Screening and diagnosis of cardiovascular disease using artificial intelligence-enabled cardiac magnetic resonance imaging.

Cardiac magnetic resonance imaging (CMR) is the gold standard for cardiac function assessment and plays a crucial role in diagnosing cardiovascular disease (CVD). However, its widespread application has been limited by the heavy resource burden of CMR interpretation. Here, to address this challenge, we developed and validated computerized CMR interpretation for screening and diagnosis of 11 types of CVD in 9,719 patients. We propose a two-stage paradigm consisting of noninvasive cine-based CVD screening followed by cine and late gadolinium enhancement-based diagnosis. The screening and diagnostic models achieved high performance (area under the curve of 0.988 ± 0.3% and 0.991 ± 0.0%, respectively) in both internal and external datasets. Furthermore, the diagnostic model outperformed cardiologists in diagnosing pulmonary arterial hypertension, demonstrating the ability of artificial intelligence-enabled CMR to detect previously unidentified CMR features. This proof-of-concept study holds the potential to substantially advance the efficiency and scalability of CMR interpretation, thereby improving CVD screening and diagnosis.

Corrigendum: OPN promotes pro-inflammatory cytokine expression via ERK/JNK pathway and M1 macrophage polarization in Rosacea.

[This corrects the article DOI: 10.3389/fimmu.2023.1285951.].

Towards trustworthy seizure onset detection using workflow notes.

A major barrier to deploying healthcare AI is trustworthiness. One form of trustworthiness is a model's robustness across subgroups: while models may exhibit expert-level performance on aggregate metrics, they often rely on non-causal features, leading to errors in hidden subgroups. To take a step closer towards trustworthy seizure onset detection from EEG, we propose to leverage annotations that are produced by healthcare personnel in routine clinical workflows-which we refer to as workflow notes-that include multiple event descriptions beyond seizures. Using workflow notes, we first show that by scaling training data to 68,920 EEG hours, seizure onset detection performance significantly improves by 12.3 AUROC (Area Under the Receiver Operating Characteristic) points compared to relying on smaller training sets with gold-standard labels. Second, we reveal that our binary seizure onset detection model underperforms on clinically relevant subgroups (e.g., up to a margin of 6.5 AUROC points between pediatrics and adults), while having significantly higher FPRs (False Positive Rates) on EEG clips showing non-epileptiform abnormalities (+19 FPR points). To improve model robustness to hidden subgroups, we train a multilabel model that classifies 26 attributes other than seizures (e.g., spikes and movement artifacts) and significantly improve overall performance (+5.9 AUROC points) while greatly improving performance among subgroups (up to +8.3 AUROC points) and decreasing false positives on non-epileptiform abnormalities (by 8 FPR points). Finally, we find that our multilabel model improves clinical utility (false positives per 24 EEG hours) by a factor of 2×.

Investigation on the interaction mechanisms for stability of preheated whey protein isolate with anthocyanins from blueberry.

Proteins and anthocyanins coexist in complex food systems. This research mainly studied the steady-state protective design and mechanism of the preheated protein against anthocyanins. Multispectral and molecular dynamics are utilized to illustrate the interaction mechanism between preheated whey protein isolate (pre-WPI) and anthocyanins. The pre-WPI could effectively protect the stability of anthocyanins, and the effect was better than that of the natural whey protein isolate (NW). Among them, NW after preheating treatment at 55 °C showed better protection against anthocyanin stability. Fluorescence studies indicated that pre-WPI there existed a solid binding affinity and static quenching for malvidin-3-galactoside (M3G). Multispectral data showed a significant variation in the secondary structure of pre-WPI. Furthermore, molecular dynamics simulation selects AMBER18 as the protein force field, and the results showed that hydrogen bonding participated as an applied force. Compared with NW, pre-WPI could better wrap anthocyanins and avoid damage to the external environment due to tightening of the pockets. Protein protects anthocyanins from degradation, and this protective effect is influenced by the preheating temperature of protein and the structure of protein. On the basis of the above results, it is possible to pinpoint the interaction mechanism between preheated proteins and anthocyanins.

Real-time detection of acetone gas molecules at ppt levels in an air atmosphere using a partially suspended graphene surface acoustic wave skin gas sensor.

To improve the quality of modern life in the current society, low-power, highly sensitive, and reliable healthcare technology is necessary to monitor human health in real-time. In this study, we fabricated partially suspended monolayer graphene surface acoustic wave gas sensors (G-SAWs) with a love-mode wave to effectively detect ppt-level acetone gas molecules at room temperature. The sputtered SiO2 thin film on the surface of a black 36°YX-LiTaO3 (B-LT) substrate acted as a guiding layer, effectively reducing the noise and insertion loss. The G-SAWs exhibited enhanced gas response towards acetone gas molecules (800 ppt) in a real-time atmosphere. The high sensitivity of the G-SAW sensor can be attributed to the elasticity and surface roughness of the SiO2 film. In addition, the G-SAW sensor exhibited rapid response and recovery at room temperature. This study provides a potential strategy for diagnosing different stages of diabetes in the human body.

Chiroptical Response Inversion and Enhancement of Room-Temperature Exciton-Polaritons Using 2D Chirality in Perovskites.

Although chiral semiconductors have shown promising progress in direct circularly polarized light (CPL) detection and emission, they still face potential challenges. A chirality-switching mechanism or approach integrating two enantiomers is needed to discriminate the handedness of a given CPL; additionally, a large material volume is required for sufficient chiroptical interaction. These two requirements pose significant obstacles to the simplification and miniaturization of the devices. Here, room-temperature chiral polaritons fulfilling dual-handedness functions and exhibiting a more-than-two-order enhancement of the chiroptical signal are demonstrated, by embedding a 40 nm-thick perovskite film with a 2D chiroptical effect into a Fabry-Pérot cavity. By mixing chiral perovskites with different crystal structures, a pronounced 2D chiroptical effect is accomplished in the perovskite film, featured by an inverted chiroptical response for counter-propagating CPL. This inversion behavior matches the photonic handedness switch during CPL circulation in the Fabry-Pérot cavity, thus harvesting giant enhancement of the chiroptical response. Furthermore, affected by the unique quarter-wave-plate effects, the polariton emission achieves a chiral dissymmetry of ±4% (for the emission from the front and the back sides). The room-temperature polaritons with the strong dissymmetric chiroptical interaction shall have implications on a fundamental level and future on-chip applications for biomolecule analysis and quantum computing.

Fusion of imaging and non-imaging data for disease trajectory prediction for coronavirus disease 2019 patients.

Purpose: Our study investigates whether graph-based fusion of imaging data with non-imaging electronic health records (EHR) data can improve the prediction of the disease trajectories for patients with coronavirus disease 2019 (COVID-19) beyond the prediction performance of only imaging or non-imaging EHR data. Approach: We present a fusion framework for fine-grained clinical outcome prediction [discharge, intensive care unit (ICU) admission, or death] that fuses imaging and non-imaging information using a similarity-based graph structure. Node features are represented by image embedding, and edges are encoded with clinical or demographic similarity. Results: Experiments on data collected from the Emory Healthcare Network indicate that our fusion modeling scheme performs consistently better than predictive models developed using only imaging or non-imaging features, with area under the receiver operating characteristics curve of 0.76, 0.90, and 0.75 for discharge from hospital, mortality, and ICU admission, respectively. External validation was performed on data collected from the Mayo Clinic. Our scheme highlights known biases in the model prediction, such as bias against patients with alcohol abuse history and bias based on insurance status. Conclusions: Our study signifies the importance of the fusion of multiple data modalities for the accurate prediction of clinical trajectories. The proposed graph structure can model relationships between patients based on non-imaging EHR data, and graph convolutional networks can fuse this relationship information with imaging data to effectively predict future disease trajectory more effectively than models employing only imaging or non-imaging data. Our graph-based fusion modeling frameworks can be easily extended to other prediction tasks to efficiently combine imaging data with non-imaging clinical data.

Artificial Intelligence Predictive Model for Hormone Therapy Use in Prostate Cancer.

Background: Androgen deprivation therapy (ADT) with radiotherapy can benefit patients with localized prostate cancer. However, ADT can negatively impact quality of life and there remain no validated predictive models to guide its use. Methods: Digital pathology image and clinical data from pre-treatment prostate tissue from 5,727 patients enrolled on five phase III randomized trials treated with radiotherapy +/- ADT were used to develop and validate an artificial intelligence (AI)-derived predictive model to assess ADT benefit with the primary endpoint of distant metastasis. After the model was locked, validation was performed on NRG/RTOG 9408 (n = 1,594) that randomized men to radiotherapy +/- 4 months of ADT. Fine-Gray regression and restricted mean survival times were used to assess the interaction between treatment and predictive model and within predictive model positive and negative subgroup treatment effects. Results: In the NRG/RTOG 9408 validation cohort (14.9 years of median follow-up), ADT significantly improved time to distant metastasis (subdistribution hazard ratio [sHR] = 0.64, 95%CI [0.45-0.90], p = 0.01). The predictive model-treatment interaction was significant (p-interaction = 0.01). In predictive model positive patients (n = 543, 34%), ADT significantly reduced the risk of distant metastasis compared to radiotherapy alone (sHR = 0.34, 95%CI [0.19-0.63], p < 0.001). There were no significant differences between treatment arms in the predictive model negative subgroup (n = 1,051, 66%; sHR = 0.92, 95%CI [0.59-1.43], p = 0.71). Conclusions: Our data, derived and validated from completed randomized phase III trials, show that an AI-based predictive model was able to identify prostate cancer patients, with predominately intermediate-risk disease, who are likely to benefit from short-term ADT.

Predicting 30-day all-cause hospital readmission using multimodal spatiotemporal graph neural networks.

Reduction in 30-day readmission rate is an important quality factor for hospitals as it can reduce the overall cost of care and improve patient post-discharge outcomes. While deep-learning-based studies have shown promising empirical results, several limitations exist in prior models for hospital readmission prediction, such as: (a) only patients with certain conditions are considered, (b) do not leverage data temporality, (c) individual admissions are assumed independent of each other, which ignores patient similarity, (d) limited to single modality or single center data. In this study, we propose a multimodal, spatiotemporal graph neural network (MM-STGNN) for prediction of 30-day all-cause hospital readmission, which fuses in-patient multimodal, longitudinal data and models patient similarity using a graph. Using longitudinal chest radiographs and electronic health records from two independent centers, we show that MM-STGNN achieved an area under the receiver operating characteristic curve (AUROC) of 0.79 on both datasets. Furthermore, MM-STGNN significantly outperformed the current clinical reference standard, LACE+ (AUROC=0.61), on the internal dataset. For subset populations of patients with heart disease, our model significantly outperformed baselines, such as gradient-boosting and Long Short-Term Memory models (e.g., AUROC improved by 3.7 points in patients with heart disease). Qualitative interpretability analysis indicated that while patients' primary diagnoses were not explicitly used to train the model, features crucial for model prediction may reflect patients' diagnoses. Our model could be utilized as an additional clinical decision aid during discharge disposition and triaging high-risk patients for closer post-discharge follow-up for potential preventive measures.

Atrial fibrillation ablation outcome prediction with a machine learning fusion framework incorporating cardiac computed tomography.

BACKGROUND: Structural changes in the left atrium (LA) modestly predict outcomes in patients undergoing catheter ablation for atrial fibrillation (AF). Machine learning (ML) is a promising approach to personalize AF management strategies and improve predictive risk models after catheter ablation by integrating atrial geometry from cardiac computed tomography (CT) scans and patient-specific clinical data. We hypothesized that ML approaches based on a patient's specific data can identify responders to AF ablation. METHODS: Consecutive patients undergoing AF ablation, who had preprocedural CT scans, demographics, and 1-year follow-up data, were included in the study for a retrospective analysis. The inputs of models were CT-derived morphological features from left atrial segmentation (including the shape, volume of the LA, LA appendage, and pulmonary vein ostia) along with deep features learned directly from raw CT images, and clinical data. These were merged intelligently in a framework to learn their individual importance and produce the optimal classification. RESULTS: Three hundred twenty-one patients (64.2 ± 10.6 years, 69% male, 40% paroxysmal AF) were analyzed. Post 10-fold nested cross-validation, the model trained to intelligently merge and learn appropriate weights for clinical, morphological, and imaging data (AUC 0.821) outperformed those trained solely on clinical data (AUC 0.626), morphological (AUC 0.659), or imaging data (AUC 0.764). CONCLUSION: Our ML approach provides an end-to-end automated technique to predict AF ablation outcomes using deep learning from CT images, derived structural properties of LA, augmented by incorporation of clinical data in a merged ML framework. This can help develop personalized strategies for patient selection in invasive management of AF.

Reconfigurable magnon interference by on-chip dynamic wavelength conversion.

Spin waves (SWs), an ultra-low power magnetic excitation in ferro or antiferromagnetic media, have tremendous potential as transport less data carriers for post-CMOS technology using their wave interference properties. The concept of magnon interference originates from optical interference, resulting in a historical taboo of maintaining an identical wavevector for magnon interference-based devices. This makes the attainment of on-chip design reconfigurability challenging owing to the difficulty in phase tuning via external fields. Breaking the taboo, this study explores a novel technique to systematically control magnon interference using asymmetric wavevectors from two different SW modes (magnetostatic surface SWs and backward volume magnetostatic SWs) in a microstructured yttrium iron garnet crossbar. Using this system, we demonstrate phase reconfigurability in the interference pattern by modulating the thermal landscape, modifying the dispersion of the interfering SW modes. Thus, we manifest that such a tunable interference can be used to implement reconfigurable logic gates operating between the XNOR and XOR modes by using symmetric and asymmetric interference, respectively.

Artificial Intelligence Predictive Model for Hormone Therapy Use in Prostate Cancer.

BACKGROUND: Androgen deprivation therapy (ADT) with radiotherapy can benefit patients with localized prostate cancer. However, ADT can negatively impact quality of life, and there remain no validated predictive models to guide its use. METHODS: We used digital pathology images from pretreatment prostate tissue and clinical data from 5727 patients enrolled in five phase 3 randomized trials, in which treatment was radiotherapy with or without ADT, as our data source to develop and validate an artificial intelligence (AI)­derived predictive patient-specific model that would determine which patients would develop the primary end point of distant metastasis. The model used baseline data to provide a binary output that a given patient will likely benefit from ADT or not. After the model was locked, validation was performed using data from NRG Oncology/Radiation Therapy Oncology Group (RTOG) 9408 (n=1594), a trial that randomly assigned men to radiotherapy plus or minus 4 months of ADT. Fine­Gray regression and restricted mean survival times were used to assess the interaction between treatment and the predictive model and within predictive model­positive, i.e., benefited from ADT, and ­negative subgroup treatment effects. RESULTS: Overall, in the NRG/RTOG 9408 validation cohort (14.9 years of median follow-up), ADT significantly improved time to distant metastasis. Of these enrolled patients, 543 (34%) were model positive, and ADT significantly reduced the risk of distant metastasis compared with radiotherapy alone. Of 1051 patients who were model negative, ADT did not provide benefit. CONCLUSIONS: Our AI-based predictive model was able to identify patients with a predominantly intermediate risk for prostate cancer likely to benefit from short-term ADT. (Supported by a grant [U10CA180822] from NRG Oncology Statistical and Data Management Center, a grant [UG1CA189867] from NCI Community Oncology Research Program, a grant [U10CA180868] from NRG Oncology Operations, and a grant [U24CA196067] from NRG Specimen Bank from the National Cancer Institute and by Artera, Inc. ClinicalTrials.gov numbers NCT00767286, NCT00002597, NCT00769548, NCT00005044, and NCT00033631.)

OPN promotes pro-inflammatory cytokine expression via ERK/JNK pathway and M1 macrophage polarization in Rosacea.

Rosacea is a chronic inflammatory dermatosis that involves dysregulation of innate and adaptive immune systems. Osteopontin (OPN) is a phosphorylated glycoprotein produced by a broad range of immune cells such as macrophages, keratinocytes, and T cells. However, the role of OPN in rosacea remains to be elucidated. In this study, it was found that OPN expression was significantly upregulated in rosacea patients and LL37-induced rosacea-like skin inflammation. Transcriptome sequencing results indicated that OPN regulated pro-inflammatory cytokines and promoted macrophage polarization towards M1 phenotype in rosacea-like skin inflammation. In vitro, it was demonstrated that intracellular OPN (iOPN) promoted LL37-induced IL1B production through ERK1/2 and JNK pathways in keratinocytes. Moreover, secreted OPN (sOPN) played an important role in keratinocyte-macrophage crosstalk. In conclusion, sOPN and iOPN were identified as key regulators of the innate immune system and played different roles in the pathogenesis of rosacea.

Comparative study on alleviating effect of kiwi berry (Actinidia arguta) polysaccharide and polyphenol extracts on constipated mice.

Kiwi berry (Actinidia arguta) is beneficial for relieving constipation, but the mechanism of easing constipation is still unknown. The alleviating effects of kiwi berry polysaccharide and polyphenol extracts on loperamide induced constipation were studied. Administration with polysaccharide extract of kiwi berry in loperamide-induced constipation mice distinctly decreased the body weight gain by 124.0%, the number and the water content of feces was decreased by 152.4% and 107.0% respectively, gastrointestinal (GI) transit rate was decreased by 39.5% and the time to the first dark stool was largen by 56.2% as compared with those in the loperamide group, respectively. The levels of excitability neurotransmitters were increased, and the inhibitory neurotransmitter was decreased in the kiwi berry extracts groups compared with the loperamide group. The levels of aquaporins were decreased to ameliorate constipation. Moreover, kiwi berry extracts can protect colon smooth muscle cells from apoptosis and help to restore colon health. Interstitial cells of Cajal (ICC) and animal experiments suggested that kiwi berry extracts can up-regulate the expression levels of stem cell factors (SCF)/c-kit protein. Kiwi berry can remodel the structure of microbial communities. All findings suggest that kiwi berry polysaccharide and polyphenol especially its polysaccharide extract, can effectively alleviate constipation induced by loperamide. Kiwi berry is a promising food supplement that can be used to relieve constipation.

Graph-based Fusion Modeling and Explanation for Disease Trajectory Prediction.

We propose a relational graph to incorporate clinical similarity between patients while building personalized clinical event predictors with a focus on hospitalized COVID-19 patients. Our graph formation process fuses heterogeneous data, i.e., chest X-rays as node features and non-imaging EHR for edge formation. While node represents a snap-shot in time for a single patient, weighted edge structure encodes complex clinical patterns among patients. While age and gender have been used in the past for patient graph formation, our method incorporates complex clinical history while avoiding manual feature selection. The model learns from the patient's own data as well as patterns among clinically-similar patients. Our visualization study investigates the effects of 'neighborhood' of a node on its predictiveness and showcases the model's tendency to focus on edge-connected patients with highly suggestive clinical features common with the node. The proposed model generalizes well by allowing edge formation process to adapt to an external cohort.

An updated review on the stability of anthocyanins regarding the interaction with food proteins and polysaccharides.

The health benefits of anthocyanins are compromised by their chemical instability and susceptibility to external stress. Researchers found that the interaction between anthocyanins and macromolecular components such as proteins and polysaccharides substantially determines the stability of anthocyanins during food processing and storage. The topic thus has attracted much attention in recent years. This review underlines the new insights gained in our current study of physical and chemical properties and functional properties in complex food systems. It examines the interaction between anthocyanins and food proteins or polysaccharides by focusing on the "structure-stability" relationship. Furthermore, multispectral and molecular computing simulations are used as the chief instruments to explore the interaction's mechanism. During processing and storage, the stability of anthocyanins is generally influenced by the adverse characteristics of food and beverage, including temperature, light, oxygen, enzymes, pH. While the action modes and types between protein/polysaccharide and anthocyanins mainly depend on their structures, the noncovalent interaction between them is the key intermolecular force that increases the stability of anthocyanins. Our goal is to provide the latest understanding of the stability of anthocyanins under food processing conditions and further improve their utilization in food industries. Practical Application: This review provides support for the steady-state protection of active substances.

Machine Learning-Enabled Multimodal Fusion of Intra-Atrial and Body Surface Signals in Prediction of Atrial Fibrillation Ablation Outcomes.

BACKGROUND: Machine learning is a promising approach to personalize atrial fibrillation management strategies for patients after catheter ablation. Prior atrial fibrillation ablation outcome prediction studies applied classical machine learning methods to hand-crafted clinical scores, and none have leveraged intracardiac electrograms or 12-lead surface electrocardiograms for outcome prediction. We hypothesized that (1) machine learning models trained on electrograms or electrocardiogram (ECG) signals can perform better at predicting patient outcomes after atrial fibrillation ablation than existing clinical scores and (2) multimodal fusion of electrogram, ECG, and clinical features can further improve the prediction of patient outcomes. METHODS: Consecutive patients who underwent catheter ablation between 2015 and 2017 with panoramic left atrial electrogram before ablation and clinical follow-up for at least 1 year following ablation were included. Convolutional neural network and a novel multimodal fusion framework were developed for predicting 1-year atrial fibrillation recurrence after catheter ablation from electrogram, ECG signals, and clinical features. The models were trained and validated using 10-fold cross-validation on patient-level splits. RESULTS: One hundred fifty-six patients (64.5±10.5 years, 74% male, 42% paroxysmal) were analyzed. Using electrogram signals alone, the convolutional neural network achieved an area under the receiver operating characteristics curve (AUROC) of 0.731, outperforming the existing APPLE scores (AUROC=0.644) and CHA2DS2-VASc scores (AUROC=0.650). Similarly using 12-lead ECG alone, the convolutional neural network achieved an AUROC of 0.767. Combining electrogram, ECG, and clinical features, the fusion model achieved an AUROC of 0.859, outperforming single and dual modality models. CONCLUSIONS: Deep neural networks trained on electrogram or ECG signals improved the prediction of catheter ablation outcome compared with existing clinical scores, and fusion of electrogram, ECG, and clinical features further improved the prediction. This suggests the promise of using machine learning to help treatment planning for patients after catheter ablation.

Preparing for the next pandemic via transfer learning from existing diseases with hierarchical multi-modal BERT: a study on COVID-19 outcome prediction.

Developing prediction models for emerging infectious diseases from relatively small numbers of cases is a critical need for improving pandemic preparedness. Using COVID-19 as an exemplar, we propose a transfer learning methodology for developing predictive models from multi-modal electronic healthcare records by leveraging information from more prevalent diseases with shared clinical characteristics. Our novel hierarchical, multi-modal model ([Formula: see text]) integrates baseline risk factors from the natural language processing of clinical notes at admission, time-series measurements of biomarkers obtained from laboratory tests, and discrete diagnostic, procedure and drug codes. We demonstrate the alignment of [Formula: see text]'s predictions with well-established clinical knowledge about COVID-19 through univariate and multivariate risk factor driven sub-cohort analysis. [Formula: see text]'s superior performance over state-of-the-art methods shows that leveraging patient data across modalities and transferring prior knowledge from similar disorders is critical for accurate prediction of patient outcomes, and this approach may serve as an important tool in the early response to future pandemics.

Sensitivity enhancement in magnetic sensor using CoFeB/Y3Fe5O12 resonator.

Magnonics, an emerging research field that uses the quanta of spin waves as data carriers, has a potential to dominate the post-CMOS era owing to its intrinsic property of ultra-low power operation. Spin waves can be manipulated by a wide range of parameters; thus, they are suitable for sensing applications in a wide range of physical fields. In this study, we designed a highly sensitive, simple structure, and ultra-low power magnetic sensor using a simple CoFeB/Y3Fe5O12 bilayer structure. We demonstrated that the CoFeB/Y3Fe5O12 bilayer structure can create a sharp rejection band in its spin-wave transmission spectra. The lowest point of this strong rejection band allows the detection of a small frequency shift owing to the external magnetic field variation. Experimental observations revealed that such a bilayer magnetic sensor exhibits 20 MHz frequency shifts upon the application of an external magnetic field of 0.5 mT. Considering the lowest full width half maximum, which is about 2 MHz, a sensitivity of 10-2 mT order can be experimentally achieved. Furthermore, the higher sensitivity in the order of 10-6 T (µT) has been demonstrated using the sharp edge of the rejection band of the CoFeB/Y3Fe5O12 bilayer device. A Y-shaped spin waves interference device with two input arms consisting of CoFeB/Y3Fe5O12 and Y3Fe5O12 has been theoretically investigated. We proposed that such a structure can demonstrate a magnetic sensitivity in the range of [Formula: see text] T (nT) at room temperature. The sensitivity of the sensor can be further enhanced by tuning the width of the CoFeB metal stripe.

Graph-based Fusion Modeling and Explanation forDisease Trajectory Prediction.

We propose a relational graph to incorporate clinical similarity between patients while building personalized clinical event predictors with a focus on hospitalized COVID-19 patients. Our graph formation process fuses heterogeneous data, i.e., chest X-rays as node features and non-imaging EHR for edge formation. While node represents a snap-shot in time for a single patient, weighted edge structure encodes complex clinical patterns among patients. While age and gender have been used in the past for patient graph formation, our method incorporates complex clinical history while avoiding manual feature selection. The model learns from the patient's own data as well as patterns among clinically-similar patients. Our visualization study investigates the effects of 'neighborhood' of a node on its predictiveness and showcases the model's tendency to focus on edge-connected patients with highly suggestive clinical features common with the node. The proposed model generalizes well by allowing edge formation process to adapt to an external cohort.