In Situ Photo-Crosslinkable Protein Bioadhesive for Bone Graft Fixation

Bone grafting is a fundamental dental surgery procedure widely used for implant placement and periodontal disease management treatments. Despite its broad applications, vertical bone augmentation presents unique challenges, including the risk of graft displacement due to gravitational and masticatory forces. Traditional physical stabilization methods introduce additional complexities and risks, underscoring the need for innovative fixation technologies. This study aimed to develop an in situ photo-crosslinkable bioadhesive hydrogel (iPBAH) as a multifunctional bone graft binder to enhance the process of bone reconstruction. The bioadhesive is composed of mussel-derived adhesive protein (MAP) fused with the cell-adhesive peptide RGD. The numerous tyrosine residues in MAP facilitate rapid photo-crosslinking, enabling efficient hydrogel formation using visible blue light. Subsequently, iPBAH underwent comprehensive characterization to evaluate its suitability as a multifunctional bone graft binder. iPBAH efficiently underwent in situ crosslinking through harmless exposure to visible light within minutes and displayed several exceptional properties, including a microporous structure, underwater adhesion, extended durability, high compressive strength, and biocompatibility. In vivo assessments, using male Sprague-Dawley rats, demonstrated that iPBAH binder significantly enhanced bone regeneration in a rat calvarial bone defect model. The in situ crosslinking of the iPBAH binder during bone graft transplantation can effectively fill irregular and complex defect shapes while simultaneously preventing graft material leakage. The improved physical attributes of the bound graft material can enhance its resistance to external forces, thereby ensuring sustained retention over time. Moreover, the interaction between iPBAH and surrounding tissues promotes adhesion and integration of the graft material with host tissues in the defect area. In addition, the included RGD peptide in iPBAH can augment inherent cell recruitment, adhesion, and growth, consequently expediting osteogenesis.

Evaluation of retrieval accuracy and visual similarity in content-based image retrieval of chest CT for obstructive lung disease

The aim of our study was to assess the performance of content-based image retrieval (CBIR) for similar chest computed tomography (CT) in obstructive lung disease. This retrospective study included patients with obstructive lung disease who underwent volumetric chest CT scans. The CBIR database included 600 chest CT scans from 541 patients. To assess the system performance, follow-up chest CT scans of 50 patients were evaluated as query cases, which showed the stability of the CT findings between baseline and follow-up chest CT, as confirmed by thoracic radiologists. The CBIR system retrieved the top five similar CT scans for each query case from the database by quantifying and comparing emphysema extent and size, airway wall thickness, and peripheral pulmonary vasculatures in descending order from the database. The rates of retrieval of the same pairs of query CT scans in the top 1-5 retrievals were assessed. Two expert chest radiologists evaluated the visual similarities between the query and retrieved CT scans using a five-point scale grading system. The rates of retrieving the same pairs of query CTs were 60.0% (30/50) and 68.0% (34/50) for top-three and top-five retrievals. Radiologists rated 64.8% (95% confidence interval 58.8-70.4) of the retrieved CT scans with a visual similarity score of four or five and at least one case scored five points in 74% (74/100) of all query cases. The proposed CBIR system for obstructive lung disease integrating quantitative CT measures demonstrated potential for retrieving chest CT scans with similar imaging phenotypes. Further refinement and validation in this field would be valuable.

Machine learning-based clinical decision support system for treatment recommendation and overall survival prediction of hepatocellular carcinoma: a multi-center study

The treatment decisions for patients with hepatocellular carcinoma are determined by a wide range of factors, and there is a significant difference between the recommendations of widely used staging systems and the actual initial treatment choices. Herein, we propose a machine learning-based clinical decision support system suitable for use in multi-center settings. We collected data from nine institutions in South Korea for training and validation datasets. The internal and external datasets included 935 and 1750 patients, respectively. We developed a model with 20 clinical variables consisting of two stages: the first stage which recommends initial treatment using an ensemble voting machine, and the second stage, which predicts post-treatment survival using a random survival forest algorithm. We derived the first and second treatment options from the results with the highest and the second-highest probabilities given by the ensemble model and predicted their post-treatment survival. When only the first treatment option was accepted, the mean accuracy of treatment recommendation in the internal and external datasets was 67.27% and 55.34%, respectively. The accuracy increased to 87.27% and 86.06%, respectively, when the second option was included as the correct answer. Harrell's C index, integrated time-dependent AUC curve, and integrated Brier score of survival prediction in the internal and external datasets were 0.8381 and 0.7767, 91.89 and 86.48, 0.12, and 0.14, respectively. The proposed system can assist physicians by providing data-driven predictions for reference from other larger institutions or other physicians within the same institution when making treatment decisions.

Immature bovine cartilage wear is due to fatigue failure from repetitive compressive forces and not reciprocating frictional forces

OBJECTIVE

Wear of articular cartilage is not well understood. We hypothesize that cartilage wears due to fatigue failure in repetitive compression instead of reciprocating friction.

DESIGN

This study compares reciprocating sliding of immature bovine articular cartilage against glass in two testing configurations: (1) a stationary contact area configuration (SCA), which results in static compression, interstitial fluid depressurization, and increasing friction coefficient during reciprocating sliding, and (2) a migrating contact area configuration (MCA), which maintains pressurization and low friction while producing repetitive compressive loading in addition to reciprocating sliding. Contact pressure, sliding duration, and sliding distance were controlled to be similar between test groups.

RESULTS

SCA tests exhibited an average friction coefficient of μ=0.084±0.032, while MCA tests exhibited a lower average friction coefficient of μ=0.020±0.008 (p<10-4). Despite the lower friction, MCA cartilage samples exhibited clear surface damage with a significantly greater average surface deviation from a fitted plane after wear testing (Rq=0.125±0.095 mm) than cartilage samples slid in a SCA configuration (Rq=0.044±0.017 mm, p=0.002), which showed minimal signs of wear. Polarized light microscopy confirmed that delamination damage occurred between the superficial and middle zones of the articular cartilage in MCA samples.

CONCLUSIONS

The greatest wear was observed in the group with lowest friction coefficient, subjected to cyclical instead of static compression, implying that friction is not the primary driver of cartilage wear. Delamination between superficial and middle zones implies the main mode of wear is fatigue failure under cyclical compression, not fatigue or abrasion due to reciprocating frictional sliding.

Technical Feasibility of Quantitative Measurement of Various Degrees of Small Bowel Motility Using Cine Magnetic Resonance Imaging

OBJECTIVE

Cine magnetic resonance imaging (MRI) has emerged as a noninvasive method to quantitatively assess bowel motility. However, its accuracy in measuring various degrees of small bowel motility has not been extensively evaluated. We aimed to draw a quantitative small bowel motility score from cine MRI and evaluate its performance in a population with varying degrees of small bowel motility.

MATERIALS AND METHODS

A total of 174 participants (28.5 ± 7.6 years; 135 males) underwent a 22-second-long cine MRI sequence (2-dimensional balanced turbo-field echo; 0.5 seconds per image) approximately 5 minutes after being intravenously administered 10 mg of scopolamine-N-butyl bromide to deliberately create diverse degrees of small bowel motility. In a manually segmented area of the small bowel, motility was automatically quantified using a nonrigid registration and calculated as a quantitative motility score. The mean value (MV) of motility grades visually assessed by two radiologists was used as a reference standard. The quantitative motility score's correlation (Spearman's ρ) with the reference standard and performance (area under the receiver operating characteristics curve [AUROC], sensitivity, and specificity) for diagnosing adynamic small bowel (MV of 1) were evaluated.

RESULTS

For the MV of the quantitative motility scores at grades 1, 1.5, 2, 2.5, and 3, the mean ± standard deviation values were 0.019 ± 0.003, 0.027 ± 0.010, 0.033 ± 0.008, 0.032 ± 0.009, and 0.043 ± 0.013, respectively. There was a significant positive correlation between the quantitative motility score and the MV (ρ = 0.531, P < 0.001). The AUROC value for diagnosing a MV of 1 (i.e., adynamic small bowel) was 0.953 (95% confidence interval, 0.923-0.984). Moreover, the optimal cutoff for the quantitative motility score was 0.024, with a sensitivity of 100% (15/15) and specificity of 89.9% (143/159).

CONCLUSION

The quantitative motility score calculated from a cine MRI enables diagnosis of an adynamic small bowel, and potentially discerns various degrees of bowel motility.

Deep Learning for Automated Triaging of Stable Chest Radiographs in a Follow-up Setting

Background Most artificial intelligence algorithms that interpret chest radiographs are restricted to an image from a single time point. However, in clinical practice, multiple radiographs are used for longitudinal follow-up, especially in intensive care units (ICUs). Purpose To develop and validate a deep learning algorithm using thoracic cage registration and subtraction to triage pairs of chest radiographs showing no change by using longitudinal follow-up data. Materials and Methods A deep learning algorithm was retrospectively developed using baseline and follow-up chest radiographs in adults from January 2011 to December 2018 at a tertiary referral hospital. Two thoracic radiologists reviewed randomly selected pairs of "change" and "no change" images to establish the ground truth, including normal or abnormal status. Algorithm performance was evaluated using area under the receiver operating characteristic curve (AUC) analysis in a validation set and temporally separated internal test sets (January 2019 to August 2021) from the emergency department (ED) and ICU. Threshold calibration for the test sets was conducted, and performance with 40% and 60% triage thresholds was assessed. Results This study included 3 304 996 chest radiographs in 329 036 patients (mean age, 59 years ± 14 [SD]; 170 433 male patients). The training set included 550 779 pairs of radiographs. The validation set included 1620 pairs (810 no change, 810 change). The test sets included 533 pairs (ED; 265 no change, 268 change) and 600 pairs (ICU; 310 no change, 290 change). The algorithm had AUCs of 0.77 (validation), 0.80 (ED), and 0.80 (ICU). With a 40% triage threshold, specificity was 88.4% (237 of 268 pairs) and 90.0% (261 of 290 pairs) in the ED and ICU, respectively. With a 60% triage threshold, specificity was 79.9% (214 of 268 pairs) and 79.3% (230 of 290 pairs) in the ED and ICU, respectively. For urgent findings (consolidation, pleural effusion, pneumothorax), specificity was 78.6%-100% (ED) and 85.5%-93.9% (ICU) with a 40% triage threshold. Conclusion The deep learning algorithm could triage pairs of chest radiographs showing no change while detecting urgent interval changes during longitudinal follow-up. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Czum in this issue.

Significance of Radiomics in Predicting Local Control for Patients with Malignant Liver Tumors Treated Using Stereotactic Body Radiotherapy

PURPOSE/OBJECTIVE(S)

We routinely deliver Stereotactic Body Radiotherapy (SBRT) in malignant liver tumors using planning computed tomography (CT) and Magnetic Resonance Images (MRI) to aid target definition. In this study, we extracted radiomic features from the MR images to predict local control (LC) post-SBRT.

MATERIALS/METHODS

We retrospectively analyzed patients with either hepatocellular cancers (HCCs) or liver metastases (Mets) treated with SBRT between Aug 2014 and Aug 2020. All patients had CT simulation followed by 1.5 Tesla planning MRI in treatment position. Contrast enhanced T1 VIBE and T2 Haste MR sequences were registered with planning CT for target definition. Radiomic features were extracted from Gross Tumor Volumes (GTV) masked out of 60 seconds post contrast T1 VIBE MR images using the Radiomics calculator tool RaCaT. The output included 480 (408 textural, 50 intensity and 22 morphological) features for each target. Principal Component Analysis of the outputs obtained from all the targets yielded 20 radiomic feature clusters after computational prioritization. These clusters were correlated to LC outcomes at various time points post-SBRT. LC was defined as non-progressive disease. Accuracy of predictions was measured by area under (AUC) receiver operating characteristic curve. Cox regression analysis was done to find univariate and multivariate clinical [HCCs vs. Mets, single vs. multiple lesions, previous local therapy (yes vs. no), GTV volume (≤40 vs. >40 cc)], radiomic and dosimetric predictors (continuous) of LC.

RESULTS

In total, 97 patients received SBRT to 122 lesions. The median dose prescribed was 45 Gy (range, 30-50 Gy). Median age was 69 years (interquartile range, IQR 61-73 yrs.). 59 patients had HCCs and 38 had Mets. 24 lesions had prior ablative therapy. 75 patients had one target, and 22 had multiple targets. Median GTV was 43.5 cc (IQR 23.4-78.6 cc). Median follow up was 16.6 months (IQR 9.7-27.2 mths). Median LC was 13.6 months (IQR 8.0-23.5 mths). On univariate analysis, histology (HCCs vs. Mets; Hazard ratio (HR) 2.9, 95% CI 1.4-6.4; p < 0.006), radiomic clusters (p < 0.006) and the max., mean, and min. doses to GTV and Planning Target Volumes correlated with improved LC (all p-values < 0.05). On multivariate analysis, histology (HCCs vs. Mets; HR 4.4, 95% CI 1.6-12.3; p = 0.004), radiomic clusters (p = 0.034) and prescription dose (p = 0.048) were significant covariates. Specifically, the 20 radiomic clusters were predictive of LC, and the accuracy of predictions showed promise with AUC values of 0.74, 0.80, and 0.81 at 12, 24, and 36 months post-SBRT, respectively. AUC values for LC in HCCs vs. Mets at 12, 24, and 36 months were 0.83, 0.77, and 0.70, and 0.66, 0.77, and 0.88, respectively.

CONCLUSION

MR-based radiomics predict LC post-SBRT in patients with malignant liver tumors. Further research focused on independent validation of the model is required to explore its clinical use.

MuSiC-ViT: A multi-task Siamese convolutional vision transformer for differentiating change from no-change in follow-up chest radiographs

A major responsibility of radiologists in routine clinical practice is to read follow-up chest radiographs (CXRs) to identify changes in a patient's condition. Diagnosing meaningful changes in follow-up CXRs is challenging because radiologists must differentiate disease changes from natural or benign variations. Here, we suggest using a multi-task Siamese convolutional vision transformer (MuSiC-ViT) with an anatomy-matching module (AMM) to mimic the radiologist's cognitive process for differentiating baseline change from no-change. MuSiC-ViT uses the convolutional neural networks (CNNs) meet vision transformers model that combines CNN and transformer architecture. It has three major components: a Siamese network architecture, an AMM, and multi-task learning. Because the input is a pair of CXRs, a Siamese network was adopted for the encoder. The AMM is an attention module that focuses on related regions in the CXR pairs. To mimic a radiologist's cognitive process, MuSiC-ViT was trained using multi-task learning, normal/abnormal and change/no-change classification, and anatomy-matching. Among 406 K CXRs studied, 88 K change and 115 K no-change pairs were acquired for the training dataset. The internal validation dataset consisted of 1,620 pairs. To demonstrate the robustness of MuSiC-ViT, we verified the results with two other validation datasets. MuSiC-ViT respectively achieved accuracies and area under the receiver operating characteristic curves of 0.728 and 0.797 on the internal validation dataset, 0.614 and 0.784 on the first external validation dataset, and 0.745 and 0.858 on a second temporally separated validation dataset. All code is available at https://github.com/chokyungjin/MuSiC-ViT.

Generative Adversarial Network-Based Image Conversion Among Different Computed Tomography Protocols and Vendors: Effects on Accuracy and Variability in Quantifying Regional Disease Patterns of Interstitial Lung Disease

OBJECTIVE

To assess whether computed tomography (CT) conversion across different scan parameters and manufacturers using a routable generative adversarial network (RouteGAN) can improve the accuracy and variability in quantifying interstitial lung disease (ILD) using a deep learning-based automated software.

MATERIALS AND METHODS

This study included patients with ILD who underwent thin-section CT. Unmatched CT images obtained using scanners from four manufacturers (vendors A-D), standard- or low-radiation doses, and sharp or medium kernels were classified into groups 1-7 according to acquisition conditions. CT images in groups 2-7 were converted into the target CT style (Group 1: vendor A, standard dose, and sharp kernel) using a RouteGAN. ILD was quantified on original and converted CT images using a deep learning-based software (Aview, Coreline Soft). The accuracy of quantification was analyzed using the dice similarity coefficient (DSC) and pixel-wise overlap accuracy metrics against manual quantification by a radiologist. Five radiologists evaluated quantification accuracy using a 10-point visual scoring system.

RESULTS

Three hundred and fifty CT slices from 150 patients (mean age: 67.6 ± 10.7 years; 56 females) were included. The overlap accuracies for quantifying total abnormalities in groups 2-7 improved after CT conversion (original vs. converted: 0.63 vs. 0.68 for DSC, 0.66 vs. 0.70 for pixel-wise recall, and 0.68 vs. 0.73 for pixel-wise precision; P < 0.002 for all). The DSCs of fibrosis score, honeycombing, and reticulation significantly increased after CT conversion (0.32 vs. 0.64, 0.19 vs. 0.47, and 0.23 vs. 0.54, P < 0.002 for all), whereas those of ground-glass opacity, consolidation, and emphysema did not change significantly or decreased slightly. The radiologists' scores were significantly higher (P < 0.001) and less variable on converted CT.

CONCLUSION

CT conversion using a RouteGAN can improve the accuracy and variability of CT images obtained using different scan parameters and manufacturers in deep learning-based quantification of ILD.

Automated detection of intracranial aneurysms using skeleton-based 3D patches, semantic segmentation, and auxiliary classification for overcoming data imbalance in brain TOF-MRA

Accurate and reliable detection of intracranial aneurysms is vital for subsequent treatment to prevent bleeding. However, the detection of intracranial aneurysms can be time-consuming and even challenging, and there is great variability among experts, especially in the case of small aneurysms. This study aimed to detect intracranial aneurysms accurately using a convolutional neural network (CNN) with 3D time-of-flight magnetic resonance angiography (TOF-MRA). A total of 154 3D TOF-MRA datasets with intracranial aneurysms were acquired, and the gold standards were manually drawn by neuroradiologists. We also obtained 113 subjects from a public dataset for external validation. These angiograms were pre-processed by using skull-stripping, signal intensity normalization, and N4 bias correction. The 3D patches along the vessel skeleton from MRA were extracted. Values of the ratio between the aneurysmal and the normal patches ranged from 1:1 to 1:5. The semantic segmentation on intracranial aneurysms was trained using a 3D U-Net with an auxiliary classifier to overcome the imbalance in patches. The proposed method achieved an accuracy of 0.910 in internal validation and external validation accuracy of 0.883 with a 2:1 ratio of normal to aneurysmal patches. This multi-task learning method showed that the aneurysm segmentation performance was sufficient to be helpful in an actual clinical setting.

Deep Learning of Time-Signal Intensity Curves from Dynamic Susceptibility Contrast Imaging Enables Tissue Labeling and Prediction of Survival in Glioblastoma

BACKGROUND AND PURPOSE

An autoencoder can learn representative time-signal intensity patterns to provide tissue heterogeneity measures using dynamic susceptibility contrast MR imaging. The aim of this study was to investigate whether such an autoencoder-based pattern analysis could provide interpretable tissue labeling and prognostic value in isocitrate dehydrogenase (IDH) wild-type glioblastoma.

MATERIALS AND METHODS

Preoperative dynamic susceptibility contrast MR images were obtained from 272 patients with IDH wild-type glioblastoma (training and validation, 183 and 89 patients, respectively). The autoencoder was applied to the dynamic susceptibility contrast MR imaging time-signal intensity curves of tumor and peritumoral areas. Representative perfusion patterns were defined by voxelwise K-means clustering using autoencoder latent features. Perfusion patterns were labeled by comparing parameters with anatomic reference tissues for baseline, signal drop, and percentage recovery. In the validation set (n = 89), a survival model was created from representative patterns and clinical predictors using Cox proportional hazard regression analysis, and its performance was calculated using the Harrell C-index.

RESULTS

Eighty-nine patients were enrolled. Five representative perfusion patterns were used to characterize tissues as high angiogenic tumor, low angiogenic/cellular tumor, perinecrotic lesion, infiltrated edema, and vasogenic edema. Of these, the low angiogenic/cellular tumor (hazard ratio, 2.18; P = .047) and infiltrated edema patterns (hazard ratio, 1.88; P = .009) in peritumoral areas showed significant prognostic value. The combined perfusion patterns and clinical predictors (C-index, 0.72) improved prognostication when added to clinical predictors (C-index, 0.55).

CONCLUSIONS

The autoencoder perfusion pattern analysis enabled tissue characterization of peritumoral areas, providing heterogeneity and dynamic information that may provide useful prognostic information in IDH wild-type glioblastoma.

[The status and influencing factors of presenteeism among clinical nurses: a systematic review]

Objective: To systematically review the status and factors influencing presenteeism among clinical nurses. Methods: In December 2021, CNKI, CBM, Wanfang, VIP, Web of Science, PubMed, Embase, The Cochrane Library, CINAHL, PsyclNFO and other databases were electronically searched to cross sectional studies on the current situation and factors influencing the occurrence of presenteeism among clinical nurses. The search terms mainly included presenteeism, sick at work, Stanford Presenteeism Scale, nurse, level, risk factor, influence, et al. And the search time was from the establishment of the database to November 30, 2021. Literature screening, data extraction and evaluation of the risk of bias in the included literature were done independently by two researchers, and meta-analysis was performed using Stata 15.1 software. Results: A total of 29 studies involving 13 535 clinical nurses were included.The results of the meta-analysis showed that the score of presenteeism was 17.99 [95% CI (17.02-18.95), P =0.000]. Subgroup analysis showed that presenteeism scores were higher in articles published before 2020 (ES=19.28, 95%CI: 18.41-20.15, P=0.000) and in the group of nurses aged 36 to 40 years (ES=19.27, 95%CI: 17.35~21.19, P=0.000), female (ES= 17.04, 95%CI: 14.70-19.39, P=0.000), secondary school education (ES=21.01, 95%CI: 17.76-24.26, P= 0.007), married (ES=17.49, 95%CI: 15.13-19.85, P=0.000), working for 5 to 10 years (ES=17.78, 95%CI: 16.54-19.02, P=0.000), contract (ES=17.05, 95%CI: 15.23-18.87, P=0.000), working in pediatrics (ES= 16.65, 95% CI: 15.31-17.99, P=0.000) and European region (ES =21.21, 95% CI: 20.50-21.93, P=0.000) . Conclusion: Current evidence suggests that clinical nurses are at high risk of presenteeism, which is affected by variety of factors. The managers should pay attention to the physical and mental health of nurses, identify high-risk factors as early as possible and take measures to reduce the occurrence of presenteeism and improve the quality of nursing.

Abstract 273: Synergistically suppressive effects by combination of glycolytic inhibitor with TRAP1-depletion in colorectal cancer

Tumor necrosis factor receptor-associated protein 1 (TRAP1) is a mitochondrial homolog of HSP90 chaperones. It plays an important role in protection against oxidative stress and apoptosis by regulating reactive oxidative species (ROS). Recent studies have suggested TRAP1 as a regulator of metabolism and mitochondrial homeostasis in diverse pathological states, including cancer. To further elucidate the mechanistic role of TRAP1 in regulating tumor cell survival, we used gamitrinib-triphenylphosphonium (G-TPP) to inhibit TRAP1 signaling pathways in colon cancer. Inhibition of TRAP1 disrupted redox homeostasis and increased ER stress in colon cancer. To confirm the G-TPP effects of TRAP1 inhibition in colon cancer, we performed TRAP1 CRISPR/Cas9 to deplete TRAP1 expression in colon cancer cell line CT26. TRAP1-depleted cells were found to acquire several distinct phenotypes: induction of reactive oxidative species (ROS) level, reduction of cell growth, and lower pH value of culture medium. In addition, proteomic data showed that depletion of TRAP1 induced significant metabolism alteration in CT26 cells. Using seahorse assay to access mitochondrial respiration and glycolysis, we found that TRAP1-depleted cells were insensitive to Oxidative Phosphorylation (OXPHOS) inhibitor treatment and significantly increased extracellular acidification rate (ECAR) compared to parental cells. Furthermore, higher glucose consumption rate and higher lactic acid level of culture medium were observed in TRAP1-depleted cells. On the other hand, Glycolytic inhibitor (2DG) led to excessive ROS production, reduction of glucose uptake and exacerbated cell death in TRAP1-depleted cell compared to parental cells. Taken together, depletion of TRAP1 induces mitochondrial dysfunction and modulates metabolic switch toward aerobic glycolysis in TRAP1-depleted colon cancer cells. Dual inhibition of TRAP1 and glycolytic signaling may increase therapeutic efficiency through induction of excessive ROS production in colon cancer.

Citation Format: Hong-Yuan Tsai, Miao-Hsueh Chen, Herath Lakmini Senavirathna, Jihye Yun, Sheng Pan, Ru Chen. Synergistically suppressive effects by combination of glycolytic inhibitor with TRAP1-depletion in colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 273.

Deep Learning-based Approach to Predict Pulmonary Function at Chest CT

Background Low-dose chest CT screening is recommended for smokers with the potential for lung function abnormality, but its role in predicting lung function remains unclear. Purpose To develop a deep learning algorithm to predict pulmonary function with low-dose CT images in participants using health screening services. Materials and Methods In this retrospective study, participants underwent health screening with same-day low-dose CT and pulmonary function testing with spirometry at a university affiliated tertiary referral general hospital between January 2015 and December 2018. The data set was split into a development set (model training, validation, and internal test sets) and temporally independent test set according to first visit year. A convolutional neural network was trained to predict the forced expiratory volume in the first second of expiration (FEV₁) and forced vital capacity (FVC) from low-dose CT. The mean absolute error and concordance correlation coefficient (CCC) were used to evaluate agreement between spirometry as the reference standard and deep-learning prediction as the index test. FVC and FEV₁ percent predicted (hereafter, FVC% and FEV₁%) values less than 80% and percent of FVC exhaled in first second (hereafter, FEV₁/FVC) less than 70% were used to classify participants at high risk. Results A total of 16 148 participants were included (mean age, 55 years ± 10 [SD]; 10 981 men) and divided into a development set (n = 13 428) and temporally independent test set (n = 2720). In the temporally independent test set, the mean absolute error and CCC were 0.22 L and 0.94, respectively, for FVC and 0.22 L and 0.91 for FEV₁. For the prediction of the respiratory high-risk group, FVC%, FEV₁%, and FEV₁/FVC had respective accuracies of 89.6% (2436 of 2720 participants; 95% CI: 88.4, 90.7), 85.9% (2337 of 2720 participants; 95% CI: 84.6, 87.2), and 90.2% (2453 of 2720 participants; 95% CI: 89.1, 91.3) in the same testing data set. The sensitivities were 61.6% (242 of 393 participants; 95% CI: 59.7, 63.4), 46.9% (226 of 482 participants; 95% CI: 45.0, 48.8), and 36.1% (91 of 252 participants; 95% CI: 34.3, 37.9), respectively. Conclusion A deep learning model applied to volumetric chest CT predicted pulmonary function with relatively good performance. © RSNA, 2023 Supplemental material is available for this article.

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.

Artificial Intelligence in Lung Imaging

Recently, interest and advances in artificial intelligence (AI) including deep learning for medical images have surged. As imaging plays a major role in the assessment of pulmonary diseases, various AI algorithms have been developed for chest imaging. Some of these have been approved by governments and are now commercially available in the marketplace. In the field of chest radiology, there are various tasks and purposes that are suitable for AI: initial evaluation/triage of certain diseases, detection and diagnosis, quantitative assessment of disease severity and monitoring, and prediction for decision support. While AI is a powerful technology that can be applied to medical imaging and is expected to improve our current clinical practice, some obstacles must be addressed for the successful implementation of AI in workflows. Understanding and becoming familiar with the current status and potential clinical applications of AI in chest imaging, as well as remaining challenges, would be essential for radiologists and clinicians in the era of AI. This review introduces the potential clinical applications of AI in chest imaging and also discusses the challenges for the implementation of AI in daily clinical practice and future directions in chest imaging.

Development and Validation of a Model Using Radiomics Features from an Apparent Diffusion Coefficient Map to Diagnose Local Tumor Recurrence in Patients Treated for Head and Neck Squamous Cell Carcinoma

OBJECTIVE

To develop and validate a model using radiomics features from apparent diffusion coefficient (ADC) map to diagnose local tumor recurrence in head and neck squamous cell carcinoma (HNSCC).

MATERIALS AND METHODS

This retrospective study included 285 patients (mean age ± standard deviation, 62 ± 12 years; 220 male, 77.2%), including 215 for training (n = 161) and internal validation (n = 54) and 70 others for external validation, with newly developed contrast-enhancing lesions at the primary cancer site on the surveillance MRI following definitive treatment of HNSCC between January 2014 and October 2019. Of the 215 and 70 patients, 127 and 34, respectively, had local tumor recurrence. Radiomics models using radiomics scores were created separately for T2-weighted imaging (T2WI), contrast-enhanced T1-weighted imaging (CE-T1WI), and ADC maps using non-zero coefficients from the least absolute shrinkage and selection operator in the training set. Receiver operating characteristic (ROC) analysis was used to evaluate the diagnostic performance of each radiomics score and known clinical parameter (age, sex, and clinical stage) in the internal and external validation sets.

RESULTS

Five radiomics features from T2WI, six from CE-T1WI, and nine from ADC maps were selected and used to develop the respective radiomics models. The area under ROC curve (AUROC) of ADC radiomics score was 0.76 (95% confidence interval [CI], 0.62-0.89) and 0.77 (95% CI, 0.65-0.88) in the internal and external validation sets, respectively. These were significantly higher than the AUROC values of T2WI (0.53 [95% CI, 0.40-0.67], p = 0.006), CE-T1WI (0.53 [95% CI, 0.40-0.67], p = 0.012), and clinical parameters (0.53 [95% CI, 0.39-0.67], p = 0.021) in the external validation set.

CONCLUSION

The radiomics model using ADC maps exhibited higher diagnostic performance than those of the radiomics models using T2WI or CE-T1WI and clinical parameters in the diagnosis of local tumor recurrence in HNSCC following definitive treatment.

Incidence and predictors of transthyretin cardiac amyloidosis in patients with degenerative aortic stenosis

Background

The approval of new therapies for treating transthyretin (TTR) cardiac amyloidosis has led to significant interest in identifying patients at high risk for this disease. Investigators have identified a correlation between severe degenerative aortic stenosis (AS) and TTR cardiac amyloidosis in older patients, with several studies finding up to 20% of patients who undergo transcatheter aortic valve replacement (TAVR) having TTR cardiac amyloidosis. These initial TAVR studies were conducted when TAVR was performed almost exclusively in high surgical risk populations. Therefore, the true incidence of TTR cardiac amyloidosis in an all-comers with severe degenerative AS referred for TAVR is unknown.

Purpose

To identify the true burden of TTR cardiac amyloidosis in a contemporary population of severe degenerative AS undergoing TAVR screening.

Methods

All patients ≥70 years with severe native valve degenerative AS seen in a multidisciplinary valve clinic were referred for technetium-99m pyrophosphate cardiac scintigraphy (PYP scan) for evaluation of cardiac amyloidosis. Diagnosis was made via combination of planar grade and heart to contralateral lung ratio, confirmed on single positron emission computed tomography/computed tomography (SPECT/CT). Patients with a positive PYP scan were referred to a heart failure clinic where they underwent testing for AL amyloidosis.

Results

Over a 10 month period, 247 patients seen in valve clinic underwent a PYP scan. Of this cohort, 203 patients ultimately underwent TAVR, 15 surgical aortic valve replacement, and 2 balloon valvuloplasty with 27 patients having not yet undergone a procedure. The positivity rate was 4% (10/247) with 1 patient having an equivocal result. The patients with a positive PYP scan had higher rate of low-flow low-gradient (LFLG) AS (64% vs 29%, p=0.006) and classical LFLG AS (27% vs 8%, p=0.02). Echocardiographic measures associated with a positive PYP scan include a lower global longitudinal strain (−10.0% vs −16.1%, p=0.008), lower average e' (4.5 vs 6.5, p=0.003), and an increased E/A ratio (1.8 vs 1.1, p=0.018).

Conclusions

The incidence of TTR amyloidosis in a contemporary, severe AS population ≥70 years undergoing TAVR screening appears much lower than previously described in the literature. Echocardiographic traits may be able to predict which patients with severe AS should undergo evaluation for TTR cardiac amyloidosis.

Funding Acknowledgement

Type of funding sources: None.

The influence of multiple frailty profiles on institutionalization and all-cause mortality in community-living older adults

BACKGROUND

Frailty in older adults is associated with adverse geriatric outcomes. Physical frailty is often accompanied by problems in the cognitive, psychological, and social domains. This study investigated the ability of physical frailty combined with other health domains to predict institutionalization and mortality.

METHODS

A national sample of 9171 Koreans aged 65 years or older were surveyed at baseline in 2008 and 3 year follow-up. Those who were prefrail or frail according to the Fried criteria were conceived to have physical frailty. Psychological frailty, cognitive frailty, and social frailty were defined as having depressive symptoms, cognitive impairment, and social vulnerabilities, respectively, in addition to physical frailty. Using Cox proportional hazards and competing-risks regression, the risk of mortality and institutionalization by the number and profiles of different frailty domains was analysed.

RESULTS

At baseline, the 9171 participants were aged 73.1 (±6.8) years on average (median: 72, range: 65 to 103), and 59.2% were women. Multidomain frailty was highly prevalent (49.3%), with 6.1% concurrently displaying frailty in all four domains (mixed frailty). The risk of negative health outcomes increased with frailty in a higher number of domains with a subhazard ratio (SHR) of 3.48 (95% confidence interval [CI]: 1.83, 6.62; P < 0.001) for institutionalization and a hazard ratio (HR) of 3.95 (95% CI: 2.62, 5.93; P < 0.001) for mortality among those presenting mixed frailty. Psychological frailty (depressive symptoms combined with physical frailty) was strongly predictive of institutionalization (SHR = 2.85; 95% CI: 1.45, 5.59; P = 0.002) and mortality (HR = 2.47; 95% CI: 1.61, 3.78; P < 0.001). When combined with physical frailty and either depressive symptoms or social vulnerabilities, cognitive impairment also exhibited a significantly elevated risk of negative events. Physical frailty alone was not a strong predictor of adverse events, especially for mortality (HR = 1.13; 95% CI: 0.77, 1.67; P = 0.53).

CONCLUSIONS

Co-occurrence of physical frailty with other domains is common in late life. The presence of frailty in multiple domains raises the risk of adverse outcomes, with the effects varying by multidimensional profiles.

Multi-domain CT translation by a routable translation network

Objective. To unify the style of CT images from multiple sources, we propose a novel multi-domain image translation network to convert CT images from different scan parameters and manufacturers by simply changing a routing vector. Approach. Unlike the existing multi-domain translation techniques, our method is based on a shared encoder and a routable decoder architecture to maximize the expressivity and conditioning power of the network. Main results. Experimental results show that the proposed CT image conversion can minimize the variation of image characteristics caused by imaging parameters, reconstruction algorithms, and hardware designs. Quantitative results and clinical evaluation from radiologists also show that our method can provide accurate translation results. Significance. Quantitative evaluation of CT images from multi-site or longitudinal studies has been a difficult problem due to the image variation depending on CT scan parameters and manufacturers. The proposed method can be utilized to address this for the quantitative analysis of multi-domain CT images.

Genetic diversity and population structure of Mongolian regional horses with 14 Microsatellite markers

Objective

This study aimed to identify the genetic diversity and population structure of Mongolian horse populations according to the province of residence (Khentii, KTP; Uvs, USP; Omnogovi and Dundgovi, GOP; Khovsgol, KGP) using 14 microsatellite (MS) markers.

Methods

A total of 269 whole blood samples were obtained from the four populations (KTP, USP, GOP, KGP) geographically distinct provinces. Multiplex polymerase chain reaction (PCR) was conducted using 14 MS markers (AHT4, ASB2, ASB17, ASB23, CA425, HMS1, HMS2, HMS3, HMS6, HMS7, HTG4, HTG6, HTG7, and VHL20), as recommended by the International Society for Animal Genetics. Capillary electrophoresis was conducted using the amplified PCR products, alleles were determined. Alleles were used for statistical analysis of genetic variability, Nei’s DA genetic distance, principal coordinate analysis (PCoA), factorial corresponding analysis (FCA), and population structure.

Results

On average, the number of alleles, expected heterozygosity (H_Exp), observed heterozygosity (H_Obs), and polymorphic information content among all populations were 11.43, 0.772, 0.757, and 0.737, respectively. In the PCoA and FCA, GOP, and KGP were genetically distinct from other populations, and the KTP and USP showed a close relationship. The two clusters identified using Nei’s DA genetic distance analysis and population structure highlighted the presence of structurally clear genetic separation.

Conclusion

Overall, the results of this study suggest that genetic diversity between KTP and USP was low, and that between GOP and KGP was high. It is thought that these results will help in the effective preservation and improvement of Mongolian horses through genetic diversity analysis and phylogenetic relationships.

CTRP3 exacerbates tendinopathy by dysregulating tendon stem cell differentiation and altering extracellular matrix composition

Tendinopathy, the most common disorder affecting tendons, is characterized by chronic disorganization of the tendon matrix, which leads to tendon tear and rupture. The goal was to identify a rational molecular target whose blockade can serve as a potential therapeutic intervention for tendinopathy. We identified C1q/TNF-related protein-3 (CTRP3) as a markedly up-regulated cytokine in human and rodent tendinopathy. Overexpression of CTRP3 enhanced the progression of tendinopathy by accumulating cartilaginous proteoglycans and degenerating collagenous fibers in the mouse tendon, whereas CTRP3 knockdown suppressed the tendinopathy pathogenesis. Functional blockade of CTRP3 using a neutralizing antibody ameliorated overuse-induced tendinopathy of the Achilles and rotator cuff tendons. Mechanistically, CTRP3 elicited a transcriptomic pattern that stimulates abnormal differentiation of tendon stem/progenitor cells and ectopic chondrification as an effect linked to activation of Akt signaling. Collectively, we reveal an essential role for CTRP3 in tendinopathy and propose a potential therapeutic strategy for the treatment of tendinopathy.

An Open Medical Platform to Share Source Code and Various Pre-Trained Weights for Models to Use in Deep Learning Research

Deep learning-based applications have great potential to enhance the quality of medical services. The power of deep learning depends on open databases and innovation. Radiologists can act as important mediators between deep learning and medicine by simultaneously playing pioneering and gatekeeping roles. The application of deep learning technology in medicine is sometimes restricted by ethical or legal issues, including patient privacy and confidentiality, data ownership, and limitations in patient agreement. In this paper, we present an open platform, MI2RLNet, for sharing source code and various pre-trained weights for models to use in downstream tasks, including education, application, and transfer learning, to encourage deep learning research in radiology. In addition, we describe how to use this open platform in the GitHub environment. Our source code and models may contribute to further deep learning research in radiology, which may facilitate applications in medicine and healthcare, especially in medical imaging, in the near future. All code is available at https://github.com/mi2rl/MI2RLNet.

Impact of doppler velocity index after transcatheter aortic valve replacement using Sapien-3 valve – a single centre experience

Background

Little is known about the hemodynamic performance of Edwards Sapien-3 (S3) valve after Transcatheter Aortic Valve Replacement (TAVR). Doppler velocity index (DVI) is a better indicator of prosthetic valve function as it is independent of valve size and flow, unlike mean gradient and peak velocity which are flow-dependent.

Methods

In this study, we compare outcomes based on differences in DVI among a consecutive series of patients who underwent S3 TAVR between April 2015 and December 2018. Our institutional review board approved the study and informed consents were obtained from the subjects.

Results

Among 921 patients who had follow-up echocardiograms within 30 days after TAVR, 60.8% had DVI ≤0.5, while 39.2% had DVI &gt;0.5. The median 30-day DVI was 0.47 with a standard deviation of 0.11 and mean 0.49 and interquartile range 0.41–0.55. The baseline clinical and procedural characteristics were similar between both groups with the exception of less post-dilation (36.8% vs. 47.4%; p=0.001) and greater implantation depth (2.59±1.99 vs. 2.31±1.9mm; p=0.031) with DVI ≤0.5. The rates of aortic valve calcification, pre dilation, pre-TAVR aortic regurgitation (AR) were similar. At baseline, there were no differences between both groups in mean or peak gradients or aortic velocity time integral (VTI). At 1 year, mean gradients were higher with DVI ≤0.5 (12.7±5.6 vs. 11.1±4.6 mmHg; p=0.001). DVI ≤0.5 was associated with higher peak gradients (24.2±10.2 vs. 21.4±8.7 mmHg; p=0.002), and aortic VTI (51.4±13.5 vs. 46.8±12.2 cm; p&lt;0.001) at 1 year, especially with the 26mm and 29mm prostheses. Compared with DVI&gt;0.5 group, patients in DVI&lt;0.5 group had lower baseline left ventricular ejection fraction (LVEF) (54.5±12.2% vs. 58.9±11.2%; p&lt;0.001), higher left ventricular end-diastolic volume index (LVEDVi) (54.3±20.9 vs. 49.4±17.4 ml/m2; p=0.001), higher LV end-systolic volume index (LVESVi) (25.2±16.5 vs. 21.3±12.7 ml/m2; p=0.001), and similar LV mass index (110.7±31.9 vs. 106.9±32.7 g/m2; p=0.134). 1-year mortality rates among patients who had DVI ≤0.5 compared to DVI &gt;0.5 were lower (6.6% vs. 10.6%; log-rank p=0.033), however no difference was noted at both 2 years (17.3% vs. 20.1%; log-rank p=0.151), and 3 years after TAVR (30.7% vs. 31.2%; log-rank p=0.333).

Conclusions

DVI&lt;0.5 was associated with higher peak gradients and lower baseline LVEF. DVI &lt;0.5 group patients had lower 1-year mortality but similar mortality at 2 and 3-years of follow up.

Funding Acknowledgement

Type of funding sources: None. Figure 1. All-cause SurvivalFigure 2. Hemodynamic Data

Late Clinical and Hemodynamic Outcomes in patients with degenerated bioprosthetic aortic valves undergoing transcatheter valve-in-valve implantation

Background

Clinical studies have shown promising early outcomes for valve-in-valve transcatheter aortic valve implantation (ViV-TAVI); however, the late outcomes of this procedure remain under-investigated.

Purpose

We performed the present analysis to assess the late clinical and hemodynamic outcomes of ViV-TAVI in patients with degenerated bioprosthetic aortic valves.

Methods

A comprehensive chart review was performed for eligible patients to retrieve data on procedural characteristics, admission details following the procedure, and echocardiographic parameters. Clinical outcomes included all-cause mortality, heart failure hospitalization and structural valve deterioration (SVD), as defined by VARC-II criteria, up to 5 years of follow-up. To assess the trends in mean and peak transvalvular gradients, data from the follow-up echocardiographic reports were analyzed using Syngo Dynamics imaging software.

Results

A total of 188 patients were included with a mean age of 75.8±10.4 years. Balloon- and self-expandable valves were used in 155 (82.4%) and 33 (17.6%) patients, respectively. At 30 days, 3 (1.6%) patients died and 8 (4.2%) required hospitalization for heart failure, while at 5 years, both events were recorded in 29 (15.4%) and 37 (19.7) patients, respectively. Kaplan-Meier survival analysis showed that patients with smaller surgical valves (internal diameter ≤21 mm) had a significantly higher mortality rate (log-rank p=0.021) than those with larger valves; however, no significant difference (log-rank p=0.59) was detected between different transcatheter valves (self vs. balloon-expandable). Three patients underwent re-intervention, performed via a transcatheter approach. Further, assessment of follow-up echocardiographic reports revealed 9 (4.8%) cases of SVD, as well as stable mean (16.3±6.9 at discharge and 16.9±11.3) and peak (30.3±12.1 at discharge and 30.7±18.4 at 5 years) transvalvular gradients. No difference (p&gt;0.05) was observed based on transcatheter valve type or surgical valve internal diameter in terms of mean and peak transvalvular gradients throughout the follow-up period.

Conclusion

The present study showed good clinical outcomes among patients undergoing VIV-TAVI, with stable VIV performance over a five-year period. Future long-term studies are warranted to analyze the predictors of outcomes following ViV-TAVI and explore the role of this treatment option in the life-long management of aortic stenosis.

Funding Acknowledgement

Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Jennifer and Robert McNeil Donation to the Heart, Thoracic, and Vascular Institute at Cleveland Clinic. Figure 1

Outcomes of transfemoral-transcatheter aortic valve replacement with Sapien-3 valve in liver cirrhosis patients

Background

Little is known about the outcomes of liver cirrhosis patients with severe aortic stenosis (AS) who undergo transcatheter aortic valve replacement (TAVR).

Methods

We undertook a retrospective analysis of consecutive patients with severe symptomatic AS who underwent transfemoral-TAVR with Sapien-3 valve at our Clinic between April 2015 and December 2018, yielding 32 patients with liver cirrhosis on imaging including ultrasound and/or computed tomography. Their baseline characteristics, procedural and long-term outcomes after TAVR with the non-cirrhotic group were compared, along with their management strategies as per the hepatology team.

Results

Among 1028 patients, 32 were assigned to the cirrhosis, and 996 were assigned to the non-cirrhosis (control) group. Compared with the control group cirrhotic patients were slightly younger in age (74.5 vs 81.2 years), had a slightly higher BMI (31.3 vs 29.3), and had a higher incidence of prior history of myocardial infarction (38% vs 33%). Baseline variables including the history of smoking, hypertension, diabetes, and atrial fibrillation were comparable in both groups. Among cirrhotic patients (n=32), the most common etiologies were non-alcoholic steatohepatitis (NASH) (37.5%), Alcoholism (18.75%), and Hepatitis C (12.5%). The mean MELD-NA score was 11.8 and 67% of patients were Child PUGH Class A and 33% were Child PUGH Class B and all patients had a Child PUGH score of ≥5. 53% of patients (n=17) in the cirrhosis group were evaluated by Hepatology and 12.5% (n=4) were evaluated for a liver transplant but only 1 patient had a liver transplant post-TAVR. Compared with the control group cirrhotic patients had similar 1-year mortality (12% vs 12%, p=1), had a lower rate of 30-day new pacemaker post tavr (6% vs 9% p=0.85), had a higher rate of 1-year readmission for heart failure (12% vs 5% p=0.12) and similar 1-year major adverse cardiac and cerebrovascular event (MACCE) rate (15% vs 14% p=0.98)

Conclusion

Patients with severe AS undergoing TAVR with concomitant liver cirrhosis demonstrate comparable outcomes compared with their non- cirrhotic counterparts. NASH followed by alcoholic cirrhosis was found to be most common etiology.

Funding Acknowledgement

Type of funding sources: None. Figure 1Figure 2

Content-based Image Retrieval by Using Deep Learning for Interstitial Lung Disease Diagnosis with Chest CT

Background Evaluation of interstitial lung disease (ILD) at CT is a challenging task that requires experience and is subject to substantial interreader variability. Purpose To investigate whether a proposed content-based image retrieval (CBIR) of similar chest CT images by using deep learning can aid in the diagnosis of ILD by readers with different levels of experience. Materials and methods This retrospective study included patients with confirmed ILD after multidisciplinary discussion and available CT images identified between January 2000 and December 2015. Database was composed of four disease classes: usual interstitial pneumonia (UIP), nonspecific interstitial pneumonia (NSIP), cryptogenic organizing pneumonia, and chronic hypersensitivity pneumonitis. Eighty patients were selected as queries from the database. The proposed CBIR retrieved the top three similar CT images with diagnosis from the database by comparing the extent and distribution of different regional disease patterns quantified by a deep learning algorithm. Eight readers with varying experience interpreted the query CT images and provided their most probable diagnosis in two reading sessions 2 weeks apart, before and after applying CBIR. Diagnostic accuracy was analyzed by using McNemar test and generalized estimating equation, and interreader agreement was analyzed by using Fleiss κ. Results A total of 288 patients were included (mean age, 58 years ± 11 [standard deviation]; 145 women). After applying CBIR, the overall diagnostic accuracy improved in all readers (before CBIR, 46.1% [95% CI: 37.1, 55.3]; after CBIR, 60.9% [95% CI: 51.8, 69.3]; P < .001). In terms of disease category, the diagnostic accuracy improved after applying CBIR in UIP (before vs after CBIR, 52.4% vs 72.8%, respectively; P < .001) and NSIP cases (before vs after CBIR, 42.9% vs 61.6%, respectively; P < .001). Interreader agreement improved after CBIR (before vs after CBIR Fleiss κ, 0.32 vs 0.47, respectively; P = .005). Conclusion The proposed content-based image retrieval system for chest CT images with deep learning improved the diagnostic accuracy of interstitial lung disease and interreader agreement in readers with different levels of experience. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Wielpütz in this issue.

Radiomics approach for survival prediction in chronic obstructive pulmonary disease

OBJECTIVES

To apply radiomics analysis for overall survival prediction in chronic obstructive pulmonary disease (COPD), and evaluate the performance of the radiomics signature (RS).

METHODS

This study included 344 patients from the Korean Obstructive Lung Disease (KOLD) cohort. External validation was performed on a cohort of 112 patients. In total, 525 chest CT-based radiomics features were semi-automatically extracted. The five most useful features for survival prediction were selected by least absolute shrinkage and selection operation (LASSO) Cox regression analysis and used to generate a RS. The ability of the RS for classifying COPD patients into high or low mortality risk groups was evaluated with the Kaplan-Meier survival analysis and Cox proportional hazards regression analysis.

RESULTS

The five features remaining after the LASSO analysis were %LAA_-950, AWT_Pi10_6^th, AWT_Pi10_heterogeneity, %WA_heterogeneity, and VA_18mm. The RS demonstrated a C-index of 0.774 in the discovery group and 0.805 in the validation group. Patients with a RS greater than 1.053 were classified into the high-risk group and demonstrated worse overall survival than those in the low-risk group in both the discovery (log-rank test, < 0.001; hazard ratio [HR], 5.265) and validation groups (log-rank test, < 0.001; HR, 5.223). For both groups, RS was significantly associated with overall survival after adjustments for patient age and body mass index.

CONCLUSIONS

A radiomics approach for survival prediction and risk stratification in COPD patients is feasible, and the constructed radiomics model demonstrated acceptable performance. The RS derived from chest CT data of COPD patients was able to effectively identify those at increased risk of mortality.

KEY POINTS

• A total of 525 chest CT-based radiomics features were extracted and the five radiomics features of %LAA_-950, AWT_Pi10_6^th, AWT_Pi10_heterogeneity, %WA_heterogeneity, and VA_18mm were selected to generate a radiomics model. • A radiomics model for predicting survival of COPD patients demonstrated reliable performance with a C-index of 0.774 in the discovery group and 0.805 in the validation group. • Radiomics approach was able to effectively identify COPD patients with an increased risk of mortality, and patients assigned to the high-risk group demonstrated worse overall survival in both the discovery and validation groups.

Value of bone suppression software in chest radiographs for improving image quality and reducing radiation dose

OBJECTIVES

To compare image quality and radiation dose between dual-energy subtraction (DES)-based bone suppression images (D-BSIs) and software-based bone suppression images (S-BSIs).

METHODS

Chest radiographs (CXRs) of forty adult patients were obtained with the two X-ray devices, one with DES and one with bone suppression software. Three image quality metrics (relative mean absolute error (RMAE), peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM)) between original CXR and BSI for each of D-BSI and S-SBI groups were calculated for each bone and soft tissue areas. Two readers rated the visual image quality for original CXR and BSI for each of D-BSI and S-SBI groups. The dose area product (DAP) values were recorded. Paired t test was used to compare the image quality and DAP values between D-BSI and S-BSI groups.

RESULTS

In bone areas, S-BSIs had better SSIM values than D-BSI (94.57 vs. 87.77) but worse RMAE and PSNR values (0.50 vs. 0.20; 20.93 vs. 34.37) (all p < 0.001). In soft tissue areas, S-BSIs had better SSIM values than D-BSI (97.56 vs. 91.42) but similar RMAE and PSNR values (0.29 vs. 0.27; 31.35 vs. 29.87) (all p < 0.001). Both readers gave S-BSIs significantly higher image quality scores than D-BSI (p < 0.001). The mean DAP in software-related images (0.98 dGy·cm²) was significantly lower than that in the DES-related images (1.48 dGy·cm²) (p < 0.001).

CONCLUSION

Bone suppression software significantly improved the image quality of bone suppression images with a relatively lower radiation dose, compared with dual-energy subtraction technique.

KEY POINTS

• Bone suppression software preserves structure similarity of soft tissues better than dual-energy subtraction technique in bone suppression images. • Bone suppression software achieves superior image quality for lung lesions than dual-energy subtraction technique in bone suppression images. • Bone suppression software can decrease the radiation dose over the hardware-based image processing technique.

Comparison of uniaxial and triaxial accelerometer outputs among individuals with and without Down syndrome

BACKGROUND

Using uniaxial accelerometry approach in measuring physical activity levels of individuals with Down syndrome (DS) might underestimate their energy expenditure due to the unique mediolateral walking pattern. Therefore, the purpose of this study was to examine and compare the relationship between two types of accelerometer outputs, uniaxial and triaxial, and energy expenditure in individuals with and without DS.

METHODS

Eighteen participants with DS and 19 participants without DS wore a GT3X+ accelerometer and a portable metabolic system in three different walking conditions.

RESULTS

Correlations between V̇O2 and each of the two accelerometer outputs (uniaxial: r = 0.75, triaxial: r = 0.75) were not significantly different among individuals without DS (z = 0.14, P = 0.89); however, significant differences in the relationship between V̇O2 and accelerometer outputs (uniaxial: r = 0.53, triaxial: r = 0.64) were observed among individuals with DS (z = -1.72, P < 0.046).

CONCLUSIONS

The findings suggest that when using accelerometers to measure physical activity levels for individuals with DS, triaxial outputs may better predict physical activity levels.

Morolic Acid 3-O-Caffeate Inhibits Adipogenesis by Regulating Epigenetic Gene Expression

Obesity causes a wide range of metabolic diseases including diabetes, cardiovascular disease, and kidney disease. Thus, plenty of studies have attempted to discover naturally derived compounds displaying anti-obesity effects. In this study, we evaluated the inhibitory effects of morolic acid 3-O-caffeate (MAOC), extracted from Betula schmidtii, on adipogenesis. Treatment of 3T3-L1 cells with MAOC during adipogenesis significantly reduced lipid accumulation and decreased the expression of adiponectin, a marker of mature adipocytes. Moreover, the treatment with MAOC only during the early phase (day 0-2) sufficiently inhibited adipogenesis, comparable with the inhibitory effects observed following MAOC treatment during the whole processes of adipogenesis. In the early phase of adipogenesis, the expression level of Wnt6, which inhibits adipogenesis, increased by MAOC treatment in 3T3-L1 cells. To identify the gene regulatory mechanism, we assessed alterations in histone modifications upon MAOC treatment. Both global and local levels on the Wnt6 promoter region of histone H3 lysine 4 trimethylation, an active transcriptional histone marker, increased markedly by MAOC treatment in 3T3-L1 cells. Our findings identified an epigenetic event associated with inhibition of adipocyte generation by MAOC, suggesting its potential as an efficient therapeutic compound to cure obesity and metabolic diseases.

Diffusion and perfusion MRI radiomics obtained from deep learning segmentation provides reproducible and comparable diagnostic model to human in post-treatment glioblastoma

OBJECTIVES

Deep learning-based automatic segmentation (DLAS) helps the reproducibility of radiomics features, but its effect on radiomics modeling is unknown. We therefore evaluated whether DLAS can robustly extract anatomical and physiological MRI features, thereby assisting in the accurate assessment of treatment response in glioblastoma patients.

METHODS

A DLAS model was trained on 238 glioblastomas and validated on an independent set of 98 pre- and 86 post-treatment glioblastomas from two tertiary hospitals. A total of 1618 radiomics features from contrast-enhanced T1-weighted images (CE-T1w) and histogram features from apparent diffusion coefficient (ADC) and cerebral blood volume (CBV) mapping were extracted. The diagnostic performance of radiomics features and ADC and CBV parameters for identifying treatment response was tested using area under the curve (AUC) from receiver operating characteristics analysis. Feature reproducibility was tested using a 0.80 cutoff for concordance correlation coefficients.

RESULTS

Reproducibility was excellent for ADC and CBV features (ICC, 0.82-0.99) and first-order features (pre- and post-treatment, 100% and 94.1% remained), but lower for texture (79.0% and 69.1% remained) and wavelet-transformed (81.8% and 74.9% remained) features of CE-T1w. DLAS-based radiomics showed similar performance to human-performed segmentations in internal validation (AUC, 0.81 [95% CI, 0.64-0.99] vs. AUC, 0.81 [0.60-1.00], p = 0.80), but slightly lower performance in external validation (AUC, 0.78 [0.61-0.95] vs. AUC, 0.65 [0.46-0.84], p = 0.23).

CONCLUSION

DLAS-based feature extraction showed high reproducibility for first-order features from anatomical and physiological MRI, and comparable diagnostic performance to human manual segmentations in the identification of pseudoprogression, supporting the utility of DLAS in quantitative MRI analysis.

KEY POINTS

• Deep learning-based automatic segmentation (DLAS) enables fast and robust feature extraction from diffusion- and perfusion-weighted MRI. • DLAS showed high reproducibility in first-order feature extraction from anatomical, diffusion, and perfusion MRI across two centers. • DLAS-based radiomics features showed comparable diagnostic accuracy to manual segmentations in post-treatment glioblastoma.

Oropharyngeal squamous cell carcinoma: radiomic machine-learning classifiers from multiparametric MR images for determination of HPV infection status

We investigated the ability of machine-learning classifiers on radiomics from pre-treatment multiparametric magnetic resonance imaging (MRI) to accurately predict human papillomavirus (HPV) status in patients with oropharyngeal squamous cell carcinoma (OPSCC). This retrospective study collected data of 60 patients (48 HPV-positive and 12 HPV-negative) with newly diagnosed histopathologically proved OPSCC, who underwent head and neck MRIs consisting of axial T1WI, T2WI, CE-T1WI, and apparent diffusion coefficient (ADC) maps from diffusion-weighted imaging (DWI). The median age was 59 years (the range being 35 to 85 years), and 83.3% of patients were male. The imaging data were randomised into a training set (32 HPV-positive and 8 HPV-negative OPSCC) and a test set (16 HPV-positive and 4 HPV-negative OPSCC) in each fold. 1618 quantitative features were extracted from manually delineated regions-of-interest of primary tumour and one definite lymph node in each sequence. After feature selection by using the least absolute shrinkage and selection operator (LASSO), three different machine-learning classifiers (logistic regression, random forest, and XG boost) were trained and compared in the setting of various combinations between four sequences. The highest diagnostic accuracies were achieved when using all sequences, and the difference was significant only when the combination did not include the ADC map. Using all sequences, logistic regression and the random forest classifier yielded higher accuracy compared with the that of the XG boost classifier, with mean area under curve (AUC) values of 0.77, 0.76, and 0.71, respectively. The machine-learning classifier of non-invasive and quantitative radiomics signature could guide the classification of the HPV status.

Development of machine learning-based clinical decision support system for hepatocellular carcinoma

There is a significant discrepancy between the actual choice for initial treatment option for hepatocellular carcinoma (HCC) and recommendations from the currently used BCLC staging system. We develop a machine learning-based clinical decision support system (CDSS) for recommending initial treatment option in HCC and predicting overall survival (OS). From hospital records of 1,021 consecutive patients with HCC treated at a single centre in Korea between January 2010 and October 2010, we collected information on 61 pretreatment variables, initial treatment, and survival status. Twenty pretreatment key variables were finally selected. We developed the CDSS from the derivation set (N = 813) using random forest method and validated it in the validation set (N = 208). Among the 1,021 patients (mean age: 56.9 years), 81.8% were male and 77.0% had positive hepatitis B BCLC stages 0, A, B, C, and D were observed in 13.4%, 26.0%, 18.0%, 36.6%, and 6.3% of patients, respectively. The six multi-step classifier model was developed for treatment decision in a hierarchical manner, and showed good performance with 81.0% of accuracy for radiofrequency ablation (RFA) or resection versus not, 88.4% for RFA versus resection, and 76.8% for TACE or not. We also developed seven survival prediction models for each treatment option. Our newly developed HCC-CDSS model showed good performance in terms of treatment recommendation and OS prediction and may be used as a guidance in deciding the initial treatment option for HCC.

Infection of Brain Organoids and 2D Cortical Neurons with SARS-CoV-2 Pseudovirus

Since the global outbreak of SARS-CoV-2 (COVID-19), infections of diverse human organs along with multiple symptoms continue to be reported. However, the susceptibility of the brain to SARS-CoV-2, and the mechanisms underlying neurological infection are still elusive. Here, we utilized human embryonic stem cell-derived brain organoids and monolayer cortical neurons to investigate infection of brain with pseudotyped SARS-CoV-2 viral particles. Spike-containing SARS-CoV-2 pseudovirus infected neural layers within brain organoids. The expression of ACE2, a host cell receptor for SARS-CoV-2, was sustained during the development of brain organoids, especially in the somas of mature neurons, while remaining rare in neural stem cells. However, pseudotyped SARS-CoV-2 was observed in the axon of neurons, which lack ACE2. Neural infectivity of SARS-CoV-2 pseudovirus did not increase in proportion to viral load, but only 10% of neurons were infected. Our findings demonstrate that brain organoids provide a useful model for investigating SARS-CoV-2 entry into the human brain and elucidating the susceptibility of the brain to SARS-CoV-2.