Ankit Agrawal

India’s ambitious renewable energy targets and its vision for energy self-reliance critically depend on a resilient and competitive solar photovoltaic (PV) manufacturing ecosystem. However, India’s heavy reliance on imported solar modules, primarily from China, poses significant risks to energy security, economic stability, and long-term sustainability. This study investigates the Indian Solar PV manufacturing sector to identify structural gaps and propose actionable strategies for achieving self-reliance. The paper offers a multidimensional assessment of the industry’s competitiveness and supply chain vulnerabilities by employing a hybrid research methodology combining Porter’s Five Forces Framework and expert interviews. The findings reveal that traditional localization strategies have not sufficiently empowered domestic manufacturers to scale, integrate, or compete globally. In response, the study introduces the concept of deep localization. This strategic framework extends beyond surface-level sourcing to include vertical and horizontal integration, Micro, Small, and Medium Enterprises participation, innovation, regulatory coherence, and socio-cultural alignment. The research outlines a phased roadmap for operationalizing deep localization: Phase 1 focuses on foundation building, Phase 2 strengthens the supply chain and technology, and Phase 3 emphasises system efficiency and sustainability. The qualitative nature of this study provides an in-depth assessment of complex structural and strategic dimensions within India’s solar PV sector. It also complements the methodological rigour by offering a conceptual foundation for future quantitative validation. The paper concludes that deep localization is not just an industrial strategy but a transformative approach to self-reliance and enhancing global competitiveness, aligning with India’s sustainable goals.

Mixture-of-Experts (MoE) architectures achieve scalable learning by routing inputs to specialized subnetworks through conditional computation. However, conventional MoE designs assume homogeneous expert capability and domain-agnostic routing-assumptions that are fundamentally misaligned with medical imaging, where anatomical structure and regional disease heterogeneity govern pathological patterns. We introduce Regional Expert Networks (REN), the first anatomically-informed MoE framework for medical image classification. REN encodes anatomical priors by training seven specialized experts, each dedicated to a distinct lung lobe or bilateral lung combination, enabling precise modeling of region-specific pathological variation. Multi-modal gating mechanisms dynamically integrate radiomics biomarkers with deep learning (DL) features extracted by convolutional (CNN), Transformer (ViT), and state-space (Mamba) architectures to weight expert contributions at inference. Applied to interstitial lung disease (ILD) classification on a 597-patient, 1,898-scan longitudinal cohort, REN achieves consistently superior performance: the radiomics-guided ensemble attains an average AUC of 0.8646 ± 0.0467, a +12.5% improvement over the SwinUNETR single-model base-line (AUC 0.7685, p = 0.031). Lower-lobe experts reach AUCs of 0.88-0.90, outperforming DL baselines (CNN: 0.76-0.79) and mirroring known patterns of basal ILD progression. Evaluated under rigorous patient-level cross-validation, REN demonstrates strong generalizability and clinical interpretability, establishing a scalable, anatomically-guided framework potentially extensible to other structured medical imaging tasks. Code is available on our GitHub https://github.com/NUBagciLab/MoE-REN.

Small Unmanned Aerial Systems (sUAS) are increasingly deployed across a wide range of applications, yet ensuring their safe and reliable performance in complex wind conditions remains a major challenge. Existing simulation platforms, such as AirSim and Gazebo, often fail to capture the unpredictable wind dynamics that occur in real-world environments, where variations in building structures and natural terrains strongly influence airflow. Turbulent wind flows, characterized by gusts, shears, and rapid directional changes, are among the leading factors contributing to sUAS malfunctions and accidents. Consequently, there is a clear need for advanced simulation tools and testing methods that allow developers to evaluate how unpredictable wind conditions affect sUAS performance early in the development process. To address this need, we present DroneWind Simulation (DroneWiS), a novel simulation testing framework that leverages Computational Fluid Dynamics (CFD) to generate realistic wind flow patterns within digital twin environments. DroneWiS enables developers to systematically test and assess the impact of diverse wind scenarios on sUAS stability and flight behavior throughout the design and validation lifecycle. Comparative studies show that DroneWiS produces more realistic flight responses to complex wind conditions than current state-of-the-art platforms. Scenario-based testing allows DroneWiS to recreate real-world incidents influenced by wind and to support identification of performance limitations and safety risks during development. An initial comparison with measured wind data at one location across two timestamps indicates pipeline feasibility, while broader validation remains for future work.

Determining lattice parameters in two-dimensional (2D) materials is essential for materials characterization and discovery. In this work, we propose a deep-learning-driven pipeline that addresses the regression task of estimating the lattice constants a and b and the angle γ directly from 2D images using computer vision. We evaluate our approach on three different 2D material data sets: JARVIS-2D (JV2D) and Computational 2D Materials Database (C2DB), and a newly created data set derived from the Alexandria database. Multiple architectures are compared, including DenseNet121, Vision Transformers (ViT-L/14) paired with Multi Layer Perceptron (MLP) heads, and GoogleNet. DenseNet121 achieves accurate performance, with mean absolute errors as low as 0.18 Å for the Alexandria-based data set and 0.17 Å for C2DB and 0.59 Å for JV2D, as well as up to 96% accuracy in classifying Bravais lattice types for the Alexandria-based data set.

Thermal management challenges in modern electronics have motivated the use of nanofluids, colloidal suspensions of nanoparticles in a base fluid. They have high thermal properties such as high thermal conductivity and high heat transfer coefficients, which are suitable in the area of applications that require efficient heat dissipation. In this chapter, the use of nanofluids in different electronics cooling applications is explored including microprocessors, power electronics, light-emitting diodes, data centers, batteries, and 3D integrated circuits. By effectively managing heat, nanofluids help maintain optimal operating temperatures, improve device performance, and extend the lifespan of electronic components. However, challenges, such as stability, cost, and material compatibility, must be overcome to fully realize their potential. With the increasing thermal demands in industry, the adoption of nanofluids in cooling systems offers a pathway to meet the thermal demands in electronic devices becoming more powerful and more compact.

Stress–strain curves capture the mechanical behavior of materials but are computationally expensive to generate using crystal-plasticity finite-element (CPFE) models, due to the profound nonlinearity of the response, and its relationship to crystallographic orientation. We propose an AI-driven framework for predicting bilinear approximate stress–strain curves of metallic alloys using supervised machine learning models. We evaluate its performance on three representative materials: aluminum (Al), nickel (Ni), and copper (Cu), commonly used in aerospace engineering. Trained on just 100 fully resolved CPFE curves per material, our model accurately reconstructs entire curves using features extracted after only a single CPFE step, effectively leading to orders of magnitude speedup with respect to CPFE simulation for predicting stress-strain curves of new orientations unseen by the AI model. The resulting predictions achieve a mean absolute error fraction of around 1.53% for nickel, 1.43% for aluminum, and 2.86% for copper, while producing orientation-specific stress–strain curves several times faster than conventional CPFE simulations.

High-level I/O libraries, such as PnetCDF and HDF5, are commonly used by large-scale scientific applications to perform I/O tasks in parallel. These I/O libraries store the metadata of data objects in files along with their raw data. To ensure metadata consistency during parallel data object creation, they require applications to call the metadata APIs collectively using consistent metadata. Such a requirement can result in an expensive consistency check, as its cost increases with the metadata volume and the number of processes. To address this limitation, we propose a new file header format, which uses partitioned metadata blocks to enable independent data object creation and reduce the objects required for consistency check. Our performance evaluation shows that this new design achieves a scalable performance, cutting data object creation times by up to 196 × when running on 4096 MPI processes to create 5,684,800 data objects in parallel.

Background: Cardiogenic shock (CS) remains a life-threatening complication in heart failure (HF), often requiring rapid hemodynamic support. Mechanical circulatory support (MCS) devices including intra-aortic balloon pumps (IABP), percutaneous ventricular assist devices (pVADs), and extracorporeal membrane oxygenation (ECMO) have reshaped acute care, yet contemporary national data on utilization trends, outcomes, and disparities in HF-related CS are limited. Objective: To evaluate national trends, outcomes, and disparities in MCS use among patients hospitalized with HF complicated by CS from 2016–2022. Methods: We conducted a retrospective cohort study using the National Inpatient Sample (2016–2022). Adult hospitalizations with a diagnosis of HF and concurrent CS were identified using validated ICD-10-CM codes. MCS use (IABP, Impella, ECMO) was captured and stratified by year. Outcomes included in-hospital mortality, length of stay (LOS), and inflation-adjusted hospitalization costs. Multivariable regression adjusted for demographics, comorbidities, and hospital characteristics. Kaplan-Meier analysis was performed for hospitalizations >30 days. Results: Among 265,910 weighted HF-CS admissions, MCS use increased from 11.4% in 2016 to 16.5% in 2022 (p<0.001). In-hospital mortality remained high but was lower with MCS (31.8%) vs. non-MCS (39.7%). Kaplan-Meier analysis showed improved survival for MCS patients with prolonged hospital stays (log-rank p=0.036). MCS use was associated with longer LOS (13 vs. 10 days, p<0.001) and higher costs ($84,000 vs. $41,000, p<0.001). MCS was more frequently used in younger patients (67.6 vs. 69.7 years, p<0.001) with fewer comorbidities. However, in-hospital deaths were more common in older patients (71 vs. 68 years, p<0.001). Use was higher in urban-teaching hospitals, the Midwest, and among White patients and those in higher income quartiles (p<0.05). Use was lowest in rural and small hospitals. Mechanical ventilation was the strongest mortality predictor in MCS recipients (aOR 2.45; 95% CI, 2.27–2.64). Conclusions: MCS use in HF-related CS has increased nationally, but mortality benefits remain modest, especially in older and ventilated patients. Significant disparities by geography, race, and income persist. These findings underscore the need for equitable access, improved patient selection, and prospective trials to define optimal MCS strategies.

Background: Diabetes mellitus is a known risk factor for adverse cardiovascular outcomes, particularly in patients with atrial fibrillation. While left atrial appendage occlusion (LAAO) has emerged as a viable alternative to long-term anticoagulation, the influence of glycemic control on post-procedural outcomes remains inadequately defined. This study aimed to evaluate cardiovascular outcomes among nondiabetic (NDM), well-controlled diabetic (WCDM; HbA1c <7%), and poorly controlled diabetic (PCDM; HbA1c ≥7%) patients undergoing LAAO. Research Question: Does diabetes and its glycemic control status affect adverse cardiovascular outcomes following LAAO in atrial fibrillation patients? Methods: Using the TriNetX Global Collaborative Network, we identified adult AF patients who underwent LAAO between 2018 and 2024. Patients were stratified into three comparisons: Diabetics vs NDM, WCDM vs NDM, and PCDM vs NDM. Propensity score matching (PSM) was employed to balance 26 baseline covariates across cohorts. Outcomes including major adverse cardiovascular events (MACE), stroke and all-cause mortality were assessed over a 5-year follow-up. Risk analysis and Kaplan-Meier survival curves were used to derive hazard ratios (HR) and 95% confidence intervals (CI). Results: In the overall analysis, major adverse cardiovascular events (MACE) occurred in 13.4% of diabetics versus 9.0% of non-diabetics (HR, 1.44; 95% CI, 1.28–1.62) and all-cause mortality in 15.2% versus 12.3% (HR, 1.15; 95% CI, 1.05–1.25). Stroke occurred in 5.1% versus 3.6% (HR, 1.30; 95% CI, 1.09–1.54), pulmonary embolism in 2.8% versus 2.1% (HR, 1.25; 95% CI, 1.02–1.54), and ventricular fibrillation in 1.0% versus 0.5% (HR, 1.81; 95% CI, 1.22–2.70). In the HbA1c >7% subgroup, MACE risk was further elevated (HR, 1.59; 95% CI, 1.25–2.00), whereas diabetics with HbA1c <7% still experienced increased risk (HR, 1.39; 95% CI, 1.21–1.60) compared with non-diabetics. Conclusion: Poor glycemic control (HbA1c ≥7%) in diabetic patients with atrial fibrillation undergoing LAAO is associated with a significantly increased long-term risk of MACE, stroke, and mortality. Notably, both suboptimal and even controlled glycemic states appear to confer elevated cardiovascular risk, underscoring the complex interplay between diabetes and outcomes post-LAAO. These findings highlight the critical importance of individualized glycemic management and comprehensive cardiovascular risk reduction strategies in this high-risk population.

Medical Publisher in Delhi, India - IP Innovative Publication Pvt Limited, Medical Journals Publication, Open Access Journals, Print Journals,Indian Journals, Surgery, Microbiology, Anaesthesia, Anatomy Biochemistry Books and Journals, Skills for Employability,Best Journal Publishers-IP Innovative Publication Pvt Limit

Interstitial lung disease (ILD) is present in over 60% of patients with systemic sclerosis (SSc) and is the leading cause of SSc-related deaths. Profibrotic monocyte-derived alveolar macrophages (MoAM) play a causal role in the pathogenesis of pulmonary fibrosis in animal models where their persistence in the niche requires signaling through Colony Stimulating Factor 1 Receptor (CSF1R). We hypothesized that the presence and proportion of MoAM in bronchoalveolar lavage (BAL) fluid from patients with SSc-ILD may be a biomarker of ILD severity. To test this hypothesis, we analyzed BAL fluid from 9 prospectively enrolled patients with SSc-ILD and 13 healthy controls using flow cytometry and single-cell RNA sequencing. Patients with SSc-ILD had more MoAM and interstitial macrophages in BAL fluid than healthy controls, and their abundance was associated with lung fibrosis severity. We identified changes in the MoAM transcriptome as a function of treatment with mycophenolate, an effective therapy for SSc-ILD. In SSc-ILD lung explants, spatial transcriptomics identified an expanded population of interstitial macrophages spilling into the alveolar space. Our findings suggest that the proportion of profibrotic MoAM and interstitial macrophages in BAL fluid may serve as a biomarker of SSc-ILD and credential them as possible targets for therapy.

OBJECTIVE: Rule-based structured data algorithms and natural language processing (NLP) approaches applied to unstructured clinical notes have limited accuracy and poor generalizability for identifying immunosuppression. Large language models (LLMs) may effectively identify patients with heterogenous types of immunosuppression from unstructured clinical notes. We compared the performance of LLMs applied to unstructured notes for identifying patients with immunosuppressive conditions or immunosuppressive medication use against 2 baselines: (1) structured data algorithms using diagnosis codes and medication orders and (2) NLP approaches applied to unstructured notes. MATERIALS AND METHODS: We used hospital admission notes from a primary cohort of 827 intensive care unit (ICU) patients at Northwestern Memorial Hospital and a validation cohort of 200 ICU patients at Beth Israel Deaconess Medical Center, along with diagnosis codes and medication orders from the primary cohort. We evaluated the performance of structured data algorithms, NLP approaches, and LLMs in identifying 7 immunosuppressive conditions and 6 immunosuppressive medications. RESULTS: In the primary cohort, structured data algorithms achieved peak F1 scores ranging from 0.30 to 0.97 for identifying immunosuppressive conditions and medications. NLP approaches achieved peak F1 scores ranging from 0 to 1. GPT-4o outperformed or matched structured data algorithms and NLP approaches across all conditions and medications, with F1 scores ranging from 0.51 to 1. GPT-4o also performed impressively in our validation cohort (F1 = 1 for 8/13 variables). DISCUSSION: LLMs, particularly GPT-4o, outperformed structured data algorithms and NLP approaches in identifying immunosuppressive conditions and medications with robust external validation. CONCLUSION: LLMs can be applied for improved cohort identification for research purposes.

Criteria to identify patients who are ready to be liberated from mechanical ventilation (MV) are imprecise, often resulting in prolonged MV or reintubation, both of which are associated with adverse outcomes. Daily protocol-driven assessment of the need for MV leads to earlier extubation but requires dedicated personnel. We sought to determine whether machine learning (ML) applied to the electronic health record could predict next-day extubation. We examined 37 clinical features aggregated from 12AM-8AM on each patient-ICU-day from a single-center prospective cohort study of patients in our quaternary care medical ICU who received MV. We also tested our models on an external test set from a community hospital ICU in our health care system. We used three data encoding/imputation strategies and built XGBoost, LightGBM, logistic regression, LSTM, and RNN models to predict next-day extubation. We compared model predictions and actual events to examine how model-driven care might have differed from actual care. Our internal cohort included 448 patients and 3,095 ICU days, and our external test cohort had 333 patients and 2,835 ICU days. The best model (LSTM) predicted next-day extubation with an AUROC of 0.870 (95% CI 0.834-0.902) on the internal test cohort and 0.870 (95% CI 0.848-0.885) on the external test cohort. Across multiple model types, measures previously demonstrated to be important in determining readiness for extubation were found to be most informative, including plateau pressure and Richmond Agitation Sedation Scale (RASS) score. Our model often predicted patients to be stable for extubation in the days preceding their actual extubation, with 63.8% of predicted extubations occurring within three days of true extubation. Our findings suggest that an ML model may serve as a useful clinical decision support tool rather than complete replacement of clinical judgement. However, any ML-based model should be compared with protocol-based practice in a prospective, randomized controlled trial to determine improvement in outcomes while maintaining safety as well as cost effectiveness.

Wilson's disease (WD) is a hepatolenticular degeneration caused by mutations in the ATP7B gene.It is an autosomal recessive disorder resulting in excessive copper deposition, particularly in liver, eyes, and brain.This occurs due to a deficiency of ceruloplasmin, a copper binding protein, often leading to liver cirrhosis.A characteristic ocular sign of Wilson disease is the Kayser-Fleischer (KF) ring, represent copper deposition in Descemet's membrane of the cornea.KF ring is most common in neuro WD, detected in 78-85% of cases.Its prevalence in hepatic WD ranges from 36-62%, while it may be detected in 10% of asymptomatic patients. 1Traditionally KF rings are detected using slit-lamp biomicroscopy.However anterior segment optical coherence tomography (AS-OCT) has emerged as a promising tool, offering a non-invasive, objective, high-resolution cross-sectional imaging tool that can detect subtle KF ring especially in early stage or uncooperative patients challenges with slit lamp examination Copper deposits primarily in the anterior chamber angle at Schwalbe's line within the Descemet's membrane.Early detection with a standard slit-lamp is challenging since the angle view is obscured by the corneal limbus.Copper deposits only become visible after crossing the limbus. role of as-octAS-OCT is a high-resolution non-contact optical imaging technique using infrared light (wavelength 1310 nm) permitting greater penetration, allowing better visualisation of the structures at the angle.The KF ring appears as a distinct hyperreflective band at the level of the Descemet's membrane on AS-OCT.It displays greenish-yellow to brownish coloration on the imaging scale, signifying copper accumulation at the corneal periphery.AS-OCT captures high-resolution, reproducible images without requiring corneal anaesthesia, with each scan taking less than 20 seconds

Severe community-acquired pneumonia (CAP) remains a major cause of critical illness, yet there are no validated early clinical criteria to predict short-term treatment outcomes in these patients. Short-term pneumonia treatment outcomes are less affected by confounding factors introduced by a prolonged hospital course, and early prediction of short-term treatment outcomes can help physicians identify those who are likely to fail the current treatment and implement adjustments to existing diagnostic and therapeutic plans. Traditional clinical stability criteria such as Halm's criteria are not calibrated for early outcome prediction in critically ill severe pneumonia patients. We applied the XGBoost algorithm to predict pneumonia cure by day 7-8 post-intubation with clinical features from days 1-3 in mechanically ventilated patients with severe CAP from the Successful Clinical Response in Pneumonia Therapy (SCRIPT) study, a prospective cohort study at a tertiary academic center. Pneumonia episodes were adjudicated for day 7-8 cure status by a panel of critical care physicians using a structured review process. Clinical features that inform Halm's criteria, including vital signs, oxygenation parameters, mental status, and vasopressor use, were extracted from the electronic health record. We also examined model performance by including additional features, such as laboratory data, ventilator settings, and medications. Basic demographic characteristics including age and BMI were also incorporated. Among 85 patients, 42 (49.4%) were cured by day 7-8. The best-performing model, which used Halm's clinical features and ventilator features from days 1-3, achieved a cross-validated AUROC of 0.757. Inclusion of lab and medication data did not significantly improve performance. Key predictors included GCS, norepinephrine requirement, and BMI. We prove the feasibility of using ML models to predict short-term treatment outcomes of severe CAP among critically ill patients with basic clinical features. Future studies should focus on external validation and clinical integration to inform prognosis and early reevaluation of treatment strategy in patients with predicted poor outcomes.

In this study, we present an artificial intelligence (AI)-driven framework for predicting the microstructural texture of polycrystalline materials after a specific deformation process. The microstructural texture is defined in terms of the orientation distribution function (ODF) which indicates the volume density of crystal orientations. Our approach leverages an encoder-decoder model with Long Short-Term Memory (LSTM) layers to model the relationship between processing conditions and material properties. As a case study, we apply our framework to copper, generating a dataset of 3125 unique processing parameter combinations and their corresponding ODF vectors. The resulting predictions enable the calculation of homogenized properties. Our AI-driven framework outperforms traditional material processing simulations, yielding faster results with limited error rates (< 0.3% for both the elastic matrix C and the compliance matrix S), making it a promising tool for the expedited design of microstructures with tailored properties.

Understanding the large deformation behavior of materials under external forces is crucial for reliable engineering applications. The mechanical properties of materials depend on their underlying microstructures, which change over time during deformation. Experimental observation of these processes is time-consuming and influenced by various conditions. Therefore, we developed MicroProcSim, a physics-based simulation tool to replicate the deformation process of cubic microstructures. MicroProcSim can predict the evolution of texture, represented by the orientation distribution function (ODF), over time under various loads and strain rates. This software package can be run on both Windows and Linux operating systems. Unlike conventional crystal plasticity finite element software, MicroProcSim offers a distinct advantage by rapidly generating deformed textures, as it bypasses incorporating grain morphology. Furthermore, comparisons with existing experimental and computational studies on texture evolution have demonstrated that this software seamlessly replicates real-world material processing conditions through a simple modification of a single input matrix. Editor’s Video Summary: The online version of this article (10.1007/s40192-025-00405-6) contains an Editor's Video Summary, which is available to authorized users.

High-level I/O libraries, such as HDF5 and PnetCDF, are commonly used by large-scale scientific applications to perform I/O tasks in parallel. These I/O libraries store the metadata such as data types and dimensionality along with the raw data in the same files. While these libraries are well-optimized for concurrent access to the raw data, they are designed neither to handle a large number of data objects efficiently nor to create different data objects independently by multiple processes, as they require applications to call data object creation APIs collectively with consistent metadata among all processes. Applications that process data gathered from remote sensors, such as particle collision experiments in high-energy physics, may generate data of different sizes from different sensors and desire to store them as separate data objects. For such applications, the I/O library's requirement on collective data object creation can become very expensive, as the cost of metadata consistency check increases with the metadata volume as well as the number of processes. To address this limitation, using PnetCDF as an experimental platform, we investigate solutions in this paper that abide the netCDF file format, as well as propose a new file header format that enables independent data object creation. The proposed file header consists of two sections, an index table and a list of metadata blocks. The index table contains the reference to the metadata blocks and each block stores metadata of objects that can be created collectively or independently. The new design achieves a scalable performance, cutting data object creation times by up to 582x when running on 4096 MPI processes to create 5,684,800 data objects in parallel. Additionally, the new method reduces the memory footprints, with each process requiring an amount of memory space inversely proportional to the number of processes.

Abstract Background Interstitial lung disease (ILD) affects &gt; 40% of patients with systemic sclerosis (SSc/scleroderma) and is the leading cause of disease-related mortality. Although therapies may slow progression, outcomes remain poor, partly because ILD is often detected after irreversible lung injury has occurred. Although chest computed tomography (CT) is a sensitive tool for ILD detection and is recommended at SSc diagnosis, it is oftentimes not performed and even less often performed serially. We sought to develop tools to predict ILD and mortality in patients with SSc using data routinely available in the electronic health record (EHR) to inform medical decision-making. Methods We analyzed longitudinal EHR data from two SSc cohorts: Northwestern University (1,169 participants; derivation cohort) and Yale University (376 participants; validation cohort). We identified clinical features from existing cohort-linked EHR queries composing a convenience sample of data from participants spanning decades rather than employing a single unified data collection effort. Three ILD experts independently reviewed CT reports and classified each as having or lacking ILD. To explore derivation cohort data structure, patients with &gt; =3 forced vital capacity (FVC) results available were identified and stratified according to prevalent or absent ILD. Using unsupervised trajectory-based clustering exploratory analyses, we determined standardized patterns across groups. ML models were then developed using clinical EHR data as predictor variables and prevalent ILD and all-cause mortality as outcome variables. Model performance was assessed using area under the receiver operating characteristic curve (AUC). Results Seventy-four clinical features with low missingness, including demographic, vital sign, laboratory, and pulmonary function test data, were utilized for analyses. Four robust PFT trajectory clusters were identified that were associated with ILD prevalence and mortality in exploratory analyses. A ML model for ILD detection achieved an AUC of 0.832 and retained performance in the Yale cohort (AUC 0.754). In addition to established predictors such as autoantibodies and pulmonary function, the model identified routine laboratory measurements, including red cell distribution width (RDW), white blood cell count, and serum chloride, as important contributors. One-year mortality prediction achieved AUCs of 0.904 in the North-western cohort and 0.910 in the Yale cohort. Among patients with SSc-ILD, one-year mortality was predicted with AUCs of 0.744 and 0.902 in the Northwestern and Yale cohorts, respectively. Unexpectedly, we found that subtle laboratory abnormalities (such as change in RDW) contributed to predicting mortality. Conclusions Our prediction models comprised of widely available EHR data are useful tools to identify SSc patients at high risk for prevalent ILD and all-cause mortality. Integration of these models into clinical practice could enable scalable risk stratification and inform individualized ILD screening and monitoring strategies for SSc patients.

Background: Interstitial lung disease (ILD) affects 40-75% of patients with systemic sclerosis (SSc) and is the leading cause of SSc-related deaths. In murine models, profibrotic monocyte-derived alveolar macrophages (MoAM) play a causal role in the pathogenesis of pulmonary fibrosis. In humans, profibrotic MoAM have been identified in explanted lung tissue from patients with fibrotic ILD yet are absent in healthy donor lung. Objectives: We aimed to determine the localization of these profibrotic MoAM in lung explants from patients with SSc-ILD and to assess the association between profibrotic MoAM number and phenotype and SSc-ILD disease severity. Methods: Explants from patients with SSc-ILD were profiled with the Xenium spatial transcriptomic assay. Nine subjects with SSc-ILD and thirteen lung donors were recruited from three academic centers. Clinical, spirometric, and radiologic characteristics of those with SSc-ILD were obtained. Subjects underwent flexible fiberoptic bronchoscopy with bronchoalveolar lavage (BAL). Cell populations from BAL fluid (BALF) were analyzed by single-cell RNA sequencing. Their proportions and transcriptional profiles were tested for association with lung function and extent of pulmonary fibrosis on chest high-resolution computed tomography (HRCT) as quantified by the Kazerooni method, which quantifies lung fibrosis and inflammation. Results: Single-cell spatial transcriptomic analysis of explanted lung tissue from patients with SSc-ILD demonstrated that profibrotic MoAM localized exclusively to airspaces. Single-cell RNA-sequencing analysis of immune lung cells from BALF showed increased proportions of profibrotic MoAM in patients with SSc-ILD compared to healthy controls (Figure 1A, 1B). The abundance of these MoAM inversely correlated with lung function (Figure 1C) and positively correlated with the extent of fibrosis on HRCT (Figure 1D). Additionally, we identified genes that significantly correlated with lung function across all alveolar macrophages (AM) subsets and could potentially be translated into clinical biomarkers of disease severity. Differential expression analysis between patients with SSc-ILD and controls revealed significant transcriptomic changes in tissue-resident alveolar macrophages, MoAM, and T cell subsets. Furthermore, we identified a transcriptomic signature in SSc-ILD AM, T cells and dendritic cells that was associated with current, compared to previous or absent, mycophenolate treatment. We detected reduction of interferon gamma gene expression in T cells in patients with current mycophenolate treatment. Bioinformatic analysis suggested that profibrotic MoAM represent a distinct differentiation branch rather than a stalled intermediate state between monocytes and mature AM. Conclusion: Profibrotic MoAM are localized to the airspace of patients with SSc-ILD and can be safely sampled by BAL. Abundance of these MoAM and gene expression profiles are associated with functional and radiologic disease severity. Our data suggest that profibrotic MoAM may serve as a biomarker of SSc-ILD and a potential therapeutic target. REFERENCES: [1] Denton CP, Wells AU, Coghlan JG. Major lung complications of systemic sclerosis. Nat Rev Rheumatol. 2018 Sep;14(9):511–527. PMID: 30111804. [2] Khanna D, Tashkin DP, Denton CP, Renzoni EA, Desai SR, Varga J. Etiology, Risk Factors, and Biomarkers in Systemic Sclerosis with Interstitial Lung Disease. Am J Respir Crit Care Med. 2020 Mar 15;201(6):650–660. PMCID: PMC7068837. [3] Misharin AV, Morales-Nebreda L, Reyfman PA, Cuda CM, Walter JM, McQuattie-Pimentel AC, Chen CI, Anekalla KR, Joshi N, Williams KJN, Abdala-Valencia H, Yacoub TJ, Chi M, Chiu S, Gonzalez-Gonzalez FJ, Gates K, Lam AP, Nicholson TT, Homan PJ, Soberanes S, Dominguez S, Morgan VK, Saber R, Shaffer A, Hinchcliff M, Marshall SA, Bharat A, Berdnikovs S, Bhorade SM, Bartom ET, Morimoto RI, Balch WE, Sznajder JI, Chandel NS, Mutlu GM, Jain M, Gottardi CJ, Singer BD, Ridge KM, Bagheri N, Shilatifard A, Budinger GRS, Perlman H. Monocyte-derived alveolar macrophages drive lung fibrosis and persist in the lung over the life span. J Exp Med. 2017 Aug 7;214(8):2387–2404. PMCID: PMC5551573. Acknowledgements: Special thanks to Alyssa Williams, Nicolas Page, Sophia E. Kujawski, William Odell, Baran Ilayda Gunes, Michelle Chung and Crystal Cheung who recruited patients at Yale School of Medicine. Disclosure of Interests: Nikolay Markov: None declared , Anthony Esposito: None declared , Karolina Senkow: None declared , Maxwell Schleck: None declared , Luisa Kusick: None declared , Yuliana Sokolenko: None declared , Zhan Yu: None declared , Estefani Diaz: None declared , Emmy Jonasson: None declared , Suchitra Swaminathan: None declared , Ziyan Lu: None declared , Radmila Nafikova: None declared , Samuel Fenske: None declared , Alec Peltekian: None declared , Ankit Agrawal: None declared , Stephanie Perez: None declared , Shannon Teaw: None declared , Ian Cumming: None declared , Robert Tinghe: None declared , Hatice Savas: None declared , Hadijat Makinde: None declared , Carla Cuda: None declared , Matthew Dapas: None declared , Carrie Richardson: None declared , Harris Perlman: None declared , Cara Gottardi: None declared , Scott Budinger: None declared , Alexander Misharin: None declared , Monique Hinchcliff Merck, AbbVie, Boehringer Ingelheim, Kadmon and Boehringer Ingelheim. © The Authors 2025. This abstract is an open access article published in Annals of Rheumatic Diseases under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Neither EULAR nor the publisher make any representation as to the accuracy of the content. The authors are solely responsible for the content in their abstract including accuracy of the facts, statements, results, conclusion, citing resources etc.

Recent developments in materials science have made it possible to synthesize millions of individual nanoparticles on a chip. However, many steps in the characterization process still require extensive human input. To address this challenge, we present an automated image processing pipeline that optimizes high-throughput nanoparticle characterization using intelligent image segmentation and coordinate generation. The proposed method can rapidly analyze each image and return optimized acquisition coordinates suitable for multiple analytical STEM techniques, including 4D-STEM, EELS, and EDS. The pipeline employs computer vision and unsupervised learning to remove the image background, segment the particle into areas of interest, and generate acquisition coordinates. This approach eliminates the need for uniform grid sampling, focusing data collection on regions of interest. We validated our approach using a diverse dataset of over 900 high-resolution grayscale nanoparticle images, achieving a 96.0% success rate based on expert-validated criteria. Using established 4D-STEM acquisition times as a baseline, our method demonstrates a 25.0 to 29.1-fold reduction in total processing time. By automating this crucial preprocessing step and optimizing data acquisition, our pipeline significantly accelerates materials characterization workflows while reducing unnecessary data collection.

Abstract Rationale: Interstitial lung disease (ILD) is a common complication of systemic sclerosis (SSc) associated with significant morbidity and mortality. Professional society guidelines vary but some recommend screening all patients with SSc at diagnosis with computed tomography (CT). Early detection of ILD, however, remains challenging, as subtle CT features can be of unclear significance. Radiomics, which involves extracting quantitative features from information-rich medical imaging, can capture subtle textural and structural changes that are not easily observable to radiologists. We hypothesize that by leveraging radiomic features, machine learning models can aid clinicians in early ILD detection and risk stratification in SSc. Methods: We analyzed CT scans performed between 2015-2024 on patients with SSc enrolled in the Northwestern Scleroderma Registry. Radiomic feature extraction yielded 107 first-order features that detailed the lung's texture, shape, and intensity patterns. Principal component analysis (PCA) was utilized to reduce these features to 12 principal components, capturing the majority of variance in the data. Optuna, an optimization library, was employed to build models to detect ILD using radiomic features. These models included XGBoost, random forest, logistic regression, LightGBM, and gradient boosting. Receiver operator characteristic analysis was implemented to assess the accuracy of model predictions using an 64:16:20 train:validation:test split. Large Language Models (LLMs) were used to assist in plot generation with authors reviewing output. Results: 1221 CT scans were available for analysis from 434 patients with SSc (83% female, median age 58yr [IQR: 50, 66]). 808 CTs (66%) had evidence of ILD reported by a thoracic radiologist. PCA plots of distinct patterns of radiomic features from CT scans clustered patients with and without ILD (Figure 1A). Using ML algorithms via the Optuna framework, radiomic features successfully detected ILD in SSc patients. We used the highest validation area under the receiver operating characteristic curve (AUROC) to select the model and apply on an unseen test set and obtained an AUROC of 0.88 (Figure 1B). Conclusion: Distinct radiomic patterns were identified through ML algorithms that detect ILD in SSc patients with good discrimination. Our findings underscore the utility of radiomics in diagnosis in this high-risk population. Future research will integrate clinical data with radiomic features via deep learning models to improve predictive performance, thereby improving the care of patients with systemic sclerosis.

Abstract Background: Ventilator-associated pneumonia (VAP) is one of the deadliest hospital-acquired infections, with a mortality rate ranging from 25-70%. Currently, clinical models to predict VAP development are limited. Development of such a model would potentially allow physicians to intervene earlier with diagnostics and treatment, thereby improving patient outcomes. Methods: We examined VAP episodes from patients enrolled in the SCRIPT study, a cohort study of patients on mechanical ventilation who underwent a bronchoalveolar lavage for suspected pneumonia. A team of five attending physicians reviewed patient charts and adjudicated VAP episodes. We visualized patient-day features using hierarchical clustering with Ward's method. Clinical features such as vital signs, ventilator parameters, laboratory values, and medication data were used to develop several machine learning models trained to identify patients on a day to day basis for high risk of developing VAP within the next 7 days. LLMs were used to help coding, with all output reviewed by the team. We used five fold cross validation. For explainability, we used SHAP plots to examine which clinical features impacted model decision-making. Results: We examined ICU stay data from 688 patients, 268 of whom developed VAP. Median patient age was 62 (IQR 51-71), and 59% were male; 42% died. Our dataset had 1,296 ICU-days occurring within seven days before a VAP episode. For clean model training, we used patient days from patients who were adjudicated not to have pneumonia on enrollment into the study and who did not develop VAP during their stay to label the negative class (646 days). Visualization using hierarchical clustering showed correlation with length of hospitalization and ventilation (Figure 1A). The best-performing models using XGBoost had a mean AUROC of 0.774 with standard deviation of 0.026 (Figure 1B). Important features based on SHAP included PEEP, platelets, and day of hospitalization (Figure 1C). Conclusions: Machine learning models can predict VAP onset within the next 7 days with moderate performance. Future work will focus on revising feature selection and attempting alternative machine learning strategies such as deep learning models.

Abstract RATIONALE: Immunocompromising medications, commonly employed in the context of autoimmune diseases, cancers, and transplants, can significantly increase patient vulnerability to infection. Ascertaining use of these medications from structured electronic health record (EHR) data is difficult because medication orders are not consistently updated and may miss information for patients transferring between health systems. Large language models (LLMs) have shown promise in their ability to extract structured data elements from clinical notes. We sought to determine how an LLM would perform in extracting histories of immunosuppressive medication use from the text of clinical admission notes and to compare this to using medication order data. METHODS: We utilized data from Successful Clinical Response in Pneumonia Therapy (SCRIPT), a single-center prospective cohort study of patients at Northwestern Memorial Hospital who required mechanical ventilation and underwent bronchoalveolar lavage for suspected pneumonia. The research team reviewed each patient chart at study enrollment for history of using the following immunosuppressive medications: azathioprine, cyclosporine, cyclophosphamide, mycophenolate, rituximab, and tacrolimus. For each patient, medication order data for 6 months prior to hospitalization was extracted from the EHR. The presence of a valid order for a given medication was used to categorize patients as a user of that medication. Additionally for each patient, their hospital admission note was processed through Philter, a clinical text de-identification tool, to remove protected health information. Microsoft Azure's secure OpenAI API for batch querying gpt-4o-2024-05-13 (GPT-4o) was then used to analyze each note and extract immunosuppressive medication use. Precision, recall, and F1 scores for identification of each immunosuppressive medication by each method were computed. LLMs were used for coding with output reviewed by authors. RESULTS: There were 827 SCRIPT enrollments with medication order data, admission notes, and gold standard immunosuppressive medication use labels from chart review. The performance of medication order-based identification varied across the medications. F1 scores were 0.75 for azathioprine, 0.74 for cyclosporine, 0.3 for cyclophosphamide, 0.77 for mycophenolate, 0.68 for rituximab, and 0.81 for tacrolimus. The performance of GPT-4o-based identification was more consistent. F1 scores were 1 for azathioprine, 0.9 for cyclosporine, 0.89 for cyclophosphamide, 0.92 for mycophenolate, 0.85 for rituximab, and 0.96 for tacrolimus. CONCLUSION: GPT-4o effectively extracts histories of various immunosuppressive medication use from clinical notes. For every immunosuppressive medication examined, GPT-4o performed better than using medication orders, with F1 scores 0.85 or above for all medications. Future steps include examining broader medication classes such as myelosuppressive chemotherapies or corticosteroids.

Abstract Rationale: Interstitial lung disease (ILD) affects 40-75% of patients with systemic sclerosis (SSc) and is the leading cause of death in this population. SSc-ILD is a heterogeneous disease with a variable clinical course. Current available therapies preserve lung function; however, benefits appear to be modest and are counterbalanced by toxicity. While biomarkers have been reported for progressive disease, their utility is limited. We hypothesize that machine learning (ML) could improve mortality prediction in patients with SSc and SSc-ILD by leveraging readily available electronic health record (EHR) data. Methods: We used data from participants with SSc recruited to the Northwestern University Scleroderma Registry from 1996-2024. EHR data—clinical, laboratory, and spirometric—were extracted, and features were selected. ILD diagnosis was assigned by adjudication of chest CT reports. Multiple ML algorithms were tested to build models to predict mortality (or lung transplant) using EHR features in the entire SSc cohort and a subgroup of those with SSc-ILD. A 70:10:20 patient-wise split for training:validation:testing was implemented. Optuna was employed for hyperparameter optimization on the validation set, and the best model was selected. Receiver operating characteristic (ROC) analysis was used to evaluate each model on the held-out test set. Feature importance was assessed through ablation analysis. Results: 1,170 participants with SSc were available for analysis, encompassing 9,191 person-years of observation. 193 (16%) participants died during the observation period. 709 (61%) had CT data available, and 454 (64%) of these had SSc-ILD for subgroup analysis. 109 (24%) participants with SSc-ILD died during the observation period. EHR features predicted mortality in SSc patients within one (AUC=0.91), three (AUC=0.89), and five years (AUC=0.82) (Figure 1A). Ablation analysis identified features highly predictive of one-year mortality in SSc that are routinely assessed but rarely utilized, including blood counts and chemistries (Figure 1B). Similarly, ML algorithms used EHR features to predict mortality in a subgroup of patients with SSc-ILD within one (AUC=0.71), three (AUC=0.73), and five years (AUC=0.80) (Figure 1C). Ablation analysis again identified predictive features for one-year mortality in those with SSc-ILD (Figure 1D). Conclusions: ML analysis of readily available EHR data predicts mortality in those with SSc and SSc-ILD with high sensitivity and specificity. Ablation analyses identified features predictive of one-year mortality that are routinely collected but rarely assessed to ascertain risk. These models could assist in clinical decision-making, particularly regarding treatment. Future research will integrate quantitative imaging features and employ deep learning models to further improve model performance.

Abstract Rationale: Interstitial lung disease (ILD) affects 40-75% of patients with systemic sclerosis (SSc) and is the leading cause of death in this population. Profibrotic monocyte-derived alveolar macrophages (MoAM)—expressing SPP1, MMP9, CHIT1, CHI3L1—play a causal role in the pathogenesis of pulmonary fibrosis in animal models. In humans, these profibrotic MoAM have been identified in explanted lung tissue from patients with fibrotic ILD yet are absent in healthy donor lung. We aimed to determine the localization of profibrotic MoAM in lung explants from patients with SSc-ILD and hypothesized that the abundance and phenotype of these MoAM are associated with disease severity. Methods: Explants from patients with SSc-ILD were profiled with the Xenium spatial transcriptomic assay. Nine subjects with SSc-ILD and thirteen healthy, non-SSc controls were recruited from three academic centers. Clinical, spirometric, and radiologic characteristics were assessed. Subjects underwent flexible fiberoptic bronchoscopy with bronchoalveolar lavage (BAL). Cell populations from BAL fluid (BALF) were analyzed by single cell RNA sequencing. Their proportional prevalence and transcriptional profiles were tested for association with lung function and extent of pulmonary fibrosis on computed tomography (CT) as quantified by a thoracic radiologist using the Kazerooni score. Results: Single-cell spatial transcriptomic analysis of explanted lung tissue from patients with SSc-ILD demonstrated that profibrotic MoAM were exclusively localized to the airspace. Compared to controls, subjects with SSc-ILD had increased proportions of profibrotic MoAM in BALF (Figure 1A, 1B). The abundance of these MoAM inversely correlated with lung function (Figure 1C) and directly correlated with the extent of fibrosis on CT (Figure 1D). We additionally identified genes across all alveolar macrophage subtypes whose expression significantly correlated with lung function. Differential expression analysis revealed significant transcriptomic changes in tissue-resident alveolar macrophages, MoAM, and subsets of T cells (CD8+ tissue-resident memory, CD4+ effector memory) in subjects with SSc-ILD compared to controls. Furthermore, we identified a transcriptomic signature in BALF from subjects with SSc-ILD that was associated with treatment with mycophenolate mofetil and a reduction in interferon gamma levels. Conclusions: Profibrotic MoAM are localized to the airspace of patients with SSc-ILD and can be sampled by BAL. Abundance of these MoAM and gene expression profiles are associated with functional and radiologic severity of disease. Our data suggest that profibrotic MoAM may serve as a biomarker of SSc-ILD and a potential target for therapy.

Abstract Background: Pneumonia is a leading cause of hospitalizations and mortality worldwide. Accurate outcome prediction is crucial for timely adjustments to patient management strategies. Traditional tools examining clinical stability such as Halm's Criteria, based on vital signs, mental status, and oral intake, may overlook complex interactions in patient data and are not optimized for critically ill patients. Machine learning (ML) techniques may predict treatment success with greater precision by integrating multiple clinical features. This study aims to evaluate the performance of ML algorithms in predicting successful treatment outcomes in community-acquired pneumonia (CAP). Methods: We analyzed CAP episodes from patients enrolled in the SCRIPT study, a cohort requiring mechanical ventilation who underwent bronchoalveolar lavage (BAL) for suspected pneumonia. Clinical features from day 3 of each CAP episode were used to build several gradient-boosted ML models to predict whether the pneumonia episode would successfully be treated by day 7 (Fig 1A). These clinical features of interest were derived from the original Halm's Criteria (Fig 1B). Several binning methods for clinical features were tested, including direct processing of raw clinical data without binning, binning into quintiles, and binning into binary categories. SHAP plots were used for model interpretability. ML models were compared against two modified Halm's Criteria-based models (vital signs alone and all variables used for gradient-boosted models). Large language model was used to assist coding and writing. Results: We examined ICU stay data from 150 patients with CAP. The cohort was 53.3% male, median age 63.5. There were 81 episodes that had clinical data from the first three consecutive days and an adjudicated day 7 cure status, out of which 43 were cured by day 7. Halm's Criteria predicted cure with AUROC of 0.61. ML models universally outperformed Halm's models (Fig 1C). The best performing model was XGBoost without binning, with 5-fold cross-validation showing mean AUROC of 0.75 (Fig 1D), achieving 85.7% sensitivity, 56.5% specificity, and 70.5% accuracy for predicting non-cure by day 7 at a threshold of 0.6. Important features factored into prediction included GCS, heart rate, RASS, P/F ratio, and norepinephrine requirement (Fig 1E). Conclusions: ML models outperform traditional methods in predicting CAP treatment outcomes. Patients identified by the model to be at high risk of not being cured by day 7 may benefit from additional investigation with imaging, repeat BAL, or change in antibiotic regimen. Future research should focus on prospective validation and integration into clinical practice.

The design of novel materials hinges on the understanding of structure-property relationships. However, in recent times, our capability to synthesize a large number of materials has outpaced our speed at characterizing them. While the overall chemical constituents can be readily known during synthesis, the structural evolution and characterization of newly synthesized samples remains a bottleneck for the ultimate goal of high throughput nanomaterials discovery. Thus, scalable methods for crystal symmetry determination that can analyze a large volume of material samples within a short time-frame are especially needed. Kikuchi diffraction in the SEM is a promising technique for this due to its sensitivity to dynamical scattering, which may provide information beyond just the seven crystal systems and fourteen Bravais lattices. After diffraction patterns are collected from material samples, deep learning methods may be able to classify the space group symmetries using the patterns as input, which paired with the elemental composition, would help enable the determination of the crystal structure. To investigate the feasibility of this solution, neural networks were trained to predict the space group type of background corrected EBSD patterns. Our networks were first trained and tested on an artificial dataset of EBSD patterns of 5,148 different cubic phases, created through physics-based dynamical simulations. Next, Maximum Classifier Discrepancy, an unsupervised deep learning-based domain adaptation method, was utilized to train neural networks to make predictions for experimental EBSD patterns. We introduce a relabeling scheme, which enables our models to achieve accuracy scores higher than 90% on simulated and experimental data, suggesting that neural networks are capable of making predictions of crystal symmetry from an EBSD pattern.