0.024) and non-enhancing solid components (p = 0.003) were associated with longer OS, while older age (p < 0.001), frontal lobe center (p = 0.008), multifocality (p < 0.001), and multicentricity (p < 0.001) were associated with shorter OS.

CONCLUSION

Partial T2-FLAIR mismatch sign is highly specific for IDH mutation in WHO grade 4 gliomas.

National prevalence and determinants of COVID-19 vaccine hesitancy during the initial phase pandemic

OBJECTIVE

Although previous studies have explored the causes of COVID-19 vaccine hesitancy during the pandemic, there is a lack of generality and reproducibility in these studies. Therefore, we aimed to comprehensively identify the determinants of COVID-19 vaccine hesitancy through a representative nationwide cross-sectional study conducted in South Korea.

SUBJECTS AND METHODS

We used a nationwide, representative, and large-scale dataset from the 2021 Community Health Survey. By analyzing 193,495 participants, we investigated the nationwide incidence of COVID-19 vaccine hesitancy and the various causes thereof.

RESULTS

The national prevalence of COVID-19 vaccine hesitancy was 5.7% (95% CI, 5.5-5.8). COVID-19 vaccine hesitancy was associated with an increased incidence of the following factors: (1) demographic factors including early-middle adulthood [vs. late; odds ratio (OR), 1.51; 95% CI, 1.38-1.65] and male sex (vs. female sex; OR, 1.08; 95% CI, 1.01-1.14); (2) physically healthy subjects; (3) lower socio-economic status (vs. high household income; OR, 1.28; 95% CI, 1.19-1.38); (4) having mental illness (vs. normal mental status; OR, 1.25; 95% CI, 1.13-1.38); and (5) unhealthy habits such as current smoking (vs. non-smoking; OR, 1.22; 95% CI, 1.13-1.31); and insufficient physical activity (vs. sufficient; OR, 1.08; 95% CI, 1.01-1.17). Common reasons for vaccine hesitancy were concerns about side effects (41.34%), health problems (24.60%), and inability to select the type of vaccine (14.13%).

CONCLUSIONS

This representative large-scale nationwide study conducted in South Korea investigated the nationwide prevalence and determinants of vaccine hesitancy. Our results provide useful public health information, especially on novel aspects of vaccination strategies, for policymakers to improve the acceptance of COVID-19 vaccines.

Risk factors for musculoskeletal-related occupational disability among US Army soldiers

INTRODUCTION

Minimising temporary and permanent disability associated with musculoskeletal conditions (MSK-D) is critical to the mission of the US Army. Prior research has identified potentially actionable risk factors for overall military disability and its MSK-D subset, including elevated body mass index, tobacco use and physical fitness. However, prior work does not appear to have addressed the impact of these factors on MSK-D when controlling for a full range of factors that may affect health behaviours, including aptitude scores that may serve as a proxy for health literacy. Identifying risk factors for MSK-D when providing control for all such factors may inform efforts to improve military readiness.

METHODS

We studied 494 757 enlisted Army soldiers from 2014 to 2017 using a combined medical and administrative database. Leveraging data from the Army's digital 'eProfile' system of duty restriction records, we defined MSK-D as the first restriction associated with musculoskeletal conditions and resulting in the inability to deploy or train. We used multivariable Cox proportional hazards regression to assess the associations between incident MSK-D and selected risk factors including aptitude scores, physical fitness test scores, body mass index and tobacco use.

RESULTS

Among the subjects, 281 278 (45.14%) experienced MSK-D. In the MSK-D hazards model, the highest effect size was for failing the physical fitness test (adjusted HR=1.63, 95% CI 1.58 to 1.67, p<0.001) compared with scoring ≥290 points.

CONCLUSIONS

The analysis revealed the strongest associations between physical fitness and MSK-D. Additional efforts are warranted to determine potential mechanisms for the observed associations between selected factors and MSK-D.

17-year trends of body mass index, overweight, and obesity among adolescents from 2005 to 2021, including the COVID-19 pandemic: a Korean national representative study

OBJECTIVE

There is a lack of pediatric studies that have analyzed trends in mean body mass index (BMI) and the prevalence of obesity and overweight over a period that includes the mid-stage of the COVID-19 pandemic. Thus, we aimed to investigate trends in BMI, overweight, and obesity among Korean adolescents from 2005 to 2021, including the COVID-19 pandemic.

SUBJECTS AND METHODS

We used data from the Korea Youth Risk Behavior Web-based Survey (KYRBS), which is nationally representative of South Korea. The study included middle- and high-school students between the ages of 12 and 18. We examined trends in mean BMI and prevalence of obesity and/or overweight during the COVID-19 pandemic and compared these to those of pre-pandemic trends in each subgroup by gender, grade, and residential region.

RESULTS

Data from 1,111,300 adolescents (mean age: 15.04 years) were analyzed. The estimated weighted mean BMI was 20.48 kg/m2 (95% CI, 20.46-20.51) between 2005 and 2007, and this was 21.61 kg/m2 (95% CI, 21.54-21.68) in 2021. The prevalence of overweight and obesity was 13.1% (95% CI, 12.9-13.3%) between 2005 and 2007 and 23.4% (95% CI, 22.8-24.0%) in 2021. The mean BMI and prevalence of obesity and overweight have gradually increased over the past 17 years; however, the extent of change in mean BMI and in the prevalence of obesity and overweight during the pandemic was distinctly less than before. The 17-year trends in the mean BMI, obesity, and overweight exhibited a considerable rise from 2005 to 2021; however, the slope during the COVID-19 pandemic (2020-2021) was significantly less prominent than in the pre-pandemic (2005-2019).

CONCLUSIONS

These findings enable us to comprehend long-term trends in the mean BMI of Korean adolescents and further emphasize the need for practical prevention measures against youth obesity and overweight.

Federated learning enables big data for rare cancer boundary detection

Although machine learning (ML) has shown promise across disciplines, out-of-sample generalizability is concerning. This is currently addressed by sharing multi-site data, but such centralization is challenging/infeasible to scale due to various limitations. Federated ML (FL) provides an alternative paradigm for accurate and generalizable ML, by only sharing numerical model updates. Here we present the largest FL study to-date, involving data from 71 sites across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, reporting the largest such dataset in the literature (n = 6, 314). We demonstrate a 33% delineation improvement for the surgically targetable tumor, and 23% for the complete tumor extent, over a publicly trained model. We anticipate our study to: 1) enable more healthcare studies informed by large diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further analyses for glioblastoma by releasing our consensus model, and 3) demonstrate the FL effectiveness at such scale and task-complexity as a paradigm shift for multi-site collaborations, alleviating the need for data-sharing.

NIMG-67. MULTI-PARAMETRIC MRI-BASED MACHINE LEARNING ANALYSIS FOR PREDICTION OF NEOPLASTIC INFILTRATION AND RECURRENCE IN PATIENTS WITH GLIOBLASTOMA: UPDATES FROM THE MULTI-INSTITUTIONAL RESPOND CONSORTIUM

PURPOSE

Glioblastoma is extremely infiltrative with malignant cells extending beyond the enhancing rim where recurrence inevitably occurs, despite aggressive multimodal therapy. We hypothesize that important characteristics of peritumoral tissue heterogeneity captured and analyzed by multi-parametric MRI and artificial intelligence (AI) methods are generalizable in the updated multi-institutional ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium and predictive of neoplastic infiltration and future recurrence.

METHODS

We used the most recent update of the ReSPOND consortium to evaluate and further refine generalizability of our methods with different scanners and acquisition settings. 179 de novo glioblastoma patients with available T1, T1Gd, T2, T2-FLAIR, and ADC sequences at pre-resection baseline and after complete resection with subsequent pathology-confirmed recurrence were included. To establish generalizability of the predictive models, training and testing of the refined AI model was performed through Leave-One-Institution-Out-Cross-Validation schema. The multi-institutional cohort consisted of the Hospital of the University of Pennsylvania (UPenn, 124), Case Western Reserve University/University Hospitals (CWRU/UH, 27), New York University (NYU, 13), Ohio State University (OSU, 13), and University Hospital Río Hortega (RH, 2). Features extracted from pre-resection MRI were used to build the model predicting the spatial pattern of subsequent tumor recurrence. These predictions were evaluated against regions of pathology-confirmed post-resection recurrence.

RESULTS

Our model predicted the locations that later harbored tumor recurrence with overall odds ratio (99% CI)/AUC (99% CI), 12.0(11.8-12.2)/0.80(0.76-0.85), and per institute, CWRU/UH, 11.0(10.7-11.3)/0.80 (0.64-0.97); NYU, 7.0(6.7-7.3)/0.78(0.56-1.00); OSU, 18.3(17.5-19.1)/0.83(0.54-1.00); RH, 40.0(35.3-45.5)/0.93(0.00-1.00); UPenn, 8.00(7.7-8.3)/0.80(0.75-0.84).

CONCLUSION

This study provides extensive multi-institutional validated evidence that machine learning tools can identify peritumoral neoplastic infiltration and predict location of future recurrence, by decrypting the MRI signal heterogeneity in peritumoral tissue. Our analyses leveraged the unique dataset of the ReSPOND consortium, which aims to develop and validate AI-based biomarkers for individualized prediction and prognostication and establish generalizability in a multi-institutional setting.

NIMG-29. ASSOCIATION OF PARTIAL T2-FLAIR MISMATCH SIGN AND ISOCITRATE DEHYDROGENASE MUTATION IN WHO GRADE 4 GLIOMA/GLIOBLASTOMA: RESULTS FROM THE RESPOND CONSORTIUM

PURPOSE

T2-FLAIR mismatch (T2FM) is a highly specific imaging biomarker for isocitrate dehydrogenase (IDH) mutation in low-grade gliomas. Previous T2FM studies are inconsistent for glioblastoma (GBM)/grade-4 glioma, partly due to low IDH-mutation prevalence in high-grade gliomas. We leveraged a large multi-institutional GBM/grade-4 glioma cohort to analyze the association of partial T2FM and IDH-mutation (T2-hyperintense, FLAIR-hypointense, nonenhancing, nonedema).

METHODS

We analyzed preoperative MRI of 1500 pathologically confirmed GBM/grade-4 gliomas with known IDH-mutation status from the ReSPOND consortium, consisting of the following institutions (sample size): Ivy GBM Atlas Project (33), Catalan Institute of Oncology (132), Case Western Reserve University/University Hospitals (132), New York University (55), Ohio State University (25), University of Pennsylvania (641), University Hospital Río Hortega (16), Yonsei University Health System (118), The Cancer Imaging Archive (93), Thomas Jefferson University (48), Tata Memorial Hospital (22), University of Pittsburgh Medical Center (156), and Washington University School of Medicine in St. Louis (57). Sequences were co-registered to a common anatomic atlas. Continuous variables were compared by t-test and categorical variables by Χ 2-test.

RESULTS

71 (4.7%) were IDH-mutants, significantly younger (43±1 v. 62±12 years, p=5x10-37), and more likely to exhibit partial T2FM (20% v. 0.4%, p=1x10-43), frontal lobe predominance (68% v. 29%, p=7x10-12), nonenhancing components (T2/FLAIR-intermediate signal, nonedema; 45% v. 9%, p=1x10-22), and cystic components (smooth margins, no/minimal enhancement, homogeneous FLAIR suppression; 17% v. 3%, p=7x10-11) than IDH-wildtypes. 20 cases had partial T2FM (14 IDH-mutant, 6 IDH-wildtype). Sensitivity of partial T2FM for IDH-mutation was 19.7%, specificity 99.6%, positive predictive value 70%, and negative predictive value 96.1%. Subset analysis of 983 IDH-wildtypes with known MGMT methylation status (406 MGMT-hypermethylated) showed frontal lobe predominance was more common in MGMT-hypermethylated than MGMT-unmethylated (39.4% v. 24.3%, p=.02); other imaging characteristics did not significantly differ.

CONCLUSIONS

Partial T2FM is a highly specific imaging biomarker for IDH-mutation in GBM/grade-4 glioma.

NIMG-33. PROGNOSTIC STRATIFICATION OF DE NOVO GLIOBLASTOMA PATIENTS ACROSS 22 GEOGRAPHICALLY DISTINCT INSTITUTIONS: UPDATES FROM THE RESPOND CONSORTIUM

PURPOSE

Glioblastoma, IDH-wildtype, is the most common primary malignant adult brain tumor with median overall survival (OS) of ~14 months, with little improvement over the last 20 years. We hypothesize that AI-based integration of quantitative tumor characteristics, independent of acquisition protocol and equipment, can reveal accurate generalizable prognostic stratification. We seek an AI-based OS predictor using routine clinically acquired MRI sequences, quantitatively evaluated across institutions of the ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium.

METHODS

We identified a retrospective cohort of 2,293 diffuse glioma (IDH-wildtype/-NOS/-NEC) patients from 22 geographically distinct institutions across 3 continents, with preoperative structural MRI scans. The entire tumor burden was automatically segmented into 3 sub-compartments, i.e., enhancing, necrotic, peritumoral T2-FLAIR abnormality. We developed our AI predictor by multivariate integration of i)patient age, ii)tumor sub-compartment volume normalized to brain volume, iii)spatial distribution characteristics (tumor location, distance to the ventricles, and laterality), and iv)morphologic descriptors (major axes’ length, axes’ ratio, extent, and number of tumors). The AI predictor returns a continuous value between 0-1, defining short-, intermediate-, and long-survivors based on thresholds on the 25th and 75th percentiles. Leave-One-Site-Out-Cross-Validation was used to assess the generalizability of our stratification. Kaplan-Meier survival curves were computed for OS analysis and evaluated by a Cox proportional hazards model for statistical significance and hazard ratios.

RESULTS

Survival analysis yielded a hazard ratio of 2.07 (95%CI, 2.06-2.08, p-value= 4.8e-102) for patient stratification into short-, intermediate-, and long-survivors. Pearson correlation between the predicted and actual OS yielded an R= 0.49.

CONCLUSION

Multivariate integration of visually quantified tumor characteristics, agnostic to acquisition protocol/equipment, yields an accurate OS surrogate index. Validation of our AI model in the largest centralized glioblastoma imaging dataset, from the ReSPOND consortium, supports its generalizability across diverse patient populations and acquisition settings, potentially contributing to equitable improvements of personalized patient care.

Causal Inference in Radiomics: Framework, Mechanisms, and Algorithms

The widespread use of machine learning algorithms in radiomics has led to a proliferation of flexible prognostic models for clinical outcomes. However, a limitation of these techniques is their black-box nature, which prevents the ability for increased mechanistic phenomenological understanding. In this article, we develop an inferential framework for estimating causal effects with radiomics data. A new challenge is that the exposure of interest is latent so that new estimation procedures are needed. We leverage a multivariate version of partial least squares for causal effect estimation. The methodology is illustrated with applications to two radiomics datasets, one in osteosarcoma and one in glioblastoma.

Ion beam figuring with focused anode layer thruster

This work presents the peculiarities of cone ion beam formation with a focused thruster with anode layer (TAL) and its application to silicon carbide (SiC) ion beam figuring. Modeling results of Lorentz E × B force distribution in the discharge gap are presented. 3D particle tracing for keV Ar ions is carried out for the first time in the beam drift region of TAL with magnetic lens. Extracted ion beam full width at half maxima is about 2 mm in the focal plane, where the SiC etching rate reaches 0.5 µm/min. The SiC sputter yields are measured as a function of the Ar ion impact energy and beam incidence angle. The maximum sputter yield of 2.8 atom/ion is observed at 45° of the beam-sample angle for the Si targets. Furthermore, the maximum sputter yield value of 1.7 atom/ion is measured at 30° of the beam-sample angle for the SiC targets. The novelty of present research is in the application of focused TAL keV Ar ion beam to the SiC ion beam figuring.

NIMG-39. RADIOMIC ANALYSIS FOR NON-INVASIVE IN VIVO PROGNOSTIC STRATIFICATION OF DE NOVO GLIOBLASTOMA PATIENTS: A MULTI-INSTITUTIONAL EVALUATION FOR GENERALIZABILITY IN THE RESPOND CONSORTIUM

PURPOSE

Multi-parametric MRI based radiomic signatures have highlighted the promise of artificial intelligence (AI) in neuro-oncology. However, inter-institution heterogeneity hinders generalization to data from unseen clinical institutions. To this end, we formulated the ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium for glioblastoma. Here, we seek non-invasive generalizable radiomic signatures from routine clinically-acquired MRI for prognostic stratification of glioblastoma patients.

METHODS

We identified a retrospective cohort of 606 patients with near/gross total tumor resection ( &gt;90%), from 13 geographically-diverse institutions. All pre-operative structural MRI scans (T1,T1-Gd,T2,T2-FLAIR) were aligned to a common anatomical atlas. An automatic algorithm segmented the whole tumors (WTs) into 3 sub-compartments, i.e., enhancing (ET), necrotic core (NC), and peritumoral T2-FLAIR signal abnormality (ED). The combination of ET+NC defines the tumor core (TC). Quantitative radiomic features were extracted to generate our AI model to stratify patients into short- (&lt; 14mts) and long-survivors ( &gt;14mts). The model trained on 276 patients from a single institution was independently validated on 330 unseen patients from 12 left-out institutions, using the area-under-the-receiver-operating-characteristic-curve (AUC).

RESULTS

Each feature individually offered certain (limited but reproducible) value for identifying short-survivors: 1) TC closer to lateral ventricles (AUC=0.62); 2) larger ET/brain (AUC=0.61); 3) larger TC/brain (AUC=0.59); 4) larger WT/brain (AUC=0.55); 5) larger ET/WT (AUC=0.59); 6) smaller ED/WT (AUC=0.57); 7) larger ventricle deformations (AUC=0.6). Integrating all features and age, through a multivariate AI model, resulted in higher accuracy (AUC=0.7; 95% C.I.,0.64-0.77).

CONCLUSION

Prognostic stratification using basic radiomic features is highly reproducible across diverse institutions and patient populations. Multivariate integration yields relatively more accurate and generalizable radiomic signatures, across institutions. Our results offer promise for generalizable non-invasive in vivo signatures of survival prediction in patients with glioblastoma. Extracted features from clinically-acquired imaging, renders these signatures easier for clinical translation. Large-scale evaluation could contribute to improving patient management and treatment planning.

*Indicates equal authorship.

NIMG-22. PREDICTION OF GLIOBLASTOMA CELLULAR INFILTRATION AND RECURRENCE USING MACHINE LEARNING AND MULTI-PARAMETRIC MRI ANALYSIS: RESULTS FROM THE MULTI-INSTITUTIONAL RESPOND CONSORTIUM

PURPOSE

Multi-parametric MRI and artificial intelligence (AI) methods were previously used to predict peritumoral neoplastic cell infiltration and risk of future recurrence in glioblastoma, in single-institution studies. We hypothesize that important characteristics of peritumoral tissue heterogeneity captured, engineered/selected, and quantified by these methods relate to predictions generalizable in the multi-institutional ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium.

METHODS

To support further development, generalization, and clinical translation of our proposed method, we trained the AI model on a retrospective cohort of 29 de novo glioblastoma patients from the Hospital of the University of Pennsylvania (UPenn) (Male/Female:20/9, age:22-78 years) followed by evaluation on a prospective multi-institutional cohort of 84 glioblastoma patients (Male/Female:51/33, age:34-89 years) from Case Western Reserve University/University Hospitals (CWRU/UH, 25), New York University (NYU, 13), Ohio State University (OSU, 13), University Hospital Río Hortega (RH, 2), and UPenn (31). Features extracted from pre-resection MRI (T1, T1-Gd, T2, T2-FLAIR, ADC) were used to build our model predicting the spatial pattern of subsequent tumor recurrence. These predictions were evaluated against regions of pathology-confirmed post-resection recurrence.

RESULTS

Our model predicted the locations that later harbored tumor recurrence with sensitivity 83%, AUC 0.83 (99% CI, 0.73-0.93), and odds ratio 7.23 (99% CI, 7.09-7.37) in the prospective cohort. Odds ratio (99% CI)/AUC(99% CI) per institute were: CWRU/UH, 7.8(7.6-8.1)/0.82(0.75-0.89); NYU, 3.5(3.3-3.6)/0.84(074-0.93); OSU, 7.9(7.6-8.3)/0.8(0.67-0.94); RH, 22.7(20-25.1)/0.94(0.27-1); UPenn, 7.1(6.8-7.3)/0.83(0.76-0.91).

CONCLUSION

This is the first study that provides relatively extensive multi-institutional validated evidence that AI can provide good predictions of peritumoral neoplastic cell infiltration and future recurrence, by dissecting the MRI signal heterogeneity in peritumoral tissue. Our analyses leveraged the unique dataset of the ReSPOND consortium, which aims to develop and evaluate AI-based biomarkers for individualized prediction and prognostication, by moving from single-institution studies to generalizable, well-validated multi-institutional predictive biomarkers.

[MUC4 Silencing Inhibits TGF-β1-induced Epithelial-mesenchymal Transition VIA the ERK1/2 Pathway in Human Airway Epithelial NCI-H292 Cells]

MUC4 is a predominant membrane-tethered mucin lubricating and protecting the epithelial surface and playing various biological roles in the renewal and differentiation of epithelial cells, cell signaling, cell adhesion, and carcinogenesis. Interestingly, recent studies have demonstrated that MUC4 expression regulates the epithelial-mesenchymal transition (EMT) of cancer cells in ovarian, pancreatic, and lung cancer. However, the effects of MUC4 expression on EMT in human airway epithelial cells are not yet well known. Here, we describe the effects of transforming growth factor beta 1 (TGF-β1)-induced MUC4 expression on EMT and evaluate its downstream signaling pathway in human airway epithelial cells. In human airway epithelial NCI-H292 cells, exposure to TGF-β1 induced expression of MUC4, CDH2, VIM and SNAI1 genes and encoded by them proteins, MUC4, N-cadherin, vimentin and Snail, and reduced the level of CDH1 and its product, E-cadherin. In MUC4-knockdown cells, TGF-β1-induced expression levels of MUC4, CDH2, VIM and SNAI1 and corresponding proteins were suppressed, but CDH1 and E-cadherin levels were not. In addition, TGF-β1-induced phosphorylation of extracellular signal regulated kinase 1/2 (ERK1/2) was suppressed, but that of Smad2/3, Akt, and p38 was not. The results of this study suggest that MUC4 silencing inhibits TGF-β1 -induced EMT via the ERK1/2 pathway, and a possible role of MUC4 in the induction of EMT in human airway epithelial cells.

A Tale of Two Organ Systems: Imaging review of diseases affecting the thoracic and neurological systems. Part 1

In an era of rapidly expanding knowledge and sub-specialization, it is becoming increasingly common to focus on one organ system. However, the human body is intimately linked, and disease processes affecting one region of the body not uncommonly affect the other organ systems as well. Understanding diseases from a macroscopic perspective, rather than a narrow vantage point, enables efficient and accurate diagnosis. This tenet holds true for diseases affecting both the thoracic and neurological systems; in isolation, the radiologic appearance of disease in one organ system may be nonspecific, but viewing the pathophysiologic process in both organ systems may markedly narrow the differential considerations, and potentially lead to a definitive diagnosis. In this article, we discuss a variety of disease entities known to affect both the thoracic and neurological systems, either manifesting simultaneously or at different periods of time. Some of these conditions may show neither thoracic nor neurological manifestations. These diseases have been systematically classified into infectious, immune-mediated / inflammatory, vascular, syndromic / hereditary and neoplastic disorders. The underlying pathophysiological mechanisms linking both regions and radiologic appearances in both organ systems are discussed. When appropriate, brief clinical and diagnostic information is provided. Ultimately, accurate diagnosis will lead to expedited triage and prompt institution of potentially life-saving treatment for these groups of complex disorders.

Deep-learning-based cardiovascular risk stratification using coronary artery calcium scores predicted from retinal photographs

BACKGROUND

Coronary artery calcium (CAC) score is a clinically validated marker of cardiovascular disease risk. We developed and validated a novel cardiovascular risk stratification system based on deep-learning-predicted CAC from retinal photographs.

METHODS

We used 216 152 retinal photographs from five datasets from South Korea, Singapore, and the UK to train and validate the algorithms. First, using one dataset from a South Korean health-screening centre, we trained a deep-learning algorithm to predict the probability of the presence of CAC (ie, deep-learning retinal CAC score, RetiCAC). We stratified RetiCAC scores into tertiles and used Cox proportional hazards models to evaluate the ability of RetiCAC to predict cardiovascular events based on external test sets from South Korea, Singapore, and the UK Biobank. We evaluated the incremental values of RetiCAC when added to the Pooled Cohort Equation (PCE) for participants in the UK Biobank.

FINDINGS

RetiCAC outperformed all single clinical parameter models in predicting the presence of CAC (area under the receiver operating characteristic curve of 0·742, 95% CI 0·732-0·753). Among the 527 participants in the South Korean clinical cohort, 33 (6·3%) had cardiovascular events during the 5-year follow-up. When compared with the current CAC risk stratification (0, >0-100, and >100), the three-strata RetiCAC showed comparable prognostic performance with a concordance index of 0·71. In the Singapore population-based cohort (n=8551), 310 (3·6%) participants had fatal cardiovascular events over 10 years, and the three-strata RetiCAC was significantly associated with increased risk of fatal cardiovascular events (hazard ratio [HR] trend 1·33, 95% CI 1·04-1·71). In the UK Biobank (n=47 679), 337 (0·7%) participants had fatal cardiovascular events over 10 years. When added to the PCE, the three-strata RetiCAC improved cardiovascular risk stratification in the intermediate-risk group (HR trend 1·28, 95% CI 1·07-1·54) and borderline-risk group (1·62, 1·04-2·54), and the continuous net reclassification index was 0·261 (95% CI 0·124-0·364).

INTERPRETATION

A deep learning and retinal photograph-derived CAC score is comparable to CT scan-measured CAC in predicting cardiovascular events, and improves on current risk stratification approaches for cardiovascular disease events. These data suggest retinal photograph-based deep learning has the potential to be used as an alternative measure of CAC, especially in low-resource settings.

FUNDING

Yonsei University College of Medicine; Ministry of Health and Welfare, Korea Institute for Advancement of Technology, South Korea; Agency for Science, Technology, and Research; and National Medical Research Council, Singapore.

Fully automated hybrid approach to predict the IDH mutation status of gliomas via deep learning and radiomics

BACKGROUND

Glioma prognosis depends on isocitrate dehydrogenase (IDH) mutation status. We aimed to predict the IDH status of gliomas from preoperative MR images using a fully automated hybrid approach with convolutional neural networks (CNNs) and radiomics.

METHODS

We reviewed 1166 preoperative MR images of gliomas (grades II-IV) from Severance Hospital (n = 856), Seoul National University Hospital (SNUH; n = 107), and The Cancer Imaging Archive (TCIA; n = 203). The Severance set was subdivided into the development (n = 727) and internal test (n = 129) sets. Based on T1 postcontrast, T2, and fluid-attenuated inversion recovery images, a fully automated model was developed that comprised a CNN for tumor segmentation (Model 1) and CNN-based classifier for IDH status prediction (Model 2) that uses a hybrid approach based on 2D tumor images and radiomic features from 3D tumor shape and loci guided by Model 1. The trained model was tested on internal (a subset of the Severance set) and external (SNUH and TCIA) test sets.

RESULTS

The CNN for tumor segmentation (Model 1) achieved a dice coefficient of 0.86-0.92 across datasets. Our hybrid model achieved accuracies of 93.8%, 87.9%, and 78.8%, with areas under the receiver operating characteristic curves of 0.96, 0.94, and 0.86 and areas under the precision-recall curves of 0.88, 0.82, and 0.81 in the internal test, SNUH, and TCIA sets, respectively.

CONCLUSIONS

Our fully automated hybrid model demonstrated the potential to be a highly reproducible and generalizable tool across different datasets for the noninvasive prediction of the IDH status of gliomas.

Timing and clinical outcomes of tracheostomy in patients with COVID-19

In this retrospective multicentre cohort study that included 27 COVID-19 patients who underwent tracheostomy, the mean time between intubation and tracheostomy was 15.8 days and the negative conversion time of COVID-19 was 43.1 days. Eleven patients (40.7%) died of COVID-19 and the use of percutaneous dilatation tracheostomy was significantly associated with in-hospital death. Timely tracheostomy could be performed in COVID-19 patients, regardless of duration of intubation or positivity of COVID-19 test, with an open surgical tracheostomy as a preferable technique.

Squamous Cell Carcinoma and Lymphoma of the Oropharynx: Differentiation Using a Radiomics Approach

The purpose of this study was to evaluate the diagnostic performance of magnetic resonance (MR) radiomics-based machine learning algorithms in differentiating squamous cell carcinoma (SCC) from lymphoma in the oropharynx. MR images from 87 patients with oropharyngeal SCC (n=68) and lymphoma (n=19) were reviewed retrospectively. Tumors were semi-automatically segmented on contrast-enhanced T1-weighted images registered to T2-weighted images, and radiomic features (n=202) were extracted from contrast-enhanced T1- and T2-weighted images. The radiomics classifier was built using elastic-net regularized generalized linear model analyses with nested five-fold cross-validation. The diagnostic abilities of the radiomics classifier and visual assessment by two head and neck radiologists were evaluated using receiver operating characteristic (ROC) analyses for distinguishing SCC from lymphoma. Nineteen radiomics features were selected at least twice during the five-fold cross-validation. The mean area under the ROC curve (AUC) of the radiomics classifier was 0.750 [95% confidence interval (CI), 0.613-0.887], with a sensitivity of 84.2%, specificity of 60.3%, and an accuracy of 65.5%. Two human readers yielded AUCs of 0.613 (95% CI, 0.467-0.759) and 0.663 (95% CI, 0.531-0.795), respectively. The radiomics-based machine learning model can be useful for differentiating SCC from lymphoma of the oropharynx.

Fipronil upregulates inflammatory cytokines and MUC5AC expression in human nasal epithelial cells

BACKGROUND

Airway inflammation and excessive mucin production are pathophysiological characteristics of airway diseases. Fipronil, a pesticide, is being extensively used in agriculture and veterinary medicine worldwide. However, this compound impairs immune function in non-target organisms. The present study aimed to evaluate the effect of fipronil on pro-inflammatory cytokine and mucus production and signalling pathways in human primary nasal METHODOLOGY: The effect of fipronil on pro-inflammatory cytokine and MUC5AC expression and the signalling pathway of fipronil were investigated using real-time PCR, enzyme immunoassays, immunofluorescence, and immunoblot analysis with specific inhibitors and small interfering RNA.

RESULTS

Fipronil treatment increased pro-inflammatory cytokine interleukin (IL)-1beta, IL-6, IL-8, and MUC5AC expression in human primary nasal epithelial cells. It also induced phosphorylation of extracellular signal-regulated kinase 1/2 (ERK1/2) mitogenactivated protein kinase (MAPK), p38 MAPK, and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB). MAPK and NF-kB inhibitor treatment significantly inhibited increases in IL-1beta, IL-6, IL-8, and MUC5AC expression. Ex vivo data confirmed that fipronil-induced MUC5AC expression occurs through ERK1/2, p38, and NF-kB signalling pathways in nasal inferior turbinate tissue.

CONCLUSIONS

Fipronil induced pro-inflammatory cytokine IL-1beta, IL-6, IL-8, and MUC5AC expression via ERK1/2 MAPK, p38 MAPK, and NF-kB in human primary nasal epithelial cells.

Machine Learning Based Radiomic HPV Phenotyping of Oropharyngeal SCC: A Feasibility Study Using MRI

OBJECTIVES

To investigate whether a radiomic MRI feature-based prediction model can differentiate oropharyngeal squamous cell carcinoma (SCC) according to the human papillomavirus (HPV) status.

STUDY DESIGN

Retrospective cohort study.

METHODS

Pretreatment MRI data from 62 consecutive patients with oropharyngeal SCC were retrospectively reviewed, and chronologically allocated to training (n = 43) and test sets (n = 19). Enhancing tumors were semi-automatically segmented on each slice of the postcontrast T1WI to span the entire tumor volume, after registration of T2WI to postcontrast T1WI; 170 radiomic features were extracted from the entire tumor volume. Relevant features were selected and radiomics models were trained using least absolute shrinkage and selection operator (LASSO) logistic regression model with 10-fold cross-validation, after subsampling of training sets using synthetic minority over-sampling technique to mitigate data imbalance. The selected features, weighted by their respective coefficients, were combined linearly to yield a radiomics score. The diagnostic performance of the radiomic score was evaluated using the area under the receiver operating characteristic curve (AUC).

RESULTS

Six radiomic features, which revealed strong association with HPV status of oropharyngeal SCC, were selected using LASSO. The radiomics model yielded excellent performance on the training set (AUC, 0.982 [95% CI, 0.942-1.000]) and moderate performance on the test set (AUC, 0.744 [95% CI, 0.496-0.991]) for differentiating oropharyngeal SCC according to HPV status.

CONCLUSIONS

Radiomics-based MRI phenotyping differentiates oropharyngeal SCC according to HPV status, and thus, is a potential imaging biomarker.

LEVEL OF EVIDENCE

3 Laryngoscope, 131:E851-E856, 2021.

Radiomics MRI Phenotyping with Machine Learning to Predict the Grade of Lower-Grade Gliomas: A Study Focused on Nonenhancing Tumors

Objective

To assess whether radiomics features derived from multiparametric MRI can predict the tumor grade of lower-grade gliomas (LGGs; World Health Organization grade II and grade III) and the nonenhancing LGG subgroup.

Materials and Methods

Two-hundred four patients with LGGs from our institutional cohort were allocated to training (n = 136) and test (n = 68) sets. Postcontrast T1-weighted images, T2-weighted images, and fluid-attenuated inversion recovery images were analyzed to extract 250 radiomics features. Various machine learning classifiers were trained using the radiomics features to predict the glioma grade. The trained classifiers were internally validated on the institutional test set and externally validated on a separate cohort (n = 99) from The Cancer Genome Atlas (TCGA). Classifier performance was assessed by determining the area under the curve (AUC) from receiver operating characteristic curve analysis. An identical process was performed in the nonenhancing LGG subgroup (institutional training set, n = 73; institutional test set, n = 37; and TCGA cohort, n = 37) to predict the glioma grade.

Results

The performance of the best classifier was good in the internal validation set (AUC, 0.85) and fair in the external validation set (AUC, 0.72) to predict the LGG grade. For the nonenhancing LGG subgroup, the performance of the best classifier was good in the internal validation set (AUC, 0.82), but poor in the external validation set (AUC, 0.68).

Conclusion

Radiomics feature-based classifiers may be useful to predict LGG grades. However, radiomics classifiers may have a limited value when applied to the nonenhancing LGG subgroup in a TCGA cohort.

Machine learning and radiomic phenotyping of lower grade gliomas: improving survival prediction

BACKGROUND AND PURPOSE

Recent studies have highlighted the importance of isocitrate dehydrogenase (IDH) mutational status in stratifying biologically distinct subgroups of gliomas. This study aimed to evaluate whether MRI-based radiomic features could improve the accuracy of survival predictions for lower grade gliomas over clinical and IDH status.

MATERIALS AND METHODS

Radiomic features (n = 250) were extracted from preoperative MRI data of 296 lower grade glioma patients from databases at our institutional (n = 205) and The Cancer Genome Atlas (TCGA)/The Cancer Imaging Archive (TCIA) (n = 91) datasets. For predicting overall survival, random survival forest models were trained with radiomic features; non-imaging prognostic factors including age, resection extent, WHO grade, and IDH status on the institutional dataset, and validated on the TCGA/TCIA dataset. The performance of the random survival forest (RSF) model and incremental value of radiomic features were assessed by time-dependent receiver operating characteristics.

RESULTS

The radiomics RSF model identified 71 radiomic features to predict overall survival, which were successfully validated on TCGA/TCIA dataset (iAUC, 0.620; 95% CI, 0.501-0.756). Relative to the RSF model from the non-imaging prognostic parameters, the addition of radiomic features significantly improved the overall survival prediction accuracy of the random survival forest model (iAUC, 0.627 vs. 0.709; difference, 0.097; 95% CI, 0.003-0.209).

CONCLUSION

Radiomic phenotyping with machine learning can improve survival prediction over clinical profile and genomic data for lower grade gliomas.

KEY POINTS

• Radiomics analysis with machine learning can improve survival prediction over the non-imaging factors (clinical and molecular profiles) for lower grade gliomas, across different institutions.

Randomly Hopping Majorana Fermions in the Diluted Kitaev System α-Ru_{0.8}Ir_{0.2}Cl_{3}

dc and ac magnetic susceptibility, magnetization, specific heat, and Raman scattering measurements are combined to probe low-lying spin excitations in α-Ru_{1-x}Ir_{x}Cl_{3} (x≈0.2), which realizes a disordered spin liquid. At intermediate energies (ℏω>3  meV), Raman spectroscopy evidences linearly ω-dependent Majorana-like excitations, obeying Fermi statistics. This points to robustness of a Kitaev paramagnetic state under spin vacancies. At low energies below 3 meV, we observe power-law dependences and quantum-critical-like scalings of the thermodynamic quantities, implying the presence of a weakly divergent low-energy density of states. This scaling phenomenology is interpreted in terms of the random hoppings of Majorana fermions. Our results demonstrate an emergent hierarchy of spin excitations in a diluted Kitaev honeycomb system subject to spin vacancies and bond randomness.

Diffusion tensor imaging radiomics in lower-grade glioma: improving subtyping of isocitrate dehydrogenase mutation status

PURPOSE

To evaluate whether diffusion tensor imaging (DTI) radiomics with machine learning improves the prediction of isocitrate dehydrogenase (IDH) mutation status of lower-grade gliomas beyond radiomic features from conventional MRI and DTI histogram parameters.

METHODS

A total of 168 patients with pathologically confirmed lower-grade gliomas were retrospectively enrolled. A total of 158 and 253 radiomic features were extracted from DTI (DTI radiomics) and conventional MRI (T1-weighted image with contrast enhancement, T2-weighted image, and FLAIR [conventional radiomics]), respectively. The random forest models for predicting IDH status were trained with variable combinations as follows: (1) DTI radiomics, (2) conventional radiomics, (3) conventional radiomics + DTI radiomics, and (4) conventional radiomics + DTI histogram. The models were validated with nested cross-validation. The predictive performances of those models were compared by using area under the curve (AUC) from receiver operating characteristic analysis, and 95% confidence interval (CI) was calculated.

RESULTS

Adding DTI radiomics to conventional radiomics significantly improved the accuracy of IDH status subtyping (AUC, 0.900 [95% CI, 0.855-0.945], p = 0.006), whereas adding DTI histogram parameters yielded nonsignificant trend toward improvement (0.869 [95% CI, 0.816-0.922], p = 0.150) compared with the model with conventional radiomics alone (0.835 [95% CI, 0.773-0.896]). The performance of the model consisting of both DTI and conventional radiomics was significantly superior than that of model consisting of both DTI histogram parameters and conventional radiomics (0.900 vs 0.869, p = 0.040).

CONCLUSION

DTI radiomics with machine learning can help improve the subtyping of IDH status beyond conventional radiomics and DTI histogram parameters in patients with lower-grade gliomas.

Exotic Low-Energy Excitations Emergent in the Random Kitaev Magnet Cu_{2}IrO_{3}

We report on magnetization M(H), dc and ac magnetic susceptibility χ(T), specific heat C_{m}(T) and muon spin relaxation (μSR) measurements of the Kitaev honeycomb iridate Cu_{2}IrO_{3} with quenched disorder. In spite of the chemical disorders, we find no indication of spin glass down to 260 mK from the C_{m}(T) and μSR data. Furthermore, a persistent spin dynamics observed by the zero-field muon spin relaxation evidences an absence of static magnetism. The remarkable observation is a scaling relation of χ[H,T] and M[H,T] in H/T with the scaling exponent α=0.26-0.28, expected from bond randomness. However, C_{m}[H,T]/T disobeys the predicted universal scaling law, pointing towards the presence of additional low-lying excitations on the background of bond-disordered spin liquid. Our results signify a many-faceted impact of quenched disorder in a Kitaev spin system due to its peculiar bond character.

Gender differences in limited duty time for lower limb injury

Background

Among active-duty military personnel, lower limb musculoskeletal injuries and related conditions (injuries) frequently arise as unintended consequences of physical training. These injuries are particularly common among women. The practical impact of such injuries on temporary military occupational disability has not been estimated with precision on a large scale.

Aims

To determine the proportion of service time compromised by limited duty days attributable to lower limb injuries, characterize the time affected by these limitations in terms of specific lower limb region and compare the limited duty time between male and female soldiers.

Methods

Administrative data and individual limited duty assignments (profiles) were obtained for active-duty US Army personnel who served in 2014. Lower limb injury-related profiles were used to calculate the percent of person-time requiring duty limitations by gender and body region.

Results

The study group was 568 753 soldiers of whom 14% were women. Nearly 13% of service days for active-duty US Army soldiers required limited duty for lower limb injuries during 2014. Knee injuries were responsible for 45% of those days. Within integrated military occupations, female soldiers experienced 27-57% more time on limited duty for lower limb injuries compared with men.

Conclusions

The substantial amount of limited duty for lower limb musculoskeletal injuries among soldiers highlights the need for improvement in training-related injury screening, prevention and timely treatment with particular attention to knee injuries. The excessive impact of lower limb injuries on female soldiers' occupational functions should be a surveillance priority in the current environment of expanding gender-integrated training.

The added prognostic value of radiological phenotype combined with clinical features and molecular subtype in anaplastic gliomas

PURPOSE

To determine whether radiological phenotype can improve the predictive performance of the risk model based on molecular subtype and clinical risk factors in anaplastic glioma patients.

METHODS

This retrospective study was approved by our institutional review board with waiver of informed consent. MR images of 86 patients with pathologically diagnosed anaplastic glioma (WHO grade III) between January 2007 and February 2016 were analyzed according to the Visually Accessible Rembrandt Images (VASARI) features set. Significant imaging findings were selected to generate a radiological risk score (RRS) for overall survival (OS) and progression-free survival (PFS) using the least absolute shrinkage and selection operator (LASSO) Cox regression model. The prognostic value of RRS was evaluated with multivariate Cox regression including molecular subtype and clinical risk factors. The C-indices of multivariate models with and without RRS were compared by bootstrapping.

RESULTS

Eight VASARI features contributed to RRS for OS and six contributed to PFS. Multifocality or multicentricity was the most influential feature, followed by restricted diffusion. RRS was significantly associated with OS and PFS (P < .001), as well as age and molecular subtype. The multivariate model with RRS demonstrated a significantly higher predictive performance than the model without (C-index difference: 0.074, 95% confidence interval [CI]: 0.031, 0.148 for OS; C-index difference: 0.054, 95% CI: 0.014, 0.123 for PFS).

CONCLUSION

RRS derived from VASARI features was an independent predictor of survival in patients with anaplastic gliomas. The addition of RRS significantly improved the predictive performance of the molecular feature based model.

Radiomic MRI Phenotyping of Glioblastoma: Improving Survival Prediction

Purpose To investigate whether radiomic features at MRI improve survival prediction in patients with glioblastoma multiforme (GBM) when they are integrated with clinical and genetic profiles. Materials and Methods Data in patients with a diagnosis of GBM between December 2009 and January 2017 (217 patients) were retrospectively reviewed up to May 2017 and allocated to training and test sets (3:1 ratio). Radiomic features (n = 796) were extracted from multiparametric MRI. A random survival forest (RSF) model was trained with the radiomic features along with clinical and genetic profiles (O-6-methylguanine-DNA-methyltransferase promoter methylation and isocitrate dehydrogenase 1 mutation statuses) to predict overall survival (OS) and progression-free survival (PFS). The RSF models were validated on the test set. The incremental values of radiomic features were evaluated by using the integrated area under the receiver operating characteristic curve (iAUC). Results The 217 patients had a mean age of 57.9 years, and there were 87 female patients (age range, 22-81 years) and 130 male patients (age range, 17-85 years). The median OS and PFS of patients were 352 days (range, 20-1809 days) and 264 days (range, 21-1809 days), respectively. The RSF radiomics models were successfully validated on the test set (iAUC, 0.652 [95% confidence interval {CI}, 0.524, 0.769] and 0.590 [95% CI: 0.502, 0.689] for OS and PFS, respectively). The addition of a radiomics model to clinical and genetic profiles improved survival prediction when compared with models containing clinical and genetic profiles alone (P = .04 and .03 for OS and PFS, respectively). Conclusion Radiomic MRI phenotyping can improve survival prediction when integrated with clinical and genetic profiles and thus has potential as a practical imaging biomarker. © RSNA, 2018 Online supplemental material is available for this article. See also the editorial by Jain and Lui in this issue.

Identification of transfusion-transmitted hepatitis A through postdonation information in Korea: results of an HAV lookback (2007-2012)

BACKGROUND AND OBJECTIVES

Despite safety measures to minimize the risk of transfusion-transmitted infections, a residual risk remains. To trace and review some such cases, we ask donors to notify the blood centre if they are diagnosed with an infection after they donate blood.

MATERIALS AND METHODS

We analysed all data on postdonation cases of hepatitis A reported between 2007 and 2012. Archived specimens from these donors were tested for hepatitis A virus (HAV) using anti-HAV IgM/IgG and HAV-PCR as markers. If any of the test results were positive, we reviewed the medical records of the recipients and, if necessary, tested them for hepatitis A.

RESULTS

Fifteen blood donors notified the blood centres of having been diagnosed with hepatitis A after donation. All archived samples except for one were HAV-PCR-positive and anti-HAV IgM/IgG-negative. Of the donated components, four RBCs and 14 FFPs had not been transfused to patients and were recalled. Among 26 recipients of the implicated components, fourteen were still alive when they were notified. Two patients showed clinical symptoms of hepatitis A and had positive results with anti-HAV IgM.

CONCLUSION

Transfusion-transmitted hepatitis A is rare but exists. To reduce the risk, donors should be told to notify the blood centre if they are diagnosed with blood-borne diseases after they donate blood. Physicians should consider the possibility of transfusion-transmitted hepatitis A if a transfused patient has hepatitis A but no history of travel or route of faecal-oral infection.

Functional Communication Profiles in Children with Cerebral Palsy in Relation to Gross Motor Function and Manual and Intellectual Ability

PURPOSE

The aim of the present study was to investigate communication function using classification systems and its association with other functional profiles, including gross motor function, manual ability, intellectual functioning, and brain magnetic resonance imaging (MRI) characteristics in children with cerebral palsy (CP).

MATERIALS AND METHODS

This study recruited 117 individuals with CP aged from 4 to 16 years. The Communication Function Classification System (CFCS), Viking Speech Scale (VSS), Speech Language Profile Groups (SLPG), Gross Motor Function Classification System (GMFCS), Manual Ability Classification System (MACS), and intellectual functioning were assessed in the children along with brain MRI categorization.

RESULTS

Very strong relationships were noted among the VSS, CFCS, and SLPG, although these three communication systems provide complementary information, especially for children with mid-range communication impairment. These three communication classification systems were strongly related with the MACS, but moderately related with the GMFCS. Multiple logistic regression analysis indicated that manual ability and intellectual functioning were significantly related with VSS and CFCS function, whereas only intellectual functioning was significantly related with SLPG functioning in children with CP. Communication function in children with a periventricular white matter lesion (PVWL) varied widely. In the cases with a PVWL, poor functioning was more common on the SLPG, compared to the VSS and CFCS.

CONCLUSION

Very strong relationships were noted among three communication classification systems that are closely related with intellectual ability. Compared to gross motor function, manual ability seemed more closely related with communication function in these children.

Primary central nervous system lymphoma and atypical glioblastoma: Differentiation using radiomics approach

OBJECTIVES

To evaluate the diagnostic performance of magnetic resonance (MR) radiomics-based machine-learning algorithms in differentiating primary central nervous system lymphoma (PCNSL) from non-necrotic atypical glioblastoma (GBM).

METHODS

Seventy-seven patients (54 individuals with PCNSL and 23 with non-necrotic atypical GBM), diagnosed from January 2009 to April 2017, were enrolled in this retrospective study. A total of 6,366 radiomics features, including shape, volume, first-order, texture, and wavelet-transformed features, were extracted from multi-parametric (post-contrast T1- and T2-weighted, and fluid attenuation inversion recovery images) and multiregional (enhanced and non-enhanced) tumour volumes. These features were subjected to recursive feature elimination and random forest (RF) analysis with nested cross-validation. The diagnostic abilities of a radiomics machine-learning classifier, apparent diffusion coefficient (ADC), and three readers, who independently classified the tumours based on conventional MR sequences, were evaluated using receiver operating characteristic (ROC) analysis. Areas under the ROC curves (AUC) of the radiomics classifier, ADC value, and the radiologists were compared.

RESULTS

The mean AUC of the radiomics classifier was 0.921 (95 % CI 0.825-0.990). The AUCs of the three readers and ADC were 0.707 (95 % CI 0.622-0.793), 0.759 (95 %CI 0.656-0.861), 0.695 (95 % CI 0.590-0.800) and 0.684 (95 % CI0.560-0.809), respectively. The AUC of the radiomics-based classifier was significantly higher than those of the three readers and ADC (p< 0.001 for all).

CONCLUSIONS

Large-scale radiomics with a machine-learning algorithm can be useful for differentiating PCNSL from atypical GBM, and yields a better diagnostic performance than human radiologists and ADC values.

KEY POINTS

• Machine-learning algorithm radiomics can help to differentiate primary central PCNSL from GBM. • This approach yields a higher diagnostic accuracy than visual analysis by radiologists. • Radiomics can strengthen radiologists' diagnostic decisions whenever conventional MRI sequences are available.

Whole-Tumor Histogram and Texture Analyses of DTI for Evaluation of IDH1-Mutation and 1p/19q-Codeletion Status in World Health Organization Grade II Gliomas

BACKGROUND AND PURPOSE

Prediction of the isocitrate dehydrogenase 1 (IDH1)-mutation and 1p/19q-codeletion status of World Health Organization grade ll gliomas preoperatively may assist in predicting prognosis and planning treatment strategies. Our aim was to characterize the histogram and texture analyses of apparent diffusion coefficient and fractional anisotropy maps to determine IDH1-mutation and 1p/19q-codeletion status in World Health Organization grade II gliomas.

MATERIALS AND METHODS

Ninety-three patients with World Health Organization grade II gliomas with known IDH1-mutation and 1p/19q-codeletion status (18 IDH1 wild-type, 45 IDH1 mutant and no 1p/19q codeletion, 30 IDH1-mutant and 1p/19q codeleted tumors) underwent DTI. ROIs were drawn on every section of the T2-weighted images and transferred to the ADC and the fractional anisotropy maps to derive volume-based data of the entire tumor. Histogram and texture analyses were correlated with the IDH1-mutation and 1p/19q-codeletion status. The predictive powers of imaging features for IDH1 wild-type tumors and 1p/19q-codeletion status in IDH1-mutant subgroups were evaluated using the least absolute shrinkage and selection operator.

RESULTS

Various histogram and texture parameters differed significantly according to IDH1-mutation and 1p/19q-codeletion status. The skewness and energy of ADC, 10th and 25th percentiles, and correlation of fractional anisotropy were independent predictors of an IDH1 wild-type in the least absolute shrinkage and selection operator. The area under the receiver operating curve for the prediction model was 0.853. The skewness and cluster shade of ADC, energy, and correlation of fractional anisotropy were independent predictors of a 1p/19q codeletion in IDH1-mutant tumors in the least absolute shrinkage and selection operator. The area under the receiver operating curve was 0.807.

CONCLUSIONS

Whole-tumor histogram and texture features of the ADC and fractional anisotropy maps are useful for predicting the IDH1-mutation and 1p/19q-codeletion status in World Health Organization grade II gliomas.

Prediction of IDH1-Mutation and 1p/19q-Codeletion Status Using Preoperative MR Imaging Phenotypes in Lower Grade Gliomas

BACKGROUND AND PURPOSE

WHO grade II gliomas are divided into three classes: isocitrate dehydrogenase (IDH)-wildtype, IDH-mutant and no 1p/19q codeletion, and IDH-mutant and 1p/19q-codeleted. Different molecular subtypes have been reported to have prognostic differences and different chemosensitivity. Our aim was to evaluate the predictive value of imaging phenotypes assessed with the Visually AcceSAble Rembrandt Images lexicon for molecular classification of lower grade gliomas.

MATERIALS AND METHODS

MR imaging scans of 175 patients with lower grade gliomas with known IDH1 mutation and 1p/19q-codeletion status were included (78 grade II and 97 grade III) in the discovery set. MR imaging features were reviewed by using Visually AcceSAble Rembrandt Images (VASARI); their associations with molecular markers were assessed. The predictive power of imaging features for IDH1-wild type tumors was evaluated using the Least Absolute Shrinkage and Selection Operator. We tested the model in a validation set (40 subjects).

RESULTS

Various imaging features were significantly different according to IDH1 mutation. Nonlobar location, larger proportion of enhancing tumors, multifocal/multicentric distribution, and poor definition of nonenhancing margins were independent predictors of an IDH1 wild type according to the Least Absolute Shrinkage and Selection Operator. The areas under the curve for the prediction model were 0.859 and 0.778 in the discovery and validation sets, respectively. The IDH1-mutant, 1p/19q-codeleted group frequently had mixed/restricted diffusion characteristics and showed more pial invasion compared with the IDH1-mutant, no codeletion group.

CONCLUSIONS

Preoperative MR imaging phenotypes are different according to the molecular markers of lower grade gliomas, and they may be helpful in predicting the IDH1-mutation status.

Stroke risk among adult patients with third, fourth or sixth cranial nerve palsy: a Nationwide Cohort Study

PURPOSE

This study sought to determine whether isolated third, fourth and sixth cranial nerve palsies (NPs) are associated with increased short- and long-term risk of a subsequent stroke.

METHODS

This was a nationwide retrospective propensity score-matched cohort study. A cohort of patients with NP (n = 466) and a randomly selected, propensity-matched control cohort (n = 2281) were extracted from the Korean national insurance claim database. Subjects were tracked for 5 years total, subdivided into periods of 0-1 years, 1-3 years and 3-5 years. We assessed the risk of stroke using hazard ratios (HRs) and confidence intervals (CIs) after adjustments using Cox regression at different time intervals.

RESULTS

The median follow-up was 3.1 years. Stroke developed in 18.9% of the NP cohort and 7.5% of the control cohort. Stroke risk after NP was highest in the first year [14.7 per 100 person-year at 0-1 years (HR = 6.6), 3.1 per 100 person-year at 1-3 years (HR = 1.6) and 4.3 per 100 person-year at 3-5 years (HR = 2.8)]. Each type of NP was also associated with stroke risk: within 0-1 years, stroke risk was increased in third (HR = 7.6), fourth (HR = 6.0) and sixth (HR = 5. 84) NPs. In the 3- to 5-year period, risk was increased in sixth (HR = 4.7) and fourth (HR = 3.3) NPs, but not third (HR = 0.6) NPs.

CONCLUSION

Patients in the NP cohort were more likely to have a stroke than those in the matched control cohort; the increased risk was both time- and cranial nerve-dependent.

Morphometric analysis of the maxillary root apex positions according to crowding severity

OBJECTIVES

To determine differences in arch forms derived from the root apices locations between individuals with <2 mm maxillary crowding and controls.

SETTING AND SAMPLE POPULATION

The Department of Orthodontics, Pusan National University. Cone-beam computed tomography (CBCT) images of 102 patients in the control group and 95 patients in the crowding group.

MATERIALS AND METHODS

X, Y and Z coordinates of the tip of the crowns and the apex of the root of the maxillary teeth (except second molars) were determined on the CBCT images. The acquired three-dimensional (3D) coordinates were converted into two-dimensional (2D) coordinates via projection on the palatal plane, and the Procrustes analysis was employed to process the converted 2D coordinates. The mean shape of the arch form derived from the location of the tip of the crowns and the apex of the root was compared between groups using the statistical shape analysis.

RESULTS

There was a statistically significant difference (P = .046) between the groups for the mean shape of the root apex arch form, but the difference was small and clinically irrelevant as it is minor compared to the degree of crowding.

CONCLUSIONS

Maxillary arch from at the level of the maxillary apices only shows minor differences between crowded and non-crowded dentitions.

Amide proton transfer imaging for differentiation of benign and atypical meningiomas

PURPOSE

To investigate the difference in amide proton transfer (APT)-weighted signals between benign and atypical meningiomas and determine the value of APT imaging for differentiating the two.

METHODS

Fifty-seven patients with pathologically diagnosed meningiomas (benign, 44; atypical, 13), who underwent preoperative MRI with APT imaging between December 2014 and August 2016 were included. We compared normalised magnetisation transfer ratio asymmetry (nMTR _asym ) values between benign and atypical meningiomas on APT-weighted images. Conventional MRI features were qualitatively assessed. Both imaging features were evaluated by multivariable logistic regression analysis. The discriminative value of MRI with and without nMTR _asym was evaluated.

RESULTS

The nMTR _asym of atypical meningiomas was significantly greater than that of benign meningiomas (2.46% vs. 1.67%; P < 0.001). In conventional MR images, benign and atypical meningiomas exhibited significant differences in maximum tumour diameter, non-skull base location, and heterogeneous enhancement. On multivariable logistic regression analysis, high nMTR _asym was an independent predictor of atypical meningiomas (adjusted OR, 11.227; P = 0.014). The diagnostic performance of MRI improved with nMTR _asym for predicting atypical meningiomas.

CONCLUSION

Atypical meningiomas exhibited significantly higher APT-weighted signal intensities than benign meningiomas. The discriminative value of conventional MRI improved significantly when combined with APT imaging for diagnosis of atypical meningioma.

KEY POINTS

• APT imaging is useful for differentiating between atypical and benign meningiomas. • Atypical meningiomas exhibited high APT-weighted signal intensity than benign meningiomas. • The diagnostic performance of MRI improved with nMTR _asym for predicting atypical meningiomas.