Conditional generative learning for medical image imputation

Image imputation refers to the task of generating a type of medical image given images of another type. This task becomes challenging when the difference between the available images, and the image to be imputed is large. In this manuscript, one such application is considered. It is derived from the dynamic contrast enhanced computed tomography (CECT) imaging of the kidneys: given an incomplete sequence of three CECT images, we are required to impute the missing image. This task is posed as one of probabilistic inference and a generative algorithm to generate samples of the imputed image, conditioned on the available images, is developed, trained, and tested. The output of this algorithm is the "best guess" of the imputed image, and a pixel-wise image of variance in the imputation. It is demonstrated that this best guess is more accurate than those generated by other, deterministic deep-learning based algorithms, including ones which utilize additional information and more complex loss terms. It is also shown that the pixel-wise variance image, which quantifies the confidence in the reconstruction, can be used to determine whether the result of the imputation meets a specified accuracy threshold and is therefore appropriate for a downstream task.

Sarcopenia and body fat change as risk factors for radiologic incisional hernia following robotic nephrectomy

OBJECTIVE

To assess the effect of body muscle and fat metrics on the development of radiologic incisional hernia (IH) following robotic nephrectomy.

MATERIALS AND METHODS

We retrospectively reviewed the records of patients who underwent robotic nephrectomy for kidney tumors between 2011 and 2017. All pre- and postoperative CTs were re-reviewed by experienced radiologists for detection of radiologic IH and calculation of the following metrics using Synapse 3D software: cross-sectional psoas muscle mass at the level of L3 and L4 as well as subcutaneous and visceral fat areas. Sarcopenia was defined as psoas muscle index below the lowest quartile. Cox proportional hazard model was constructed to examine the association between muscle and fat metrics and the risk of developing radiologic IH.

RESULTS

A total of 236 patients with a median (IQR) age of 64 (54-70) years were included in this study. In a median (IQR) follow-up of 23 (14-38) months, 62 (26%) patients developed radiologic IH. On Cox proportional hazard model, we were unable to detect an association between sarcopenia and risk of IH development. In terms of subcutaneous fat change from pre-op, both lower and higher values were associated with IH development (HR (95% CI) 2.1 (1.2-3.4), p = 0.01 and 2.4 (1.4-4.1), p < 0.01 for < Q1 and ≥ Q3, respectively). Similar trend was found for visceral fat area changes from pre-op with a HR of 2.8 for < Q1 and 1.8 for ≥ Q3.

CONCLUSION

Both excessive body fat gain and loss are associated with development of radiologic IH in patients undergoing robotic nephrectomy.

Proteogenomic Analysis of Human Uterine Cervical Cancer (UCC) Reveals Treatment-Resistant Subtypes of UCC

PURPOSE/OBJECTIVE(S)

Locally advanced uterine cervical cancer (UCC) is treated by radiotherapy with concurrent chemotherapy, but heterogenous treatment responses are frequently observed. To better optimize therapeutic options based on molecular signatures, we performed proteogenomic analysis of UCC.

MATERIALS/METHODS

UCC tissue and blood samples were collected from patients who underwent primary radiotherapy ± chemotherapy at the National Cancer Center (NCC) in Korea from July 2004 to March 2020. Most samples were obtained via biopsy. Genomic DNA for WES was isolated from frozen biopsy tumor tissues and peripheral blood buffy-coat of patients. Both global proteome and phosphoproteome were analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Whole exome sequencing, RNA sequencing, global proteomics and phosphoproteomics were performed using 253, 337, and 147 cervical cancer samples, respectively. Patient-derived xenograft were established using intratongue implantation using 2 primary cell lines from sub3 and sub5, and flow cytometric analysis was performed.

RESULTS

Fourteen significantly mutated genes (SMG) were found in our study cohort which include 5 newly identified SMGs. Mutation-phosphorylation analysis revealed association with apoptosis and actin cytoskeleton pathway. Proteogenomic analysis defined 6 molecular subtypes of UCC. Of those, 3 subtypes (i.e., Sub3, Sub5 and Sub6) were associated with treatment-resistant phenotypes. The cell-type deconvolution analysis suggested activated stroma with activation of cancer-associated fibroblast in Sub 3, while Sub5 showed low levels of activated stroma and high levels of myeloid immune cells. FACS analysis of UCC mouse models established from these 2 radio-resistant primary cell lines showed high component of PDGFRA+CAF infiltration in Sub 3, and high level of PVR+CD45+ immune cells mainly composed of polymorphonuclear myeloid-derived suppressor cells (PMN-MDSC) in Sub5. For Sub6, genes and/or protein signatures represented mucin-related processes (e.g., mucin glycosylation/sugar metabolism), which are linked to metastasis-associated Tn antigen production.

CONCLUSION

The proteogenomic analysis thus suggests potential targets for radiotherapy-resistant subtypes of UCC; secretory factors from activated stroma and cancer-associated fibroblast (Sub3); RHOA signaling, PVR, and PMN-MDSCs (Sub5), and keratin/chondroitin sulfate proteoglycan and Tn antigen production (Sub6). Our study shows the importance of proteogenomic analysis in unveiling the subtype specific molecular pathways of UCC that are beyond reach by genomic data alone. The validity of our molecular pathway and cellular signatures linking these pathways should be further validated through detailed functional experiments and in larger UCC cohorts.

Radiomics Correlation to CD68+ Tumor-Associated Macrophages in Clear Cell Renal Cell Carcinoma

INTRODUCTION

Renal cell carcinoma (RCC) is the ninth most common cancer worldwide, with clear cell RCC (ccRCC) being the most frequent histological subtype. The tumor immune microenvironment (TIME) of ccRCC is an important factor to guide treatment, but current assessments are tissue-based, which can be time-consuming and resource-intensive. In this study, we used radiomics extracted from clinically performed computed tomography (CT) as a noninvasive surrogate for CD68 tumor-associated macrophages (TAMs), a significant component of ccRCC TIME.

METHODS

TAM population was measured by CD68+/PanCK+ ratio and tumor-TAM clustering was measured by normalized K function calculated from multiplex immunofluorescence (mIF). A total of 1,076 regions on mIF slides from 78 patients were included. Radiomic features were extracted from multiphase CT of the ccRCC tumor. Statistical machine learning models, including random forest, Adaptive Boosting, and ElasticNet, were used to predict TAM population and tumor-TAM clustering.

RESULTS

The best models achieved an area under the ROC curve of 0.81 (95% CI: [0.69, 0.92]) for TAM population and 0.77 (95% CI: [0.66, 0.88]) for tumor-TAM clustering, respectively.

CONCLUSION

Our study demonstrates the potential of using CT radiomics-derived imaging markers as a surrogate for assessment of TAM in ccRCC for real-time treatment response monitoring and patient selection for targeted therapies and immunotherapies.

Radiogenomic associations clear cell renal cell carcinoma: an exploratory study

OBJECTIVES

This study investigates how quantitative texture analysis can be used to non-invasively identify novel radiogenomic correlations with Clear Cell Renal Cell Carcinoma (ccRCC) biomarkers.

METHODS

The Cancer Genome Atlas-Kidney Renal Clear Cell Carcinoma (TCGA-KIRC) open-source database was used to identify 190 sets of patient genomic data that had corresponding multiphase contrast-enhanced CT images in The Cancer Imaging Archive (TCIA-KIRC). 2824 radiomic features spanning fifteen texture families were extracted from CT images using a custom-built MATLAB software package. Robust radiomic features with strong inter-scanner reproducibility were selected. Random Forest (RF), AdaBoost, and Elastic Net machine learning (ML) algorithms evaluated the ability of the selected radiomic features to predict the presence of 12 clinically relevant molecular biomarkers identified from literature. ML analysis was repeated with cases stratified by stage (I/II vs. III/IV) and grade (1/2 vs. 3/4). 10-fold cross validation was used to evaluate model performance.

RESULTS

Before stratification by tumor grade and stage, radiomics predicted the presence of several biomarkers with weak discrimination (AUC 0.60-0.68). Once stratified, radiomics predicted KDM5C, SETD2, PBRM1, and mTOR mutation status with acceptable to excellent predictive discrimination (AUC ranges from 0.70 to 0.86).

CONCLUSIONS

Radiomic texture analysis can potentially identify a variety of clinically relevant biomarkers in patients with ccRCC and may have a prognostic implication.

Weight Loss During Topiramate Treatment in a Severely Obese Adolescent with Congenital Adrenal Hyperplasia and Migraine

Youth with classical congenital adrenal hyperplasia (CAH) due to 21-hydroxylase deficiency exhibit an increased prevalence of obesity, early adiposity rebound, and increased abdominal adiposity compared to unaffected youth. Current obesity management in CAH largely focuses on lifestyle modifications. There is evidence that topiramate therapy is effective in reducing body mass index (BMI), as well as visceral adipose tissue (VAT), in unaffected adolescents with exogenous obesity. However, little is known about the efficacy of topiramate in patients with classical CAH. We report on a 17-year-old female with severe obesity and salt-wasting CAH due to 21-hydroxylase deficiency, who demonstrated reductions in BMI, as well as abdominal visceral and subcutaneous adipose tissue (SAT) while on topiramate therapy. The patient was diagnosed with classical CAH as a newborn with a 17-hydroxyprogesterone 11,000 ng/dL. She had a BMI over the 95^th percentile at 3 years of age, followed by unremitting obesity. At 17 years old, she was started on topiramate to treat chronic migraines. Following three years of topiramate therapy, her BMI z-score decreased from +2.6 to +2.1. After four years of therapy, her waist circumference decreased from 110 to 101 cm, abdominal VAT decreased substantially by 34.2%, and abdominal SAT decreased by 25.6%. Topiramate therapy was associated with effective weight loss and reduced central adiposity in an adolescent with classical CAH and severe obesity, without any side effects. Further study is warranted regarding topiramate therapy in obese youth with classical CAH and increased central adiposity, who are at higher risk for significant morbidity.

Prasugrel-based de-escalation vs. conventional therapy after percutaneous coronary intervention in ACS patients according to the renal function

Funding Acknowledgements

Type of funding sources: Private company. Main funding source(s): A consortium of six companies in Korea (Daiichi Sankyo, Boston Scientific, Terumo, Biotronik, Qualitech Korea, and Dio).

Background

Patients with coronary artery disease and impaired renal function are at higher risk for both bleeding and ischemic adverse events after percutaneous coronary intervention (PCI).

Purpose

We assessed the efficacy and safety of a prasugrel based de-escalation strategy in patients with impaired renal function.

Methods

We conducted a post-hoc analysis of the HOST-REDUCE-POLYTECH-ACS study. Patients with available estimated glomerular filtration rate (eGFR) (n=2,311) were categorized into three groups. (high eGFR: ≥90 mL/min; intermediate eGFR: ≥60 and &lt;90 mL/min; low eGFR: &lt;60 mL/min). The endpoints were bleeding outcomes (Bleeding Academic Research Consortium type 2, or higher), ischemic outcomes (cardiovascular death, myocardial infarction, stent thrombosis, repeat revascularization, and ischemic stroke), and net adverse clinical events (all cause death, BARC 2 or greater bleeding, MI, stent thrombosis, repeat revascularization, and ischemic stroke) at 1 year follow-up. The hazard ratio (HR) and 95% Confidence interval (CI) were calculated from the multivariate Cox proportional hazard regression analysis. Covariates that were considered clinically meaningful were included. The probability risk ratio was obtained by dividing ischemic hazard function from the bleeding hazard function.

Results

With respect to net adverse clinical events, prasugrel de-escalation was beneficial regardless of baseline renal function (p for interaction = 0.508). The relative reduction in bleeding risk from prasugrel de-escalation was higher in the low eGFR group compared with that from both the intermediate and high eGFR groups (relative reduction: 64% [HR 0.36, 95% CI 0.15–0.83] vs. 50% [HR 0.50, 95% CI 0.28-0.90] and 52% [HR 0.48, 95% CI 0.21-1.13] for low, intermediate, and high eGFR groups, p for interaction=0.646). Ischemic risk from prasgurel de-escalation was not significant in all eGFR groups ([HR 1.18, 95% CI 0.47-2.98], [HR 0.95, 95% CI 0.53-1.69], and [HR 0.61, 95% CI 0.26-1.39)], respectively, p for interaction=0.119). The probability risk ratio was highest in low eGFR group (1.06 vs. 1.26 vs. 1.36, for high, intermediate, and low eGFR groups, respectively, p for trend&lt;0.001), suggesting higher relative bleeding risk above ischemic risk. Within those randomized to the de-escalation strategy, the mean probability risk ratio was not significantly different according to renal function (0.89, vs. 0.84 vs. 0.80 respectively, p for trend = 0.053), which was in contrast to those randomized to the conventional strategy where the mean probability risk ratio increased significantly as renal function decreased (1.24 vs. 1.67 vs. 1.94 respectively, p for trend&lt;0.001).

Conclusion

The beneficial effect of prasugrel-based de-escalation strategy was consistent regardless of the baseline renal function, which was mostly driven by a reduction in bleeding risk which was greatest in those with low eGFR.

Feasibility of using CT radiomic signatures for predicting CD8-T cell infiltration and PD-L1 expression in renal cell carcinoma

Objectives

To identify computed tomography (CT)-based radiomic signatures of cluster of differentiation 8 (CD8)-T cell infiltration and programmed cell death ligand 1 (PD-L1) expression levels in patients with clear-cell renal cell carcinoma (ccRCC).

Methods

Seventy-eight patients with pathologically confirmed localized ccRCC, preoperative multiphase CT and tumor resection specimens were enrolled in this retrospective study. Regions of interest (ROI) of the ccRCC volume were manually segmented from the CT images and processed using a radiomics panel comprising of 1708 metrics. The extracted metrics were used as inputs to three machine learning classifiers: Random Forest, AdaBoost, and ElasticNet to create radiomic signatures for CD8-T cell infiltration and PD-L1 expression, respectively.

Results

Using a cut-off of 80 lymphocytes per high power field, 59 % were classified to CD8 highly infiltrated tumors and 41 % were CD8 non highly infiltrated tumors, respectively. An ElasticNet classifier discriminated between these two groups of CD8-T cells with an AUC of 0.68 (95 % CI, 0.55-0.80). In addition, based on tumor proportion score with a cut-off of > 1 % tumor cells expressing PD-L1, 76 % were PD-L1 positive and 24 % were PD-L1 negative. An Adaboost classifier discriminated between PD-L1 positive and PD-L1 negative tumors with an AUC of 0.8 95 % CI: (0.66, 0.95). 3D radiomics metrics of graylevel co-occurrence matrix (GLCM) and graylevel run-length matrix (GLRLM) metrics drove the performance for CD8-Tcell and PD-L1 classification, respectively.

Conclusions

CT-radiomic signatures can differentiate tumors with high CD8-T cell infiltration with moderate accuracy and positive PD-L1 expression with good accuracy in ccRCC.

A novel index reflecting both anatomical and physiologic parameters in coronary artery disease, the FFR adjusted SYNTAX score (FaSs)

Background

Various physiology-based indices have been proposed to predict adverse clinical events in patients with coronary artery disease (CAD), such as the sum of three vessel-fractional flow reserve (3v-FFR), and the functional SYNTAX score (fSS). However, these values could not fully reflect the anatomical factors, which remains as a barrier for clinical application of these indices.

Purpose

To propose a novel index which can reflect both anatomical and physiologic features in CAD patients, and evaluate the additional predictive value for cardiovascular adverse events compared to previous indices.

Methods

For an index which can reflect both anatomical and physiologic features, we proposed the FFR adjusted SYNTAX score (FaSs). The FaSs is calculated by adding the product of the SYNTAX score and `1-FFR', for all three major coronary arteries. Among the 1136 patients who enrolled at 3V FFR-FRIENDS study, we investigated 866 patients, after excluding those who had missing variables. The 3v-FFR, fSS and FaSs were calculated, derived from the baseline FFR and SYNTAX score. Patients were divided into two groups according to the median value of each index. The primary endpoint was major adverse cardiac events (MACE, a composite of cardiac death, myocardial infarction and ischemia-driven revascularization) at 2 years follow-up.

Results

Among the total population, MACE occurred in 35 (4.04%) patients. Using the median value in a multivariable COX regression model, only FaSs was associated with an increased risk of MACE, (Hazard Ratio [HR] 5.256, 95% confidence interval [CI] 2.014–13.720), while 3v-FFR (HR 1.383, 95% CI 0.685–2.790) and fSS (HR 1.640, 95% CI 0.830–3.243) were not significantly associated with a higher risk of MACE. This was also observed in the Kaplan Meier survival curve analysis (log-rank p value: p&lt;0.001 for FaSs, 0.153 for 3v-FFR, and 0.061 for fSS; Figure 1) The sensitivity and specificity of the FaSs was 85.7% and 51.6%, which was higher compared to the 3v-FFR (62.9% and 49.3%, respectively) and fSS (57.1% and 58.5%, respectively). When these indices were combined with clinical risk factors (age, sex, hypertension, diabetes, hyperlipidemia, chronic renal failure, LVEF&lt;40%), FaSs was superior compared with 3v-FFR and fSS assessed in regards of the predictive accuracy for MACE (Figure 2).

Conclusion

The FaSs, which is a novel index calculated by a formula using the SYNTAX score and FFR, showed a superior predictive value for MACE compared to previous indices. Our results confirm the importance of considering both anatomical and physiologic parameters in evaluating the patient's risk for cardiovascular adverse outcomes.

Funding Acknowledgement

Type of funding sources: None.

Benchmarking Various Radiomic Toolkit Features While Applying the Image Biomarker Standardization Initiative toward Clinical Translation of Radiomic Analysis

The image biomarkers standardization initiative (IBSI) was formed to address the standardization of extraction of quantifiable imaging metrics. Despite its effort, there remains a lack of consensus or established guidelines regarding radiomic feature terminology, the underlying mathematics and their implementation across various software programs. This creates a scenario where features extracted using different toolboxes cannot be used to build or validate the same model leading to a non-generalization of radiomic results. In this study, IBSI-established phantom and benchmark values were used to compare the variation of the radiomic features while using 6 publicly available software programs and 1 in-house radiomics pipeline. All IBSI-standardized features (11 classes, 173 in total) were extracted. The relative differences between the extracted feature values from the different software programs and the IBSI benchmark values were calculated to measure the inter-software agreement. To better understand the variations, features are further grouped into 3 categories according to their properties: 1) morphology, 2) statistic/histogram and 3)texture features. While a good agreement was observed for a majority of radiomics features across the various tested programs, relatively poor agreement was observed for morphology features. Significant differences were also found in programs that use different gray-level discretization approaches. Since these software programs do not include all IBSI features, the level of quantitative assessment for each category was analyzed using Venn and UpSet diagrams and quantified using two ad hoc metrics. Morphology features earned lowest scores for both metrics, indicating that morphological features are not consistently evaluated among software programs. We conclude that radiomic features calculated using different software programs may not be interchangeable. Further studies are needed to standardize the workflow of radiomic feature extraction.

Clinical effectiveness of house dust mite immunotherapy in mono- versus poly-sensitised patients with allergic rhinitis: a systematic review and meta-analysis

Selecting an appropriate allergen-specific immunotherapy (AIT) regimen for polysensitised allergic rhinitis (AR) patients is challenging for clinicians. Although previous studies showed comparable effectiveness of single-allergen AIT with house dust mite (HDM) extract between monosensitised and polysensitised AR patients, there is no systematic review and meta-analysis demonstrating the comparable effectiveness of HDM AIT. In this meta-analysis, we analysed nine studies to compare the clinical effectiveness of HDM AIT. The primary outcome was nasal symptom score and secondary outcomes were medication and quality of life scores. The changes in nasal symptom score after HDM AIT did not significantly differ between monosensitised and polysensitised patients. The clinical effectiveness of HDM AIT regarding medication and quality of life score was not significantly different between monosensitised and polysensitised patients). In conclusion, single-allergen AIT with HDM extract showed comparable clinical effectiveness between polysensitised and monosensitised patients with AR.

Frame to Improve the Fit of N95 Filtering Face Mask Respirators

OBJECTIVE

Test a device that can improve upon the seal of filtering face mask respirators (FFRs).

METHODS

A 3-D prototype for a fit improvement frame (FIF) was created and quantitative fit testing was performed for FFRs with and without the FIF.

RESULTS

Thirty eight volunteers underwent fit testing. The overall fit pass rate was 100% for the 3M model 1860 masks, 50% for the 3M model 8511 masks, 13% for the BYD CARE model DE2322, and 7% for the Honeywell DC300N95. When using the FIF the overall passing rate increase to 87% for the DE2322 + FIF (P < 0.01) and for the DC300N95 + FIF the passing rate increase to 73% (P < 0.01).

CONCLUSION

The FIF is effective in improving the mask fit of a common flat fold N95 masks and potentially other N95 masks.

PNPLA3 Genotype, Arachidonic Acid Intake, and Unsaturated Fat Intake Influences Liver Fibrosis in Hispanic Youth with Obesity

Non-alcoholic fatty liver disease impacts 15.2% of Hispanic adolescents and can progress to a build-up of scared tissue called liver fibrosis. If diagnosed early, liver fibrosis may be reversible, so it is necessary to understand risk factors. The aims of this study in 59 Hispanic adolescents with obesity were to: (1) identify potential biological predictors of liver fibrosis and dietary components that influence liver fibrosis, and (2) determine if the association between dietary components and liver fibrosis differs by PNPLA3 genotype, which is highly prevalent in Hispanic adolescents and associated with elevated liver fat. We examined liver fat and fibrosis, genotyped for PNPLA3 gene, and assessed diet via 24-h diet recalls. The prevalence of increased fibrosis was 20.9% greater in males, whereas participants with the GG genotype showed 23.7% greater prevalence. Arachidonic acid was associated with liver fibrosis after accounting for sex, genotype, and liver fat (β = 0.072, p = 0.033). Intakes of several dietary types of unsaturated fat have different associations with liver fibrosis by PNPLA3 genotype after accounting for sex, caloric intake, and liver fat. These included monounsaturated fat (βCC/CG = -0.0007, βGG = 0.03, p-value = 0.004), polyunsaturated fat (βCC/CG = -0.01, βGG = 0.02, p-value = 0.01), and omega-6 (βCC/CG = -0.0102, βGG = 0.028, p-value = 0.01). Results from this study suggest that reduction of arachidonic acid and polyunsaturated fatty acid intake might be important for the prevention of non-alcoholic fatty liver disease progression, especially among those with PNPLA3 risk alleles.

Identification of robust and reproducible CT-texture metrics using a customized 3D-printed texture phantom

OBJECTIVE

The objective of this study was to evaluate the robustness and reproducibility of computed tomography-based texture analysis (CTTA) metrics extracted from CT images of a customized texture phantom built for assessing the association of texture metrics to three-dimensional (3D) printed progressively increasing textural heterogeneity.

MATERIALS AND METHODS

A custom-built 3D-printed texture phantom comprising of six texture patterns was used to evaluate the robustness and reproducibility of a radiomics panel under a variety of routine abdominal imaging protocols. The phantom was scanned on four CT scanners (Philips, Canon, GE, and Siemens) to assess reproducibility. The robustness assessment was conducted by imaging the texture phantom across different CT imaging parameters such as slice thickness, field of view (FOV), tube voltage, and tube current for each scanner. The texture panel comprised of 387 features belonging to 15 subgroups of texture extraction methods (e.g., Gray-level Co-occurrence Matrix: GLCM). Twelve unique image settings were tested on all the four scanners (e.g., FOV125). Interclass correlation two-way mixed with absolute agreement (ICC3) was used to assess the robustness and reproducibility of radiomic features. Linear regression was used to test the association between change in radiomic features and increased texture heterogeneity. Results were summarized in heat maps.

RESULTS

A total of 5612 (23.2%) of 24 090 features showed excellent robustness and reproducibility (ICC ≥ 0.9). Intensity, GLCM 3D, and gray-level run length matrix (GLRLM) 3D features showed best performance. Among imaging variables, changes in slice thickness affected all metrics more intensely compared to other imaging variables in reducing the ICC3. From the analysis of linear trend effect of the CTTA metrics, the top three metrics with high linear correlations across all scanners and scanning settings were from the GLRLM 2D/3D and discrete cosine transform (DCT) texture family.

CONCLUSION

The choice of scanner and imaging protocols affect texture metrics. Furthermore, not all CTTA metrics have a linear association with linearly varying texture patterns.

Evaluation of Knee Cartilage Diurnal, Activity, and BMI-Related Variations Using Quantitative T2 Mapping MRI and Fitbit Activity Tracking

The aim of this study is to evaluate diurnal variation in knee cartilage 3 Tesla magnetic resonance imaging (MRI) T2 mapping relaxation times, as well as activity- and body mass index (BMI)-dependent variability, using quantitative analysis of T2 values from segmented regions of the weight-bearing articular surfaces of the medial and lateral femoral condyles and tibial plateaus. Ten healthy volunteers' daily activity (steps) were tracked with Fitbit pedometers. Sagittal MRI T2 maps were obtained in the morning and afternoon on days 2 and 3. Mean T2 values were analyzed for variation related to the number of steps taken (activity), time of day (diurnal variation), and BMI using mixed effect model. Significant (albeit small) differences in the medial femoral and medial tibial cartilage regions were identified between morning and afternoon scans (diurnal variation). Daily activity did not result in significant changes and increasing BMI only demonstrated a slight increase in T2 values for the lateral tibial plateau. These findings suggest that it may be necessary to control diurnal variation when using quantitative MRI T2 mapping to assess articular cartilage longitudinally in healthy participants. Further investigation is needed to confirm these findings and determine if they also apply to symptomatic patients.

Improved Glioma Grading Using Deep Convolutional Neural Networks

BACKGROUND AND PURPOSE

Accurate determination of glioma grade leads to improved treatment planning. The criterion standard for glioma grading is invasive tissue sampling. Recently, radiomic features have shown excellent potential in glioma-grade prediction. These features may not fully exploit the underlying information in MR images. The objective of this study was to investigate the performance of features learned by a convolutional neural network compared with standard radiomic features for grade prediction.

MATERIALS AND METHODS

A total of 237 patients with gliomas were included in this study. All images were resampled, registered, skull-stripped, and segmented to extract the tumors. The learned features from the trained convolutional neural network were used for grade prediction. The performance of the proposed method was compared with standard machine learning approaches, support vector machine, random forests, and gradient boosting trained with radiomic features.

RESULTS

The experimental results demonstrate that using learned features extracted from the convolutional neural network achieves an average accuracy of 87%, outperforming the methods considering radiomic features alone. The top-performing machine learning model is gradient boosting with an average accuracy of 64%. Thus, there is a 23% improvement in accuracy, and it is an efficient technique for grade prediction.

CONCLUSIONS

Convolutional neural networks are able to learn discriminating features automatically, and these features provide added value for grading gliomas. The proposed framework may provide substantial improvement in glioma-grade prediction; however, further validation is needed.

Prognostic value of pre-revascularization fractional flow reserve mediated by the post-revascularization level: a causal mediation analysis

Background

The prognostic value of pre-percutaneous coronary intervention (PCI) fractional flow reserve (FFR) can depend on that of the post-PCI FFR and their interaction. To correctly interpret the prognostic value of pre-PCI FFR, it is essential to understand to what extent the relationship between pre-PCI FFR and clinical outcome is explained by pre-PCI FFR-related post-PCI FFR.

Purpose

The aim of this study is to investigate the extent to which post-PCI FFR mediates the relationship between pre-PCI FFR and vessel-related outcomes using a global, multicenter collaboration registry.

Methods

Patient data from 4 global FFR registries were pooled and 1488 patients with pre-PCI FFR ≤0.80 were analyzed. The primary outcome was target vessel failure (TVF) during 2-years of follow-up. We evaluated the extent to which post-PCI FFR &lt;0.90 mediated the association between pre-PCI FFR &lt;0.75 and TVF employing a causal mediation analysis in a counterfactual framework.

Results

Among 1488 patients, the mean (standard deviation) age was 63.5 (9.9) years and 78% (1161 patients) were male. The median (IQR) pre-PCI and post-PCI FFR were 0.71 (0.62–0.76) and 0.88 (0.83–0.92), respectively. The direct effect of low pre-PCI FFR (&lt;0.75) on TVF was significantly elevated (OR: 1.81, 95% CI: 1.03–3.18, p=0.038), and was not mediated by post-PCI FFR&lt;0.90 (indirect effect, OR: 1.01, 95% CI: 0.98–1.05, p=0.39). In the model, post-PCI FFR explained only 2.2% of the association between pre-PCI FFR and TVF. The subgroup analysis implicated that the prognostic information of pre-PCI FFR was mainly for diffuse lesions.

Conclusions

The prognostic information of pre-PCI FFR did not greatly depend on the results of PCI assessed by post-PCI FFR. Pre-PCI FFR, as a prognostic marker, may mainly reflect the global atherosclerotic burden of the artery, not the extent of the modifiable epicardial stenosis, thus providing independent information from post-PCI FFR.

Interpretation

Funding Acknowledgement

Type of funding source: None

Radiomics and Bladder Cancer: Current Status

PURPOSE: To systematically review the current literature and discuss the applications and limitations of radiomics and machine-learning augmented radiomics in the management of bladder cancer. METHODS: Pubmed ®, Scopus ®, and Web of Science ® databases were searched systematically for all full-text English-language articles assessing the impact of Artificial Intelligence OR Radiomics OR Machine Learning AND Bladder Cancer AND (staging OR grading OR prognosis) published up to January 2020. RESULTS: Of the 686 articles that were identified, 13 studies met the criteria for quantitative analysis. Staging, Grading and Tumor Classification, Prognosis, and Therapy Response were discussed in 7, 3, 2 and 7 studies, respectively. Data on cost of implementation were not reported. CT and MRI were the most common imaging approaches. CONCLUSION: Radiomics shows potential in bladder cancer detection, staging, grading, and response to therapy, thereby supporting the physician in personalizing patient management. Extension and validation of this promising technology in large multisite prospective trials is warranted to pave the way for its clinical translation.

Deep learning based classification of solid lipid-poor contrast enhancing renal masses using contrast enhanced CT

OBJECTIVE

Establish a workflow that utilizes convolutional neural nets (CNN) to classify solid, lipid-poor, contrast enhancing renal masses using multiphase contrast enhanced CT (CECT) images and to assess the performance of the resulting network.

METHODS

In this institutional review board approved study of 143 patients with predominantly solid, lipid-poor, contrast enhancing renal lesions (46 benign and 97 malignant), patients with a pre-operative multiphase CECT of the abdomen and pelvis obtained between June 2009 and June 2015 were retrospectively queried. Benign renal masses included oncocytoma and lipid-poor angiomyolipoma and the malignant group included clear cell, papillary, and chromophobe carcinomas.Region of interests of whole tumor volumes were manually segmented, and CT phase images with the largest cross-section of the segmented tumor in the axial plane were used for assessment. Post-surgical pathological evaluation was used to establish diagnosis.The segmented images of renal masses were used as input to a CNN. The data were augmented and split into training (83.9%) and validation sets (16.1%) to determine the hyperparameters of the CNN. Thereafter. the performance of the resulting CNN was quantified using eightfold cross-validation.

RESULTS

The CNN-based classifier demonstrated an overall accuracy of 78% (95% confidence interval: 76-80%), sensitivity of 70% (95% confidence interval: 66-74%), specificity of 81% (79-83%) and an area under the curve of 0.82.

CONCLUSION

A CNN-based classifier to diagnose solid enhancing malignant renal masses based on multiphase CECT images was developed.

ADVANCES IN KNOWLEDGE

It was established that a CNN-based classifier could be trained to accurately distinguish malignant renal lesions.

An 82-year-old female with chest pain radiating to the back and flank

An 82 year-old female presents with severe, sharp right-sided chest, shoulder and flank pain, worse with deep inspiration. Cardiac workup is negative for myocardial ischemia or aortic dissection. Computed tomography demonstrates a right posterior diaphragmatic (Bochdalek) hernia, involving the right extra-renal pelvis and proximal ureter, with an associated delayed nephrogram and mild hydronephrosis. There is no obstructing nephrolithiasis. The patient is taken to the operating room and right-sided double-J ureteral stent is placed (Fig. 1), with immediate resolution of pain. Diaphragmatic hernia repair in coordination with urology and general surgery is scheduled.

Low-tech solutions for the COVID-19 supply chain crisis

A global effort is ongoing in the scientific community and in the maker movement, which focuses on creating devices and tinkering with them, to reverse-engineer commercial medical equipment and get it to healthcare workers. For these 'low-tech' solutions to have a real impact, it is important for them to coalesce around approved designs.

Physicochemical Investigation into Major League Baseballs in the Era of Unprecedented Rise in Home Runs

A major attraction of baseball is the home run. Throughout baseball's history, some seasonal increases in home run numbers have been tied to external influences, such as lowering the pitching mound or a change in the ball manufacturer. In Major League Baseball, a recent surge in home runs has led to speculation about the baseball being "juiced" or altered in a way to make it fly farther. To support multiple academic and journalistic studies, which have attempted to find evidence of the changes in the flight of the baseball, a systematic chemical analysis has been reported on the multicomponent baseballs. Thus, we undertook a study where we analyzed the core of the baseball using various chemical and physical techniques. Studies using computed tomography scans revealed that there is a drastic 56.7% difference in the density of the core of the baseballs used during the 2014 and pre-All-Star Game 2015 versus 2017 season in the Major League. Increased material porosity was observed using electron microscopy. Thermogravimetric and elemental analyses of the pill material showed a 7% difference in the ratio of organic to inorganic material and almost a 10% decrease in silicon, respectively. Overall, the data indicates a difference in the core of the baseballs between the two time periods, leaving the contribution of these measured differences in the explosion of home runs open to interpretation.

Reliability of CT-based texture features: Phantom study

Objective

To determine the intra‐, inter‐ and test‐retest variability of CT‐based texture analysis (CTTA) metrics.

Materials and methods

In this study, we conducted a series of CT imaging experiments using a texture phantom to evaluate the performance of a CTTA panel on routine abdominal imaging protocols. The phantom comprises of three different regions with various textures found in tumors. The phantom was scanned on two CT scanners viz. the Philips Brilliance 64 CT and Toshiba Aquilion Prime 160 CT scanners. The intra‐scanner variability of the CTTA metrics was evaluated across imaging parameters such as slice thickness, field of view, post‐reconstruction filtering, tube voltage, and tube current. For each scanner and scanning parameter combination, we evaluated the performance of eight different types of texture quantification techniques on a predetermined region of interest (ROI) within the phantom image using 235 different texture metrics. We conducted the repeatability (test‐retest) and robustness (intra‐scanner) test on both the scanners and the reproducibility test was conducted by comparing the inter‐scanner differences in the repeatability and robustness to identify reliable CTTA metrics. Reliable metrics are those metrics that are repeatable, reproducible and robust.

Results

As expected, the robustness, repeatability and reproducibility of CTTA metrics are variably sensitive to various scanner and scanning parameters. Entropy of Fast Fourier Transform‐based texture metrics was overall most reliable across the two scanners and scanning conditions. Post‐processing techniques that reduce image noise while preserving the underlying edges associated with true anatomy or pathology bring about significant differences in radiomic reliability compared to when they were not used.

Conclusion

Following large‐scale validation, identification of reliable CTTA metrics can aid in conducting large‐scale multicenter CTTA analysis using sample sets acquired using different imaging protocols, scanners etc.

Comprehensive radiogenomics analysis of qualitative and quantitative features of cross-sectional imaging in the TCGA project in MIBC

482

Background: Quantitative imaging descriptors derived from CT and MRI can be integrated with genomic data that may be used as non-invasive prognostic or predictive biomarkers. We report an integrated radiogenomics project designed to develop subjective and objective parameters extracted from cross-sectional imaging of MIBC from studies archived in the TCIA and linked to the TCGA project. Methods: We reported comprehensive integrated genomic analysis of 412 tumors (Cell 2017). 7 of 33 tissue source sites submitted CT scans to the TCIA (n=106). We developed 17 features describing tumor size/location, metastases sites, and tumor morphology; 9 GU radiologists reviewed the scans in a blinded manner. EH analyzed the data independent of the radiologists. We computed kappa statistics for categorical features and coverage probabilities for quantitative features (Lin et al 2002). The tumor was segmented on an axial image and the segmented image analyzed using a radiomics panel (radiomicslab.usc.edu). Associations between individual features and subtypes were assessed (Fisher’s Exact Test) for categorical features and Kruskal-Wallis Test for quantitative features. Results: Substantial agreement (k≥ 0.6) was observed in 4 features: tumor laterality, tumor within bladder diverticulum, right and left UVJ involvement and hydroureter. We observed weak agreement (95% CI <0.4) for bladder neck, posterior bladder, dome, and trigone involvement, tumor margin, internal architecture, radiographic stage, left upper tract involvement, and metastases. The coverage probability for lesion size was 0.59 (0.544-0.638) (Figure). Tumor morphology was associated with microRNA cluster, with diffuse wall thickening having a higher tendency toward Clusters 3 and 4 (p < .001). Radiomic analysis identified statistically significant associations of mutations in FGFR3, CREBBP, CASP8 and EP300 with multiple radiomic features. Conclusions: This blinded comprehensive assessment of features extracted from CT images highlights many of the ongoing challenges in staging patients with MIBC. Preliminary analysis shows promise in analyzing associations between radiomic features and mutations.