Evaluating the accuracy of a state-of-the-art large language model for prediction of admissions from the emergency room

BACKGROUND

Artificial intelligence (AI) and large language models (LLMs) can play a critical role in emergency room operations by augmenting decision-making about patient admission. However, there are no studies for LLMs using real-world data and scenarios, in comparison to and being informed by traditional supervised machine learning (ML) models. We evaluated the performance of GPT-4 for predicting patient admissions from emergency department (ED) visits. We compared performance to traditional ML models both naively and when informed by few-shot examples and/or numerical probabilities.

METHODS

We conducted a retrospective study using electronic health records across 7 NYC hospitals. We trained Bio-Clinical-BERT and XGBoost (XGB) models on unstructured and structured data, respectively, and created an ensemble model reflecting ML performance. We then assessed GPT-4 capabilities in many scenarios: through Zero-shot, Few-shot with and without retrieval-augmented generation (RAG), and with and without ML numerical probabilities.

RESULTS

The Ensemble ML model achieved an area under the receiver operating characteristic curve (AUC) of 0.88, an area under the precision-recall curve (AUPRC) of 0.72 and an accuracy of 82.9%. The naïve GPT-4's performance (0.79 AUC, 0.48 AUPRC, and 77.5% accuracy) showed substantial improvement when given limited, relevant data to learn from (ie, RAG) and underlying ML probabilities (0.87 AUC, 0.71 AUPRC, and 83.1% accuracy). Interestingly, RAG alone boosted performance to near peak levels (0.82 AUC, 0.56 AUPRC, and 81.3% accuracy).

CONCLUSIONS

The naïve LLM had limited performance but showed significant improvement in predicting ED admissions when supplemented with real-world examples to learn from, particularly through RAG, and/or numerical probabilities from traditional ML models. Its peak performance, although slightly lower than the pure ML model, is noteworthy given its potential for providing reasoning behind predictions. Further refinement of LLMs with real-world data is necessary for successful integration as decision-support tools in care settings.

Accurate prediction of neurologic changes in critically ill infants using pose AI

Importance

Infant alertness and neurologic changes are assessed by exam, which can be intermittent and subjective. Reliable, continuous methods are needed.

Objective

We hypothesized that our computer vision method to track movement, pose AI, could predict neurologic changes.

Design

Retrospective observational study from 2021-2022.

Setting

A level four urban neonatal intensive care unit (NICU).

Participants

Infants with corrected age ≤1 year, comprising 115 patients with 4,705 hours of video data linked to electroencephalograms (EEG), including 46% female and 25.2% white non-Hispanic.

Exposures

Pose AI prediction of anatomic landmark position and an XGBoost classifier trained on one-minute variance in pose.

Main outcomes and measures

Outcomes were cerebral dysfunction, diagnosed from EEG readings by an epileptologist, and sedation, defined by the administration of sedative medications. Measures of algorithm performance were receiver operating characteristic-area under the curves (ROC-AUCs) on cross-validation and on two test datasets comprised of held-out infants and held-out video frames from infants used in training.

Results

Infant pose was accurately predicted in cross-validation, held-out frames, and held-out infants (respective ROC-AUCs 0.94, 0.83, 0.89). Median movement increased with age and, after accounting for age, was lower with sedative medications and in infants with cerebral dysfunction (all P<5×10-3, 10,000 permutations). Sedation prediction had high performance on cross-validation, held-out frames, and held-out infants (ROC-AUCs 0.90, 0.91, 0.87), as did prediction of cerebral dysfunction (ROC-AUCs 0.91, 0.90, 0.76).

Conclusions and Relevance

We used pose AI to predict sedation and cerebral dysfunction in 4,705 hours of video from a large, diverse cohort of infants. Pose AI may offer a scalable, minimally invasive method for neuro-telemetry in the NICU.

Large language models for generating medical examinations: systematic review

BACKGROUND

Writing multiple choice questions (MCQs) for the purpose of medical exams is challenging. It requires extensive medical knowledge, time and effort from medical educators. This systematic review focuses on the application of large language models (LLMs) in generating medical MCQs.

METHODS

The authors searched for studies published up to November 2023. Search terms focused on LLMs generated MCQs for medical examinations. Non-English, out of year range and studies not focusing on AI generated multiple-choice questions were excluded. MEDLINE was used as a search database. Risk of bias was evaluated using a tailored QUADAS-2 tool.

RESULTS

Overall, eight studies published between April 2023 and October 2023 were included. Six studies used Chat-GPT 3.5, while two employed GPT 4. Five studies showed that LLMs can produce competent questions valid for medical exams. Three studies used LLMs to write medical questions but did not evaluate the validity of the questions. One study conducted a comparative analysis of different models. One other study compared LLM-generated questions with those written by humans. All studies presented faulty questions that were deemed inappropriate for medical exams. Some questions required additional modifications in order to qualify.

CONCLUSIONS

LLMs can be used to write MCQs for medical examinations. However, their limitations cannot be ignored. Further study in this field is essential and more conclusive evidence is needed. Until then, LLMs may serve as a supplementary tool for writing medical examinations. 2 studies were at high risk of bias. The study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Utilizing large language models in breast cancer management: systematic review

PURPOSE

Despite advanced technologies in breast cancer management, challenges remain in efficiently interpreting vast clinical data for patient-specific insights. We reviewed the literature on how large language models (LLMs) such as ChatGPT might offer solutions in this field.

METHODS

We searched MEDLINE for relevant studies published before December 22, 2023. Keywords included: "large language models", "LLM", "GPT", "ChatGPT", "OpenAI", and "breast". The risk bias was evaluated using the QUADAS-2 tool.

RESULTS

Six studies evaluating either ChatGPT-3.5 or GPT-4, met our inclusion criteria. They explored clinical notes analysis, guideline-based question-answering, and patient management recommendations. Accuracy varied between studies, ranging from 50 to 98%. Higher accuracy was seen in structured tasks like information retrieval. Half of the studies used real patient data, adding practical clinical value. Challenges included inconsistent accuracy, dependency on the way questions are posed (prompt-dependency), and in some cases, missing critical clinical information.

CONCLUSION

LLMs hold potential in breast cancer care, especially in textual information extraction and guideline-driven clinical question-answering. Yet, their inconsistent accuracy underscores the need for careful validation of these models, and the importance of ongoing supervision.

Artificial intelligence in preventive cardiology

Artificial intelligence (AI) is a field of study that strives to replicate aspects of human intelligence into machines. Preventive cardiology, a subspeciality of cardiovascular (CV) medicine, aims to target and mitigate known risk factors for CV disease (CVD). AI's integration into preventive cardiology may introduce novel treatment interventions and AI-centered clinician assistive tools to reduce the risk of CVD. AI's role in nutrition, weight loss, physical activity, sleep hygiene, blood pressure, dyslipidemia, smoking, alcohol, recreational drugs, and mental health has been investigated. AI has immense potential to be used for the screening, detection, and monitoring of the mentioned risk factors. However, the current literature must be supplemented with future clinical trials to evaluate the capabilities of AI interventions for preventive cardiology. This review discusses present examples, potentials, and limitations of AI's role for the primary and secondary prevention of CVD.

A novel method leveraging time series data to improve subphenotyping and application in critically ill patients with COVID-19

Computational subphenotyping, a data-driven approach to understanding disease subtypes, is a prominent topic in medical research. Numerous ongoing studies are dedicated to developing advanced computational subphenotyping methods for cross-sectional data. However, the potential of time-series data has been underexplored until now. Here, we propose a Multivariate Levenshtein Distance (MLD) that can account for address correlation in multiple discrete features over time-series data. Our algorithm has two distinct components: it integrates an optimal threshold score to enhance the sensitivity in discriminating between pairs of instances, and the MLD itself. We have applied the proposed distance metrics on the k-means clustering algorithm to derive temporal subphenotypes from time-series data of biomarkers and treatment administrations from 1039 critically ill patients with COVID-19 and compare its effectiveness to standard methods. In conclusion, the Multivariate Levenshtein Distance metric is a novel method to quantify the distance from multiple discrete features over time-series data and demonstrates superior clustering performance among competing time-series distance metrics.

High-throughput target trial emulation for Alzheimer's disease drug repurposing with real-world data

Target trial emulation is the process of mimicking target randomized trials using real-world data, where effective confounding control for unbiased treatment effect estimation remains a main challenge. Although various approaches have been proposed for this challenge, a systematic evaluation is still lacking. Here we emulated trials for thousands of medications from two large-scale real-world data warehouses, covering over 10 years of clinical records for over 170 million patients, aiming to identify new indications of approved drugs for Alzheimer's disease. We assessed different propensity score models under the inverse probability of treatment weighting framework and suggested a model selection strategy for improved baseline covariate balancing. We also found that the deep learning-based propensity score model did not necessarily outperform logistic regression-based methods in covariate balancing. Finally, we highlighted five top-ranked drugs (pantoprazole, gabapentin, atorvastatin, fluticasone, and omeprazole) originally intended for other indications with potential benefits for Alzheimer's patients.

A Five-Decade Text Mining Analysis of Cochlear Implant Research: Where We Started and Where We Are Heading

Background and Objectives: Since its invention in the 1970s, the cochlear implant (CI) has been substantially developed. We aimed to assess the trends in the published literature to characterize CI. Materials and Methods: We queried PubMed for all CI-related entries published during 1970-2022. The following data were extracted: year of publication, publishing journal, title, keywords, and abstract text. Search terms belonged to the patient's age group, etiology for hearing loss, indications for CI, and surgical methodological advancement. Annual trends of publications were plotted. The slopes of publication trends were calculated by fitting regression lines to the yearly number of publications. Results: Overall, 19,428 CIs articles were identified. Pediatric-related CI was the most dominant sub-population among the age groups, with the highest rate and slope during the years (slope 5.2 ± 0.3, p < 0.001), while elderly-related CIs had significantly fewer publications. Entries concerning hearing preservation showed the sharpest rise among the methods, from no entries in 1980 to 46 entries in 2021 (slope 1.7 ± 0.2, p < 0.001). Entries concerning robotic surgery emerged in 2000, with a sharp increase in recent years (slope 0.5 ± 0.1, p < 0.001). Drug-eluting electrodes and CI under local-anesthesia have been reported only in the past five years, with a gradual rise. Conclusions: Publications regarding CI among pediatrics outnumbered all other indications, supporting the rising, pivotal role of CI in the rehabilitation of children with sensorineural hearing loss. Hearing-preservation publications have recently rapidly risen, identified as the primary trend of the current era, followed by a sharp rise of robotic surgery that is evolving and could define the next revolution.

Comparing ChatGPT and GPT-4 performance in USMLE soft skill assessments

The United States Medical Licensing Examination (USMLE) has been a subject of performance study for artificial intelligence (AI) models. However, their performance on questions involving USMLE soft skills remains unexplored. This study aimed to evaluate ChatGPT and GPT-4 on USMLE questions involving communication skills, ethics, empathy, and professionalism. We used 80 USMLE-style questions involving soft skills, taken from the USMLE website and the AMBOSS question bank. A follow-up query was used to assess the models' consistency. The performance of the AI models was compared to that of previous AMBOSS users. GPT-4 outperformed ChatGPT, correctly answering 90% compared to ChatGPT's 62.5%. GPT-4 showed more confidence, not revising any responses, while ChatGPT modified its original answers 82.5% of the time. The performance of GPT-4 was higher than that of AMBOSS's past users. Both AI models, notably GPT-4, showed capacity for empathy, indicating AI's potential to meet the complex interpersonal, ethical, and professional demands intrinsic to the practice of medicine.

Adversarial attacks in radiology - A systematic review

PURPOSE

The growing application of deep learning in radiology has raised concerns about cybersecurity, particularly in relation to adversarial attacks. This study aims to systematically review the literature on adversarial attacks in radiology.

METHODS

We searched for studies on adversarial attacks in radiology published up to April 2023, using MEDLINE and Google Scholar databases.

RESULTS

A total of 22 studies published between March 2018 and April 2023 were included, primarily focused on image classification algorithms. Fourteen studies evaluated white-box attacks, three assessed black-box attacks and five investigated both. Eleven of the 22 studies targeted chest X-ray classification algorithms, while others involved chest CT (6/22), brain MRI (4/22), mammography (2/22), abdominal CT (1/22), hepatic US (1/22), and thyroid US (1/22). Some attacks proved highly effective, reducing the AUC of algorithm performance to 0 and achieving success rates up to 100 %.

CONCLUSIONS

Adversarial attacks are a growing concern. Although currently the threats are more theoretical than practical, they still represent a potential risk. It is important to be alert to such attacks, reinforce cybersecurity measures, and influence the formulation of ethical and legal guidelines. This will ensure the safe use of deep learning technology in medicine.

The promise of digital healthcare technologies

Digital health technologies have been in use for many years in a wide spectrum of healthcare scenarios. This narrative review outlines the current use and the future strategies and significance of digital health technologies in modern healthcare applications. It covers the current state of the scientific field (delineating major strengths, limitations, and applications) and envisions the future impact of relevant emerging key technologies. Furthermore, we attempt to provide recommendations for innovative approaches that would accelerate and benefit the research, translation and utilization of digital health technologies.

Deep learning on electronic medical records identifies distinct subphenotypes of diabetic kidney disease driven by genetic variations in the Rho pathway

Kidney disease affects 50% of all diabetic patients; however, prediction of disease progression has been challenging due to inherent disease heterogeneity. We use deep learning to identify novel genetic signatures prognostically associated with outcomes. Using autoencoders and unsupervised clustering of electronic health record data on 1,372 diabetic kidney disease patients, we establish two clusters with differential prevalence of end-stage kidney disease. Exome-wide associations identify a novel variant in ARHGEF18, a Rho guanine exchange factor specifically expressed in glomeruli. Overexpression of ARHGEF18 in human podocytes leads to impairments in focal adhesion architecture, cytoskeletal dynamics, cellular motility, and RhoA/Rac1 activation. Mutant GEF18 is resistant to ubiquitin mediated degradation leading to pathologically increased protein levels. Our findings uncover the first known disease-causing genetic variant that affects protein stability of a cytoskeletal regulator through impaired degradation, a potentially novel class of expression quantitative trait loci that can be therapeutically targeted.

Artificial Intelligence in Neurosurgery: A State-of-the-Art Review from Past to Future

In recent years, there has been a significant surge in discussions surrounding artificial intelligence (AI), along with a corresponding increase in its practical applications in various facets of everyday life, including the medical industry. Notably, even in the highly specialized realm of neurosurgery, AI has been utilized for differential diagnosis, pre-operative evaluation, and improving surgical precision. Many of these applications have begun to mitigate risks of intraoperative and postoperative complications and post-operative care. This article aims to present an overview of the principal published papers on the significant themes of tumor, spine, epilepsy, and vascular issues, wherein AI has been applied to assess its potential applications within neurosurgery. The method involved identifying high-cited seminal papers using PubMed and Google Scholar, conducting a comprehensive review of various study types, and summarizing machine learning applications to enhance understanding among clinicians for future utilization. Recent studies demonstrate that machine learning (ML) holds significant potential in neuro-oncological care, spine surgery, epilepsy management, and other neurosurgical applications. ML techniques have proven effective in tumor identification, surgical outcomes prediction, seizure outcome prediction, aneurysm prediction, and more, highlighting its broad impact and potential in improving patient management and outcomes in neurosurgery. This review will encompass the current state of research, as well as predictions for the future of AI within neurosurgery.

Proteomic characterization of acute kidney injury in patients hospitalized with SARS-CoV2 infection

BACKGROUND

Acute kidney injury (AKI) is a known complication of COVID-19 and is associated with an increased risk of in-hospital mortality. Unbiased proteomics using biological specimens can lead to improved risk stratification and discover pathophysiological mechanisms.

METHODS

Using measurements of ~4000 plasma proteins in two cohorts of patients hospitalized with COVID-19, we discovered and validated markers of COVID-associated AKI (stage 2 or 3) and long-term kidney dysfunction. In the discovery cohort (N = 437), we identified 413 higher plasma abundances of protein targets and 30 lower plasma abundances of protein targets associated with COVID-AKI (adjusted p < 0.05). Of these, 62 proteins were validated in an external cohort (p < 0.05, N = 261).

RESULTS

We demonstrate that COVID-AKI is associated with increased markers of tubular injury (NGAL) and myocardial injury. Using estimated glomerular filtration (eGFR) measurements taken after discharge, we also find that 25 of the 62 AKI-associated proteins are significantly associated with decreased post-discharge eGFR (adjusted p < 0.05). Proteins most strongly associated with decreased post-discharge eGFR included desmocollin-2, trefoil factor 3, transmembrane emp24 domain-containing protein 10, and cystatin-C indicating tubular dysfunction and injury.

CONCLUSIONS

Using clinical and proteomic data, our results suggest that while both acute and long-term COVID-associated kidney dysfunction are associated with markers of tubular dysfunction, AKI is driven by a largely multifactorial process involving hemodynamic instability and myocardial damage.

Automated electrocardiogram signal quality assessment based on Fourier analysis and template matching

We developed and tested a novel template matching approach for signal quality assessment on electrocardiogram (ECG) data. A computational method was developed that uses a sinusoidal approximation to the QRS complex to generate a correlation value at every point of an ECG. The strength of this correlation can be numerically adapted into a 'score' for each segment of an ECG, which can be used to stratify signal quality. The algorithm was tested on lead II ECGs of intensive care unit (ICU) patients admitted to the Mount Sinai Hospital (MSH) from January to July 2020 and on records from the MIT BIH arrhythmia database. The algorithm was found to be 98.9% specific and 99% sensitive on test data from the MSH ICU patients. The routine performs in linear O(n) time and occupies O(1) heap space in runtime. This approach can be used to lower the burden of pre-processing in ECG signal analysis. Given its runtime (O(n)) and memory (O(1)) complexity, there are potential applications for signal quality stratification and arrhythmia detection in wearable devices or smartphones.

Deep Learning for Echocardiography: Introduction for Clinicians and Future Vision: State-of-the-Art Review

Exponential growth in data storage and computational power is rapidly narrowing the gap between translating findings from advanced clinical informatics into cardiovascular clinical practice. Specifically, cardiovascular imaging has the distinct advantage in providing a great quantity of data for potentially rich insights, but nuanced interpretation requires a high-level skillset that few individuals possess. A subset of machine learning, deep learning (DL), is a modality that has shown promise, particularly in the areas of image recognition, computer vision, and video classification. Due to a low signal-to-noise ratio, echocardiographic data tend to be challenging to classify; however, utilization of robust DL architectures may help clinicians and researchers automate conventional human tasks and catalyze the extraction of clinically useful data from the petabytes of collected imaging data. The promise is extending far and beyond towards a contactless echocardiographic exam-a dream that is much needed in this time of uncertainty and social distancing brought on by a stunning pandemic culture. In the current review, we discuss state-of-the-art DL techniques and architectures that can be used for image and video classification, and future directions in echocardiographic research in the current era.

The importance of social activity to risk of major depression in older adults

BACKGROUND

Several social determinants of health (SDoH) have been associated with the onset of major depressive disorder (MDD). However, prior studies largely focused on individual SDoH and thus less is known about the relative importance (RI) of SDoH variables, especially in older adults. Given that risk factors for MDD may differ across the lifespan, we aimed to identify the SDoH that was most strongly related to newly diagnosed MDD in a cohort of older adults.

METHODS

We used self-reported health-related survey data from 41 174 older adults (50-89 years, median age = 67 years) who participated in the Mayo Clinic Biobank, and linked ICD codes for MDD in the participants' electronic health records. Participants with a history of clinically documented or self-reported MDD prior to survey completion were excluded from analysis (N = 10 938, 27%). We used Cox proportional hazards models with a gradient boosting machine approach to quantify the RI of 30 pre-selected SDoH variables on the risk of future MDD diagnosis.

RESULTS

Following biobank enrollment, 2073 older participants were diagnosed with MDD during the follow-up period (median duration = 6.7 years). The most influential SDoH was perceived level of social activity (RI = 0.17). Lower level of social activity was associated with a higher risk of MDD [hazard ratio = 2.27 (95% CI 2.00-2.50) for highest v. lowest level].

CONCLUSION

Across a range of SDoH variables, perceived level of social activity is most strongly related to MDD in older adults. Monitoring changes in the level of social activity may help identify older adults at an increased risk of MDD.

Multi-center retrospective cohort study applying deep learning to electrocardiograms to identify left heart valvular dysfunction

BACKGROUND

Aortic Stenosis and Mitral Regurgitation are common valvular conditions representing a hidden burden of disease within the population. The aim of this study was to develop and validate deep learning-based screening and diagnostic tools that can help guide clinical decision making.

METHODS

In this multi-center retrospective cohort study, we acquired Transthoracic Echocardiogram reports from five Mount Sinai hospitals within New York City representing a demographically diverse cohort of patients. We developed a Natural Language Processing pipeline to extract ground-truth labels about valvular status and paired these to Electrocardiograms (ECGs). We developed and externally validated deep learning models capable of detecting valvular disease, in addition to considering scenarios of clinical deployment.

RESULTS

We use 617,338 ECGs paired to transthoracic echocardiograms from 123,096 patients to develop a deep learning model for detection of Mitral Regurgitation. Area Under Receiver Operating Characteristic curve (AUROC) is 0.88 (95% CI:0.88-0.89) in internal testing, and 0.81 (95% CI:0.80-0.82) in external validation. To develop a model for detection of Aortic Stenosis, we use 617,338 Echo-ECG pairs for 128,628 patients. AUROC is 0.89 (95% CI: 0.88-0.89) in internal testing, going to 0.86 (95% CI: 0.85-0.87) in external validation. The model's performance increases leading up to the time of the diagnostic echo, and it performs well in validation against requirement of Transcatheter Aortic Valve Replacement procedures.

CONCLUSIONS

Deep learning based tools can increase the amount of information extracted from ubiquitous investigations such as the ECG. Such tools are inexpensive, can help in earlier disease detection, and potentially improve prognosis.

Molecular states during acute COVID-19 reveal distinct etiologies of long-term sequelae

Post-acute sequelae of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection are debilitating, clinically heterogeneous and of unknown molecular etiology. A transcriptome-wide investigation was performed in 165 acutely infected hospitalized individuals who were followed clinically into the post-acute period. Distinct gene expression signatures of post-acute sequelae were already present in whole blood during acute infection, with innate and adaptive immune cells implicated in different symptoms. Two clusters of sequelae exhibited divergent plasma-cell-associated gene expression patterns. In one cluster, sequelae associated with higher expression of immunoglobulin-related genes in an anti-spike antibody titer-dependent manner. In the other, sequelae associated independently of these titers with lower expression of immunoglobulin-related genes, indicating lower non-specific antibody production in individuals with these sequelae. This relationship between lower total immunoglobulins and sequelae was validated in an external cohort. Altogether, multiple etiologies of post-acute sequelae were already detectable during SARS-CoV-2 infection, directly linking these sequelae with the acute host response to the virus and providing early insights into their development.

Cannabis Use and CKD: Epidemiological Associations and Mendelian Randomization

Rationale & Objective

The association between cannabis use and chronic kidney disease (CKD) is controversial. We aimed to assess association of CKD with cannabis use in a large cohort study and then assess causality using Mendelian randomization with a genome-wide association study (GWAS).

Study Design

Retrospective cohort study and genome-wide association study.

Setting & Participants

The retrospective study was conducted on the All of Us cohort (N=223,354). Genetic instruments for cannabis use disorder were identified from 3 GWAS: the Psychiatric Genomics Consortium Substance Use Disorders, iPSYCH, and deCODE (N=384,032). Association between genetic instruments and CKD was investigated in the CKDGen GWAS (N > 1.2 million).

Exposure

Cannabis consumption.

Outcomes

CKD outcomes included: cystatin-C and creatinine-based kidney function, proteinuria, and blood urea nitrogen.

Analytical Approach

We conducted association analyses to test for frequency of cannabis use and CKD. To evaluate causality, we performed a 2-sample Mendelian randomization.

Results

In the retrospective study, compared to former users, less than monthly (OR, 1.01; 95% CI, 0.87-1.18; P = 0.87) and monthly cannabis users (OR, 1.15; 95% CI, 0.86-1.52; P = 0.33) did not have higher CKD odds. Conversely, weekly (OR, 1.28; 95% CI, 1.01-1.60; P = 0.04) and daily use (OR, 1.25; 95% CI, 1.04-1.50; P = 0.02) was significantly associated with CKD, adjusted for multiple confounders. In Mendelian randomization, genetic liability to cannabis use disorder was not associated with increased odds for CKD (OR, 1.00; 95% CI, 0.99-1.01; P = 0.96). These results were robust across different Mendelian randomization techniques and multiple kidney traits.

Limitations

Likely underreporting of cannabis use. In Mendelian randomization, genetic instruments were identified in the GWAS that included individuals primarily of European ancestry.

Conclusions

Despite the epidemiological association between cannabis use and CKD, there was no evidence of a causal effect, indicating confounding in observational studies.

BGLM: big data-guided LOINC mapping with multi-language support

MOTIVATION

Mapping internal, locally used lab test codes to standardized logical observation identifiers names and codes (LOINC) terminology has become an essential step in harmonizing electronic health record (EHR) data across different institutions. However, most existing LOINC code mappers are based on text-mining technology and do not provide robust multi-language support.

MATERIALS AND METHODS

We introduce a simple, yet effective tool called big data-guided LOINC code mapper (BGLM), which leverages the large amount of patient data stored in EHR systems to perform LOINC coding mapping. Distinguishing from existing methods, BGLM conducts mapping based on distributional similarity.

RESULTS

We validated the performance of BGLM with real-world datasets and showed that high mapping precision could be achieved under proper false discovery rate control. In addition, we showed that the mapping results of BGLM could be used to boost the performance of Regenstrief LOINC Mapping Assistant (RELMA), one of the most widely used LOINC code mappers.

CONCLUSIONS

BGLM paves a new way for LOINC code mapping and therefore could be applied to EHR systems without the restriction of languages. BGLM is freely available at https://github.com/Bin-Chen-Lab/BGLM.

Reversal of cancer gene expression identifies repurposed drugs for diffuse intrinsic pontine glioma

Diffuse intrinsic pontine glioma (DIPG) is an aggressive incurable brainstem tumor that targets young children. Complete resection is not possible, and chemotherapy and radiotherapy are currently only palliative. This study aimed to identify potential therapeutic agents using a computational pipeline to perform an in silico screen for novel drugs. We then tested the identified drugs against a panel of patient-derived DIPG cell lines. Using a systematic computational approach with publicly available databases of gene signature in DIPG patients and cancer cell lines treated with a library of clinically available drugs, we identified drug hits with the ability to reverse a DIPG gene signature to one that matches normal tissue background. The biological and molecular effects of drug treatment was analyzed by cell viability assay and RNA sequence. In vivo DIPG mouse model survival studies were also conducted. As a result, two of three identified drugs showed potency against the DIPG cell lines Triptolide and mycophenolate mofetil (MMF) demonstrated significant inhibition of cell viability in DIPG cell lines. Guanosine rescued reduced cell viability induced by MMF. In vivo, MMF treatment significantly inhibited tumor growth in subcutaneous xenograft mice models. In conclusion, we identified clinically available drugs with the ability to reverse DIPG gene signatures and anti-DIPG activity in vitro and in vivo. This novel approach can repurpose drugs and significantly decrease the cost and time normally required in drug discovery.

Predicting Age-related Macular Degeneration Progression with Longitudinal Fundus Images Using Deep Learning

Accurately predicting a patient's risk of progressing to late age-related macular degeneration (AMD) is difficult but crucial for personalized medicine. While existing risk prediction models for progression to late AMD are useful for triaging patients, none utilizes longitudinal color fundus photographs (CFPs) in a patient's history to estimate the risk of late AMD in a given subsequent time interval. In this work, we seek to evaluate how deep neural networks capture the sequential information in longitudinal CFPs and improve the prediction of 2-year and 5-year risk of progression to late AMD. Specifically, we proposed two deep learning models, CNN-LSTM and CNN-Transformer, which use a Long-Short Term Memory (LSTM) and a Transformer, respectively with convolutional neural networks (CNN), to capture the sequential information in longitudinal CFPs. We evaluated our models in comparison to baselines on the Age-Related Eye Disease Study, one of the largest longitudinal AMD cohorts with CFPs. The proposed models outperformed the baseline models that utilized only single-visit CFPs to predict the risk of late AMD (0.879 vs 0.868 in AUC for 2-year prediction, and 0.879 vs 0.862 for 5-year prediction). Further experiments showed that utilizing longitudinal CFPs over a longer time period was helpful for deep learning models to predict the risk of late AMD. We made the source code available at https://github.com/bionlplab/AMD_prognosis_mlmi2022 to catalyze future works that seek to develop deep learning models for late AMD prediction.

Proteomic Characterization of Acute Kidney Injury in Patients Hospitalized with SARS-CoV2 Infection

Acute kidney injury (AKI) is a known complication of COVID-19 and is associated with an increased risk of in-hospital mortality. Unbiased proteomics using biological specimens can lead to improved risk stratification and discover pathophysiological mechanisms. Using measurements of ∼4000 plasma proteins in two cohorts of patients hospitalized with COVID-19, we discovered and validated markers of COVID-associated AKI (stage 2 or 3) and long-term kidney dysfunction. In the discovery cohort (N= 437), we identified 413 higher plasma abundances of protein targets and 40 lower plasma abundances of protein targets associated with COVID-AKI (adjusted p <0.05). Of these, 62 proteins were validated in an external cohort (p <0.05, N =261). We demonstrate that COVID-AKI is associated with increased markers of tubular injury (NGAL) and myocardial injury. Using estimated glomerular filtration (eGFR) measurements taken after discharge, we also find that 25 of the 62 AKI-associated proteins are significantly associated with decreased post-discharge eGFR (adjusted p <0.05). Proteins most strongly associated with decreased post-discharge eGFR included desmocollin-2, trefoil factor 3, transmembrane emp24 domain-containing protein 10, and cystatin-C indicating tubular dysfunction and injury. Using clinical and proteomic data, our results suggest that while both acute and long-term COVID-associated kidney dysfunction are associated with markers of tubular dysfunction, AKI is driven by a largely multifactorial process involving hemodynamic instability and myocardial damage.

Enhancing convolutional neural network predictions of electrocardiograms with left ventricular dysfunction using a novel sub-waveform representation

Background

Electrocardiogram (ECG) deep learning (DL) has promise to improve the outcomes of patients with cardiovascular abnormalities. In ECG DL, researchers often use convolutional neural networks (CNNs) and traditionally use the full duration of raw ECG waveforms that create redundancies in feature learning and result in inaccurate predictions with large uncertainties.

Objective

For enhancing these predictions, we introduced a sub-waveform representation that leverages the rhythmic pattern of ECG waveforms (data-centric approach) rather than changing the CNN architecture (model-centric approach).

Results

We applied the proposed representation to a population with 92,446 patients to identify left ventricular dysfunction. We found that the sub-waveform representation increases the performance metrics compared to the full-waveform representation. We observed a 2% increase for area under the receiver operating characteristic curve and 10% increase for area under the precision-recall curve. We also carefully examined three reliability components of explainability, interpretability, and fairness. We provided an explanation for enhancements obtained by heartbeat alignment mechanism. By developing a new scoring system, we interpreted the clinical relevance of ECG features and showed that sub-waveform representation further pushes the scores towards clinical predictions. Finally, we showed that the new representation significantly reduces prediction uncertainties within subgroups that contributes to individual fairness.

Conclusion

We expect that this added control over the granularity of ECG data will improve the DL modeling for new artificial intelligence technologies in the cardiovascular space.

Automated Determination of Left Ventricular Function Using Electrocardiogram Data in Patients on Maintenance Hemodialysis

BACKGROUND AND OBJECTIVES

Left ventricular ejection fraction is disrupted in patients on maintenance hemodialysis and can be estimated using deep learning models on electrocardiograms. Smaller sample sizes within this population may be mitigated using transfer learning.

DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS

We identified patients on hemodialysis with transthoracic echocardiograms within 7 days of electrocardiogram using diagnostic/procedure codes. We developed four models: (1) trained from scratch in patients on hemodialysis, (2) pretrained on a publicly available set of natural images (ImageNet), (3) pretrained on all patients not on hemodialysis, and (4) pretrained on patients not on hemodialysis and fine-tuned on patients on hemodialysis. We assessed the ability of the models to classify left ventricular ejection fraction into clinically relevant categories of ≤40%, 41% to ≤50%, and >50%. We compared performance by area under the receiver operating characteristic curve.

RESULTS

We extracted 705,075 electrocardiogram:echocardiogram pairs for 158,840 patients not on hemodialysis used for development of models 3 and 4 and n=18,626 electrocardiogram:echocardiogram pairs for 2168 patients on hemodialysis for models 1, 2, and 4. The transfer learning model achieved area under the receiver operating characteristic curves of 0.86, 0.63, and 0.83 in predicting left ventricular ejection fraction categories of ≤40% (n=461), 41%-50% (n=398), and >50% (n=1309), respectively. For the same tasks, model 1 achieved area under the receiver operating characteristic curves of 0.74, 0.55, and 0.71, respectively; model 2 achieved area under the receiver operating characteristic curves of 0.71, 0.55, and 0.69, respectively, and model 3 achieved area under the receiver operating characteristic curves of 0.80, 0.51, and 0.77, respectively. We found that predictions of left ventricular ejection fraction by the transfer learning model were associated with mortality in a Cox regression with an adjusted hazard ratio of 1.29 (95% confidence interval, 1.04 to 1.59).

CONCLUSION

A deep learning model can determine left ventricular ejection fraction for patients on hemodialysis following pretraining on electrocardiograms of patients not on hemodialysis. Predictions of low ejection fraction from this model were associated with mortality over a 5-year follow-up period.

PODCAST

This article contains a podcast at https://www.asn-online.org/media/podcast/CJASN/2022_06_06_CJN16481221.mp3.

Machine learning to predict in‐hospital mortality among patients with severe obesity: Proof of concept study

Objectives

Hospitalized patients with severe obesity require adapted hospital management. The aim of this study was to evaluate a machine learning model to predict in‐hospital mortality among this population.

Methods

Data of unselected consecutive emergency department admissions of hospitalized patients with severe obesity (BMI ≥ 40 kg/m²) was analyzed. Data was retrieved from five hospitals from the Mount Sinai health system, New York. The study time frame was between January 2011 and December 2019. Data was used to train a gradient‐boosting machine learning model to identify in‐hospital mortality. The model was trained and evaluated based on the data from four hospitals and externally validated on held‐out data from the fifth hospital.

Results

A total of 14,078 hospital admissions of inpatients with severe obesity were included. The in‐hospital mortality rate was 297/14,078 (2.1%). In univariate analysis, albumin (area under the curve [AUC] = 0.77), blood urea nitrogen (AUC = 0.76), acuity level (AUC = 0.73), lactate (AUC = 0.72), and chief complaint (AUC = 0.72) were the best single predictors. For Youden’s index, the model had a sensitivity of 0.77 (95% CI: 0.67–0.86) with a false positive rate of 1:9.

Conclusion

A machine learning model trained on clinical measures provides proof of concept performance in predicting mortality in patients with severe obesity. This implies that such models may help to adopt specific decision support tools for this population.

Unmasking Fracture Risk in Type 2 Diabetes: The Association of Longitudinal Glycemic Hemoglobin Level and Medications

CONTEXT

Fracture risk is underestimated in people with type 2 diabetes (T2D).

OBJECTIVE

To investigate the longitudinal relationship of glycated hemoglobin (HbA1c) and common medications on fracture risk in people with T2D.

METHODS

This retrospective population-based cohort study was conducted using de-identified claims and electronic health record data obtained from the OptumLabs Data Warehouse for the period January 1, 2007, to September 30, 2015. For each individual, the study was conducted within a 2-year HbA1c observation period and a 2-year fracture follow-up period. A cohort of 157 439 individuals with T2D [age ≥ 55 years with mean HbA1c value ≥ 6%] were selected from 4 018 250 US Medicare Advantage/Commercial enrollees with a T2D diagnosis. All fractures and fragility fractures were measured.

RESULTS

With covariates adjusted, poor glycemic control in T2D individuals was associated with an 29% increase of all fracture risk, compared with T2D individuals who had adequate glycemic control (HR: 1.29; 95% CI, 1.22-1.36). Treatment with metformin (HR: 0.88; 95% CI, 0.85-0.92) and DPP4 inhibitors (HR: 0.93; 95% CI, 0.88-0.98) was associated with a reduced all fracture risk, while insulin (HR: 1.26; 95% CI, 1.21-1.32), thiazolidinediones (HR: 1.23; 95% CI, 1.18-1.29), and meglitinides (HR: 1.12; 95% CI, 1.00-1.26) were associated with an increased all fracture risk (All P value < 0.05). Bisphosphonates were associated similarly with increased fracture risk in the T2D and nondiabetic groups.

CONCLUSION

Longitudinal 2-year HbA1c is independently associated with elevated all fracture risk in T2D individuals during a 2-year follow-up period. Metformin and DPP4 inhibitors can be used for management of T2D fracture risk.

Deep phenotyping of Alzheimer's disease leveraging electronic medical records identifies sex-specific clinical associations

Alzheimer's Disease (AD) is a neurodegenerative disorder that is still not fully understood. Sex modifies AD vulnerability, but the reasons for this are largely unknown. We utilize two independent electronic medical record (EMR) systems across 44,288 patients to perform deep clinical phenotyping and network analysis to gain insight into clinical characteristics and sex-specific clinical associations in AD. Embeddings and network representation of patient diagnoses demonstrate greater comorbidity interactions in AD in comparison to matched controls. Enrichment analysis identifies multiple known and new diagnostic, medication, and lab result associations across the whole cohort and in a sex-stratified analysis. With this data-driven method of phenotyping, we can represent AD complexity and generate hypotheses of clinical factors that can be followed-up for further diagnostic and predictive analyses, mechanistic understanding, or drug repurposing and therapeutic approaches.

Artificial Intelligence for Interstitial Lung Disease Analysis on Chest Computed Tomography: A Systematic Review

RATIONALE AND OBJECTIVES

High-resolution computed tomography (HRCT) is paramount in the assessment of interstitial lung disease (ILD). Yet, HRCT interpretation of ILDs may be hampered by inter- and intra-observer variability. Recently, artificial intelligence (AI) has revolutionized medical image analysis. This technology has the potential to advance patient care in ILD. We aimed to systematically evaluate the application of AI for the analysis of ILD in HRCT.

MATERIALS AND METHODS

We searched MEDLINE/PubMed databases for original publications of deep learning for ILD analysis on chest CT. The search included studies published up to March 1, 2021. The risk of bias evaluation included tailored Quality Assessment of Diagnostic Accuracy Studies and the modified Joanna Briggs Institute Critical Appraisal checklist.

RESULTS

Data was extracted from 19 retrospective studies. Deep learning techniques included detection, segmentation, and classification of ILD on HRCT. Most studies focused on the classification of ILD into different morphological patterns. Accuracies of 78%-91% were achieved. Two studies demonstrated near-expert performance for the diagnosis of idiopathic pulmonary fibrosis (IPF). The Quality Assessment of Diagnostic Accuracy Studies tool identified a high risk of bias in 15/19 (78.9%) of the studies.

CONCLUSION

AI has the potential to contribute to the radiologic diagnosis and classification of ILD. However, the accuracy performance is still not satisfactory, and research is limited by a small number of retrospective studies. Hence, the existing published data may not be sufficiently reliable. Only well-designed prospective controlled studies can accurately assess the value of existing AI tools for ILD evaluation.

Obesity as a mortality risk factor in the medical ward: a case control study

BACKGROUND

Research regarding the association between severe obesity and in-hospital mortality is inconsistent. We evaluated the impact of body mass index (BMI) levels on mortality in the medical wards. The analysis was performed separately before and during the COVID-19 pandemic.

METHODS

We retrospectively retrieved data of adult patients admitted to the medical wards at the Mount Sinai Health System in New York City. The study was conducted between January 1, 2011, to March 23, 2021. Patients were divided into two sub-cohorts: pre-COVID-19 and during-COVID-19. Patients were then clustered into groups based on BMI ranges. A multivariate logistic regression analysis compared the mortality rate among the BMI groups, before and during the pandemic.

RESULTS

Overall, 179,288 patients were admitted to the medical wards and had a recorded BMI measurement. 149,098 were admitted before the COVID-19 pandemic and 30,190 during the pandemic. Pre-pandemic, multivariate analysis showed a "J curve" between BMI and mortality. Severe obesity (BMI > 40) had an aOR of 0.8 (95% CI:0.7-1.0, p = 0.018) compared to the normal BMI group. In contrast, during the pandemic, the analysis showed a "U curve" between BMI and mortality. Severe obesity had an aOR of 1.7 (95% CI:1.3-2.4, p < 0.001) compared to the normal BMI group.

CONCLUSIONS

Medical ward patients with severe obesity have a lower risk for mortality compared to patients with normal BMI. However, this does not apply during COVID-19, where obesity was a leading risk factor for mortality in the medical wards. It is important for the internal medicine physician to understand the intricacies of the association between obesity and medical ward mortality.

Predictive Modelling of Susceptibility to Substance Abuse, Mortality and Drug-Drug Interactions in Opioid Patients

Objective: Opioids are a class of drugs that are known for their use as pain relievers. They bind to opioid receptors on nerve cells in the brain and the nervous system to mitigate pain. Addiction is one of the chronic and primary adverse events of prolonged usage of opioids. They may also cause psychological disorders, muscle pain, depression, anxiety attacks etc. In this study, we present a collection of predictive models to identify patients at risk of opioid abuse and mortality by using their prescription histories. Also, we discover particularly threatening drug-drug interactions in the context of opioid usage.

Methods and Materials: Using a publicly available dataset from MIMIC-III, two models were trained, Logistic Regression with L2 regularization (baseline) and Extreme Gradient Boosting (enhanced model), to classify the patients of interest into two categories based on their susceptibility to opioid abuse. We’ve also used K-Means clustering, an unsupervised algorithm, to explore drug-drug interactions that might be of concern.

Results: The baseline model for classifying patients susceptible to opioid abuse has an F1 score of 76.64% (accuracy 77.16%) while the enhanced model has an F1 score of 94.45% (accuracy 94.35%). These models can be used as a preliminary step towards inferring the causal effect of opioid usage and can help monitor the prescription practices to minimize the opioid abuse.

Discussion and Conclusion: Results suggest that the enhanced model provides a promising approach in preemptive identification of patients at risk for opioid abuse. By discovering and correlating the patterns contributing to opioid overdose or abuse among a variety of patients, machine learning models can be used as an efficient tool to help uncover the existing gaps and/or fraudulent practices in prescription writing. To quote an example of one such incidental finding, our study discovered that insulin might possibly be interacting with opioids in an unfavourable way leading to complications in diabetic patients. This indicates that diabetic patients under long term opioid usage might need to take increased amounts of insulin to make it more effective. This observation backs up prior research studies done on a similar aspect. To increase the translational value of our work, the predictive models and the associated software code are made available under the MIT License.

Development of a machine learning model using electrocardiogram signals to improve acute pulmonary embolism screening

Aims

Clinical scoring systems for pulmonary embolism (PE) screening have low specificity and contribute to computed tomography pulmonary angiogram (CTPA) overuse. We assessed whether deep learning models using an existing and routinely collected data modality, electrocardiogram (ECG) waveforms, can increase specificity for PE detection.

Methods and results

We create a retrospective cohort of 21 183 patients at moderate- to high suspicion of PE and associate 23 793 CTPAs (10.0% PE-positive) with 320 746 ECGs and encounter-level clinical data (demographics, comorbidities, vital signs, and labs). We develop three machine learning models to predict PE likelihood: an ECG model using only ECG waveform data, an EHR model using tabular clinical data, and a Fusion model integrating clinical data and an embedded representation of the ECG waveform. We find that a Fusion model [area under the receiver-operating characteristic curve (AUROC) 0.81 ± 0.01] outperforms both the ECG model (AUROC 0.59 ± 0.01) and EHR model (AUROC 0.65 ± 0.01). On a sample of 100 patients from the test set, the Fusion model also achieves greater specificity (0.18) and performance (AUROC 0.84 ± 0.01) than four commonly evaluated clinical scores: Wells' Criteria, Revised Geneva Score, Pulmonary Embolism Rule-Out Criteria, and 4-Level Pulmonary Embolism Clinical Probability Score (AUROC 0.50-0.58, specificity 0.00-0.05). The model is superior to these scores on feature sensitivity analyses (AUROC 0.66-0.84) and achieves comparable performance across sex (AUROC 0.81) and racial/ethnic (AUROC 0.77-0.84) subgroups.

Conclusion

Synergistic deep learning of ECG waveforms with traditional clinical variables can increase the specificity of PE detection in patients at least at moderate suspicion for PE.

Derivation and Validation of Genome-Wide Polygenic Score for Ischemic Heart Failure

Background Despite advances in cardiovascular disease and risk factor management, mortality from ischemic heart failure (HF) in patients with coronary artery disease (CAD) remains high. Given the partial role of genetics in HF and lack of reliable risk stratification tools, we developed and validated a polygenic risk score for HF in patients with CAD, which we term HF-PRS. Methods and Results Using summary statistics from a recent genome-wide association study for HF, we developed candidate PRSs in the Mount Sinai BioMe CAD patient cohort (N=6274) by using the pruning and thresholding method and LDPred. We validated the best score in the Penn Medicine BioBank (N=7250) and performed a subgroup analysis in a high-risk cohort who had undergone coronary catheterization. We observed a significant association between HF-PRS score and ischemic HF even after adjusting for evidence of obstructive CAD in patients of European ancestry in both BioMe (odds ratio [OR], 1.14 per SD; 95% CI, 1.05-1.24; P=0.003) and Penn Medicine BioBank (OR, 1.07 per SD; 95% CI, 1.01-1.13; P=0.016). In European patients with CAD in Penn Medicine BioBank who had undergone coronary catheterization, individuals in the top 10th percentile of PRS had a 2-fold increased odds of ischemic HF (OR, 2.0; 95% CI, 1.1-3.7; P=0.02) compared with the bottom 10th percentile. Conclusions A PRS for HF enables risk stratification in patients with CAD. Future prospective studies aimed at demonstrating clinical utility are warranted for adoption in the patient setting.

Artificial intelligence in gastroenterology: A state-of-the-art review

The development of artificial intelligence (AI) has increased dramatically in the last 20 years, with clinical applications progressively being explored for most of the medical specialties. The field of gastroenterology and hepatology, substantially reliant on vast amounts of imaging studies, is not an exception. The clinical applications of AI systems in this field include the identification of premalignant or malignant lesions (e.g., identification of dysplasia or esophageal adenocarcinoma in Barrett’s esophagus, pancreatic malignancies), detection of lesions (e.g., polyp identification and classification, small-bowel bleeding lesion on capsule endoscopy, pancreatic cystic lesions), development of objective scoring systems for risk stratification, predicting disease prognosis or treatment response [e.g., determining survival in patients post-resection of hepatocellular carcinoma), determining which patients with inflammatory bowel disease (IBD) will benefit from biologic therapy], or evaluation of metrics such as bowel preparation score or quality of endoscopic examination. The objective of this comprehensive review is to analyze the available AI-related studies pertaining to the entirety of the gastrointestinal tract, including the upper, middle and lower tracts; IBD; the hepatobiliary system; and the pancreas, discussing the findings and clinical applications, as well as outlining the current limitations and future directions in this field.

Synergistic effect of hypoalbuminaemia and hypotension in predicting in-hospital mortality and intensive care admission: a retrospective cohort study

OBJECTIVE

Hypoalbuminaemia is an important prognostic factor. It may be associated with poor nutritional states, chronic heart and kidney disease, long-standing infection and cancer. Hypotension is a hallmark of circulatory failure. We evaluated hypoalbuminaemia and hypotension synergism as predictor of in-hospital mortality and intensive care unit (ICU) admission.

DESIGN

We retrospectively analysed emergency department (ED) visits from January 2011 to December 2019.

SETTING

Data were retrieved from five Mount Sinai health system hospitals, New York.

PARTICIPANTS

We included consecutive ED patients ≥18 years with albumin measurements.

PRIMARY AND SECONDARY OUTCOME MEASURES

Clinical outcomes were in-hospital mortality and ICU admission. The rates of these outcomes were stratified by systolic blood pressure (SBP) (<90 vs ≥90 mm Hg) and albumin levels. Variables included demographics, presenting vital signs, comorbidities (measured as ICD codes) and other common blood tests. Multivariable logistic regression models analysed the adjusted OR of different levels of albumin and SBP for predicting ICU admission and in-hospital mortality. The models were adjusted for demographics, vital signs, comorbidities and common laboratory results. Patients with albumin 3.5-4.5 g/dL and SBP ≥90 mm Hg were used as reference.

RESULTS

The cohort included 402 123 ED arrivals (27.9% of total adult ED visits). The rates of in-hospital mortality, ICU admission and overall admission were 1.7%, 8.4% and 47.1%, respectively. For SBP <90 mm Hg and albumin <2.5 g/dL, mortality and ICU admission rates were 34.0% and 40.6%, respectively; for SBP <90 mm Hg and albumin ≥2.5 g/dL 8.2% and 24.1%, respectively; for SBP ≥90 mm Hg and albumin <2.5 g/dL 11.4% and 18.6%, respectively; for SBP ≥90 mm Hg and albumin 3.5-4.5 g/dL 0.5% and 6.4%, respectively. Multivariable analysis showed that in patients with hypotension and albumin <2.5 g/dL the adjusted OR for in-hospital mortality was 37.1 (95% CI 32.3 to 42.6), and for ICU admission was 5.4 (95% CI 4.8 to 6.1).

CONCLUSION

Co-occurrence of hypotension and hypoalbuminaemia is associated with poor hospital outcomes.

Contrastive Learning Improves Critical Event Prediction in COVID-19 Patients

Deep Learning (DL) models typically require large-scale, balanced training data to be robust, generalizable, and effective in the context of healthcare. This has been a major issue for developing DL models for the coronavirus-disease 2019 (COVID-19) pandemic where data are highly class imbalanced. Conventional approaches in DL use cross-entropy loss (CEL) which often suffers from poor margin classification. We show that contrastive loss (CL) improves the performance of CEL especially in imbalanced electronic health records (EHR) data for COVID-19 analyses. We use a diverse EHR data set to predict three outcomes: mortality, intubation, and intensive care unit (ICU) transfer in hospitalized COVID-19 patients over multiple time windows. To compare the performance of CEL and CL, models are tested on the full data set and a restricted data set. CL models consistently outperform CEL models with differences ranging from 0.04 to 0.15 for AUPRC and 0.05 to 0.1 for AUROC.

Sex-Specific Cross Tissue Meta-Analysis Identifies Immune Dysregulation in Women With Alzheimer's Disease

Background: Alzheimer's disease (AD) is a progressive neurodegenerative disorder and the most common cause of dementia in the United States. In spite of evidence of females having a greater lifetime risk of developing Alzheimer's Disease (AD) and greater apolipoprotein E4-related (APOE ε4) AD risk compared to males, molecular signatures underlying these differences remain elusive. Methods: We took a meta-analysis approach to study gene expression in the brains of 1,084 AD patients and age-matched controls and whole blood from 645 AD patients and age-matched controls in seven independent datasets. Sex-specific gene expression patterns were investigated through use of gene-based, pathway-based and network-based approaches. The ability of a sex-specific AD gene expression signature to distinguish Alzheimer's disease from healthy controls was assessed using a linear support vector machine model. Cell type deconvolution from whole blood gene expression data was performed to identify differentially regulated cells in males and females with AD. Results: Strikingly gene-expression, network-based analysis and cell type deconvolution approaches revealed a consistent immune signature in the brain and blood of female AD patients that was absent in males. In females, network-based analysis revealed a coordinated program of gene expression involving several zinc finger nuclease genes related to Herpes simplex viral infection whose expression was modulated by the presence of the APOE ε4 allele. Interestingly, this gene expression program was missing in the brains of male AD patients. Cell type deconvolution identified an increase in neutrophils and naïve B cells and a decrease in M2 macrophages, memory B cells, and CD8+ T cells in AD samples compared to controls in females. Interestingly, among males with AD, no significant differences in immune cell proportions compared to controls were observed. Machine learning-based classification of AD using gene expression from whole blood in addition to clinical features produced an improvement in classification accuracy upon stratifying by sex, achieving an AUROC of 0.91 for females and 0.80 for males. Conclusion: These results help identify sex and APOE ε4 genotype-specific transcriptomic signatures of AD and underscore the importance of considering sex in the development of biomarkers and therapeutic strategies for AD.

Using Deep-Learning Algorithms to Simultaneously Identify Right and Left Ventricular Dysfunction From the Electrocardiogram

OBJECTIVES

This study sought to develop DL models capable of comprehensively quantifying left and right ventricular dysfunction from ECG data in a large, diverse population.

BACKGROUND

Rapid evaluation of left and right ventricular function using deep learning (DL) on electrocardiograms (ECGs) can assist diagnostic workflow. However, DL tools to estimate right- ventricular (RV) function do not exist, whereas those to estimate left ventricular (LV) function are restricted to quantification of very low LV function only.

METHODS

A multicenter study was conducted with data from 5 New York City hospitals: 4 for internal testing and 1 serving as external validation. We created novel DL models to classify left ventricular ejection fraction (LVEF) into categories derived from the latest universal definition of heart failure, estimate LVEF through regression, and predict a composite outcome of either RV systolic dysfunction or RV dilation.

RESULTS

We obtained echocardiogram LVEF estimates for 147,636 patients paired to 715,890 ECGs. We used natural language processing (NLP) to extract RV size and systolic function information from 404,502 echocardiogram reports paired to 761,510 ECGs for 148,227 patients. For LVEF classification in internal testing, area under curve (AUC) at detection of LVEF ≤40%, 40% < LVEF ≤50%, and LVEF >50% was 0.94 (95% CI: 0.94-0.94), 0.82 (95% CI: 0.81-0.83), and 0.89 (95% CI: 0.89-0.89), respectively. For external validation, these results were 0.94 (95% CI: 0.94-0.95), 0.73 (95% CI: 0.72-0.74), and 0.87 (95% CI: 0.87-0.88). For regression, the mean absolute error was 5.84% (95% CI: 5.82%-5.85%) for internal testing and 6.14% (95% CI: 6.13%-6.16%) in external validation. For prediction of the composite RV outcome, AUC was 0.84 (95% CI: 0.84-0.84) in both internal testing and external validation.

CONCLUSIONS

DL on ECG data can be used to create inexpensive screening, diagnostic, and predictive tools for both LV and RV dysfunction. Such tools may bridge the applicability of ECGs and echocardiography and enable prioritization of patients for further interventions for either sided failure progressing to biventricular disease.

Acute COVID-19 gene-expression profiles show multiple etiologies of long-term sequelae

Two years into the SARS-CoV-2 pandemic, the post-acute sequelae of infection are compounding the global health crisis. Often debilitating, these sequelae are clinically heterogeneous and of unknown molecular etiology. Here, a transcriptome-wide investigation of this new condition was performed in a large cohort of acutely infected patients followed clinically into the post-acute period. Gene expression signatures of post-acute sequelae were already present in whole blood during the acute phase of infection, with both innate and adaptive immune cells involved. Plasma cells stood out as driving at least two distinct clusters of sequelae, one largely dependent on circulating antibodies against the SARS-CoV-2 spike protein and the other antibody-independent. Altogether, multiple etiologies of post-acute sequelae were found concomitant with SARS-CoV-2 infection, directly linking the emergence of these sequelae with the host response to the virus.

Experimental and real-world evidence supporting the computational repurposing of bumetanide for APOE4-related Alzheimer's disease

The evident genetic, pathological, and clinical heterogeneity of Alzheimer's disease (AD) poses challenges for traditional drug development. We conducted a computational drug repurposing screen for drugs to treat apolipoprotein (apo) E4-related AD. We first established apoE-genotype-dependent transcriptomic signatures of AD by analyzing publicly-available human brain database. We then queried these signatures against the Connectivity Map database containing transcriptomic perturbations of >1300 drugs to identify those that best reverse apoE-genotype-specific AD signatures. Bumetanide was identified as a top drug for apoE4 AD. Bumetanide treatment of apoE4 mice without or with Aβ accumulation rescued electrophysiological, pathological, or cognitive deficits. Single-nucleus RNA-sequencing revealed transcriptomic reversal of AD signatures in specific cell types in these mice, a finding confirmed in apoE4-iPSC-derived neurons. In humans, bumetanide exposure was associated with a significantly lower AD prevalence in individuals over the age of 65 in two electronic health record databases, suggesting effectiveness of bumetanide in preventing AD.

Phe2vec: Automated disease phenotyping based on unsupervised embeddings from electronic health records

Robust phenotyping of patients from electronic health records (EHRs) at scale is a challenge in clinical informatics. Here, we introduce Phe2vec, an automated framework for disease phenotyping from EHRs based on unsupervised learning and assess its effectiveness against standard rule-based algorithms from Phenotype KnowledgeBase (PheKB). Phe2vec is based on pre-computing embeddings of medical concepts and patients' clinical history. Disease phenotypes are then derived from a seed concept and its neighbors in the embedding space. Patients are linked to a disease if their embedded representation is close to the disease phenotype. Comparing Phe2vec and PheKB cohorts head-to-head using chart review, Phe2vec performed on par or better in nine out of ten diseases. Differently from other approaches, it can scale to any condition and was validated against widely adopted expert-based standards. Phe2vec aims to optimize clinical informatics research by augmenting current frameworks to characterize patients by condition and derive reliable disease cohorts.

Deep Learning with Heterogeneous Graph Embeddings for Mortality Prediction from Electronic Health Records

Computational prediction of in-hospital mortality in the setting of an intensive care unit can help clinical practitioners to guide care and make early decisions for interventions. As clinical data are complex and varied in their structure and components, continued innovation of modelling strategies is required to identify architectures that can best model outcomes. In this work, we trained a Heterogeneous Graph Model (HGM) on electronic health record (EHR) data and used the resulting embedding vector as additional information added to a Convolutional Neural Network (CNN) model for predicting in-hospital mortality. We show that the additional information provided by including time as a vector in the embedding captured the relationships between medical concepts, lab tests, and diagnoses, which enhanced predictive performance. We found that adding HGM to a CNN model increased the mortality prediction accuracy up to 4%. This framework served as a foundation for future experiments involving different EHR data types on important healthcare prediction tasks.