Scalable and accurate deep learning with electronic health records

Early detection of pediatric health risks using maternal and child health data

Machine learning (ML)-driven diagnosis systems are particularly relevant in pediatrics given the well-documented impact of early-life health conditions on later-life outcomes. Yet, early identification of diseases and their subsequent impact on length of hospital stay for this age group has so far remained uncharacterized, likely because access to relevant health data is severely limited. Thanks to a confidential data use agreement with the California Department of Health Care Access and Information, we introduce Ped-BERT: a state-of-the-art deep learning model that accurately predicts the likelihood of 100+ conditions and the length of stay in a pediatric patient's next medical visit. We link mother-specific pre- and postnatal period health information to pediatric patient hospital discharge and emergency room visits. Our data set comprises 513.9K mother-baby pairs and contains medical diagnosis codes, length of stay, as well as temporal and spatial pediatric patient characteristics, such as age and residency zip code at the time of visit. Following the popular bidirectional encoder representations from the transformers (BERT) approach, we pre-train Ped-BERT via the masked language modeling objective to learn embedding features for the diagnosis codes contained in our data. We then continue to fine-tune our model to accurately predict primary diagnosis outcomes and length of stay for a pediatric patient's next visit, given the history of previous visits and, optionally, the mother's pre- and postnatal health information. We find that Ped-BERT generally outperforms contemporary and state-of-the-art classifiers when trained with minimum features. We also find that incorporating mother health attributes leads to significant improvements in model performance overall and across all patient subgroups in our data. Our most successful Ped-BERT model configuration achieves an area under the receiver operator curve (ROC AUC) of 0.927 and an average precision score (APS) of 0.408 for the diagnosis prediction task, and a ROC AUC of 0.855 and APS of 0.815 for the length of hospital stay task. Further, we examine Ped-BERT's fairness by determining whether prediction errors are evenly distributed across various subgroups of mother-baby demographics and health characteristics, or if certain subgroups exhibit a higher susceptibility to prediction errors.

Accurate prediction of all-cause mortality in patients with metabolic dysfunction-associated steatotic liver disease using electronic health records

INTRODUCTION AND OBJECTIVES

Despite the huge clinical burden of MASLD, validated tools for early risk stratification are lacking, and heterogeneous disease expression and a highly variable rate of progression to clinical outcomes result in prognostic uncertainty. We aimed to investigate longitudinal electronic health record-based outcome prediction in MASLD using a state-of-the-art machine learning model.

PATIENTS AND METHODS

n=940 patients with histologically-defined MASLD were used to develop a deep-learning model for all-cause mortality prediction. Patient timelines, spanning 12 years, were fully-annotated with demographic/clinical characteristics, ICD-9 and -10 codes, blood test results, prescribing data, and secondary care activity. A Transformer neural network (TNN) was trained to output concomitant probabilities of 12-, 24-, and 36-month all-cause mortality. In-sample performance was assessed using 5-fold cross-validation. Out-of-sample performance was assessed in an independent set of n=528 MASLD patients.

RESULTS

In-sample model performance achieved AUROC curve 0.74-0.90 (95% CI: 0.72-0.94), sensitivity 64%-82%, specificity 75%-92% and Positive Predictive Value (PPV) 94%-98%. Out-of-sample model validation had AUROC 0.70-0.86 (95% CI: 0.67-0.90), sensitivity 69%-70%, specificity 96%-97% and PPV 75%-77%. Key predictive factors, identified using coefficients of determination, were age, presence of type 2 diabetes, and history of hospital admissions with length of stay >14 days.

CONCLUSIONS

A TNN, applied to routinely-collected longitudinal electronic health records, achieved good performance in prediction of 12-, 24-, and 36-month all-cause mortality in patients with MASLD. Extrapolation of our technique to population-level data will enable scalable and accurate risk stratification to identify people most likely to benefit from anticipatory health care and personalized interventions.

Converting OMOP CDM to phenopackets: A model alignment and patient data representation evaluation

OBJECTIVE

This study aims to promote interoperability in precision medicine and translational research by aligning the Observational Medical Outcomes Partnership (OMOP) and Phenopackets data models. Phenopackets is an expert knowledge-driven schema designed to facilitate the storage and exchange of multimodal patient data, and support downstream analysis. The first goal of this paper is to explore model alignment by characterizing the common data models using a newly developed data transformation process and evaluation method. Second, using OMOP normalized clinical data, we evaluate the mapping of real-world patient data to Phenopackets. We evaluate the suitability of Phenopackets as a patient data representation for real-world clinical cases.

METHODS

We identified mappings between OMOP and Phenopackets and applied them to a real patient dataset to assess the transformation's success. We analyzed gaps between the models and identified key considerations for transforming data between them. Further, to improve ambiguous alignment, we incorporated Unified Medical Language System (UMLS) semantic type-based filtering to direct individual concepts to their most appropriate domain and conducted a domain-expert evaluation of the mapping's clinical utility.

RESULTS

The OMOP to Phenopacket transformation pipeline was executed for 1,000 Alzheimer's disease patients and successfully mapped all required entities. However, due to missing values in OMOP for required Phenopacket attributes, 10.2 % of records were lost. The use of UMLS-semantic type filtering for ambiguous alignment of individual concepts resulted in 96 % agreement with clinical thinking, increased from 68 % when mapping exclusively by domain correspondence.

CONCLUSION

This study presents a pipeline to transform data from OMOP to Phenopackets. We identified considerations for the transformation to ensure data quality, handling restrictions for successful Phenopacket validation and discrepant data formats. We identified unmappable Phenopacket attributes that focus on specialty use cases, such as genomics or oncology, which OMOP does not currently support. We introduce UMLS semantic type filtering to resolve ambiguous alignment to Phenopacket entities to be most appropriate for real-world interpretation. We provide a systematic approach to align OMOP and Phenopackets schemas. Our work facilitates future use of Phenopackets in clinical applications by addressing key barriers to interoperability when deriving a Phenopacket from real-world patient data.

Data mining and machine learning in HIV infection risk research: An overview and recommendations

In the contemporary era, the applications of data mining and machine learning have permeated extensively into medical research, significantly contributing to areas such as HIV studies. By reviewing 38 articles published in the past 15 years, the study presents a roadmap based on seven different aspects, utilizing various machine learning techniques for both novice researchers and experienced researchers seeking to comprehend the current state of the art in this area. While traditional regression modeling techniques have been commonly used, researchers are increasingly adopting more advanced fully supervised machine learning and deep learning techniques, which often outperform the traditional methods in predictive performance. Additionally, the study identifies nine new open research issues and outlines possible future research plans to enhance the outcomes of HIV infection risk research. This review is expected to be an insightful guide for researchers, illuminating current practices and suggesting advancements in the field.

Multimodal Machine Learning for Prediction of 30-Day Readmission Risk in Elderly Population

BACKGROUND

Readmission within 30 days is a prevalent issue among elderly patients, linked to unfavorable health outcomes. Our objective was to develop and validate multimodal machine learning models for predicting 30-day readmission risk in elderly patients discharged from internal medicine departments.

METHODS

This was a retrospective cohort study which included elderly patients aged 75 or older, who were hospitalized at the Hadassah Medical Center internal medicine departments between 2014 and 2020. Three machine learning algorithms were developed and employed to predict 30-day readmission risk. The primary measures were predictive model performance scores, specifically area under the receiver operator curve (AUROC), and average precision.

RESULTS

This study included 19,569 admissions. Of them, 3258 (16.65%) resulted in 30-day readmission. Our 3 proposed models demonstrated high accuracy and precision on an unseen test set, with AUROC values of 0.87, 0.89, and 0.93, respectively, and average precision values of 0.76, 0.78, and 0.81. Feature importance analysis revealed that the number of admissions in the past year, history of 30-day readmission, Charlson score, and admission length were the most influential variables. Notably, the natural language processing score, representing the probability of readmission according to a textual-based model trained on social workers' assessment letters during hospitalization, ranked among the top 10 contributing factors.

CONCLUSIONS

Leveraging multimodal machine learning offers a promising strategy for identifying elderly patients who are at high risk for 30-day readmission. By identifying these patients, machine learning models may facilitate the effective execution of preventive actions to reduce avoidable readmission incidents.

A multi-center study on the adaptability of a shared foundation model for electronic health records

Foundation models are transforming artificial intelligence (AI) in healthcare by providing modular components adaptable for various downstream tasks, making AI development more scalable and cost-effective. Foundation models for structured electronic health records (EHR), trained on coded medical records from millions of patients, demonstrated benefits including increased performance with fewer training labels, and improved robustness to distribution shifts. However, questions remain on the feasibility of sharing these models across hospitals and their performance in local tasks. This multi-center study examined the adaptability of a publicly accessible structured EHR foundation model (FMSM), trained on 2.57 M patient records from Stanford Medicine. Experiments used EHR data from The Hospital for Sick Children (SickKids) and Medical Information Mart for Intensive Care (MIMIC-IV). We assessed both adaptability via continued pretraining on local data, and task adaptability compared to baselines of locally training models from scratch, including a local foundation model. Evaluations on 8 clinical prediction tasks showed that adapting the off-the-shelf FMSM matched the performance of gradient boosting machines (GBM) locally trained on all data while providing a 13% improvement in settings with few task-specific training labels. Continued pretraining on local data showed FMSM required fewer than 1% of training examples to match the fully trained GBM's performance, and was 60 to 90% more sample-efficient than training local foundation models from scratch. Our findings demonstrate that adapting EHR foundation models across hospitals provides improved prediction performance at less cost, underscoring the utility of base foundation models as modular components to streamline the development of healthcare AI.

Group-informed attentive framework for enhanced diabetes mellitus progression prediction

The increasing prevalence of Diabetes Mellitus (DM) as a global health concern highlights the paramount importance of accurately predicting its progression. This necessity has propelled the use of deep learning's advanced analytical and predictive capabilities to the forefront of current research. However, this approach is confronted with significant challenges, notably the prevalence of incomplete data and the need for more robust predictive models. Our research aims to address these critical issues, leveraging deep learning to enhance the precision and reliability of diabetes progression predictions. We address the issue of missing data by first locating individuals with data gaps within specific patient clusters, and then applying targeted imputation strategies for effective data imputation. To enhance the robustness of our model, we implement strategies such as data augmentation and the development of advanced group-level feature analysis. A cornerstone of our approach is the implementation of a deep attentive transformer that is sensitive to group characteristics. This framework excels in processing a wide array of data, including clinical and physical examination information, to accurately predict the progression of DM. Beyond its predictive capabilities, our model is engineered to perform advanced feature selection and reasoning. This is crucial for understanding the impact of both individual and group-level factors on deep models' predictions, providing invaluable insights into the dynamics of DM progression. Our approach not only marks a significant advancement in the prediction of diabetes progression but also contributes to a deeper understanding of the multifaceted factors influencing this chronic disease, thereby aiding in more effective diabetes management and research.

Deep learning-based prediction of one-year mortality in Finland is an accurate but unfair aging marker

Short-term mortality risk, which is indicative of individual frailty, serves as a marker for aging. Previous age clocks focused on predicting either chronological age or longer-term mortality. Aging clocks predicting short-term mortality are lacking and their algorithmic fairness remains unexamined. We developed a deep learning model to predict 1-year mortality using nationwide longitudinal data from the Finnish population (FinRegistry; n = 5.4 million), incorporating more than 8,000 features spanning up to 50 years. We achieved an area under the curve (AUC) of 0.944, outperforming a baseline model that included only age and sex (AUC = 0.897). The model generalized well to different causes of death (AUC > 0.800 for 45 of 50 causes), including coronavirus disease 2019, which was absent in the training data. Performance varied among demographics, with young females exhibiting the best and older males the worst results. Extensive prediction fairness analyses highlighted disparities among disadvantaged groups, posing challenges to equitable integration into public health interventions. Our model accurately identified short-term mortality risk, potentially serving as a population-wide aging marker.

Explainable text-tabular models for predicting mortality risk in companion animals

As interest in using machine learning models to support clinical decision-making increases, explainability is an unequivocal priority for clinicians, researchers and regulators to comprehend and trust their results. With many clinical datasets containing a range of modalities, from the free-text of clinician notes to structured tabular data entries, there is a need for frameworks capable of providing comprehensive explanation values across diverse modalities. Here, we present a multimodal masking framework to extend the reach of SHapley Additive exPlanations (SHAP) to text and tabular datasets to identify risk factors for companion animal mortality in first-opinion veterinary electronic health records (EHRs) from across the United Kingdom. The framework is designed to treat each modality consistently, ensuring uniform and consistent treatment of features and thereby fostering predictability in unimodal and multimodal contexts. We present five multimodality approaches, with the best-performing method utilising PetBERT, a language model pre-trained on a veterinary dataset. Utilising our framework, we shed light for the first time on the reasons each model makes its decision and identify the inclination of PetBERT towards a more pronounced engagement with free-text narratives compared to BERT-base's predominant emphasis on tabular data. The investigation also explores the important features on a more granular level, identifying distinct words and phrases that substantially influenced an animal's life status prediction. PetBERT showcased a heightened ability to grasp phrases associated with veterinary clinical nomenclature, signalling the productivity of additional pre-training of language models.

Web-based artificial intelligence to predict cognitive impairment following stroke: A multicenter study

BACKGROUND AND PURPOSE

Post-stroke cognitive impairment (PSCI) is highly prevalent in modern society. However, there is limited study implying an accurate and explainable machine learning model to predict PSCI. The aim of this study is to develop and validate a web-based artificial intelligence (AI) tool for predicting PSCI.

METHODS

The retrospective cohort study design was conducted to develop and validate a web-based prediction model. Adults who experienced a stroke between January 1, 2004, and September 30, 2017, were enrolled, and patients with PSCI were followed up from the stroke index date until their last follow-up. The model's performance metrics, including accuracy, area under the curve (AUC), recall, precision, and F1 score, were compared.

RESULTS

A total of 3209 stroke patients were included in the study. The model demonstrated an accuracy of 0.8793, AUC of 0.9200, recall of 0.6332, precision of 0.9664, and F1 score of 0.7651. In the external validation phase, the accuracy improved to 0.9039, AUC to 0.9094, recall to 0.7457, precision to 0.9168, and F1 score to 0.8224. The final model can be accessed at https://psci-calculator.my.id/.

CONCLUSION

Our results are able to produce a user-friendly interface that is useful for health practitioners to perform early prediction on PSCI. These findings also suggest that the provided AI model is reliable and can serve as a roadmap for future studies using AI models in a clinical setting.

Revolutionizing Acute Cardiac Care with Artificial Intelligence: Opportunities and Challenges

This manuscript reviews the application of artificial intelligence (AI) in acute cardiac care, highlighting its potential to transform patient outcomes in the face of the global burden of cardiovascular diseases. It explores how AI algorithms can rapidly and accurately process data for the prediction and diagnosis of acute cardiac conditions. The paper examines AI's impact on patient health across various diagnostic tools such as echocardiography, electrocardiography, coronary angiography, cardiac CT, and MRI and discusses the regulatory landscape for AI in healthcare, categorizes AI algorithms by their risk levels. Furthermore, it addresses the challenges of data quality, generalizability, bias, transparency, and regulatory considerations, underscoring the necessity for inclusive data and robust validation processes. The review concludes with future perspectives on integrating AI into clinical workflows and the ongoing need for research, regulation, and innovation to harness AI's full potential in improving acute cardiac care.

Machine learning approach as an early warning system to prevent foodborne Salmonella outbreaks in northwestern Italy

Salmonellosis, one of the most common foodborne infections in Europe, is monitored by food safety surveillance programmes, resulting in the generation of extensive databases. By leveraging tree-based machine learning (ML) algorithms, we exploited data from food safety audits to predict spatiotemporal patterns of salmonellosis in northwestern Italy. Data on human cases confirmed in 2015-2018 (n = 1969) and food surveillance data collected in 2014-2018 were used to develop ML algorithms. We integrated the monthly municipal human incidence with 27 potential predictors, including the observed prevalence of Salmonella in food. We applied the tree regression, random forest and gradient boosting algorithms considering different scenarios and evaluated their predictivity in terms of the mean absolute percentage error (MAPE) and R2. Using a similar dataset from the year 2019, spatiotemporal predictions and their relative sensitivities and specificities were obtained. Random forest and gradient boosting (R2 = 0.55, MAPE = 7.5%) outperformed the tree regression algorithm (R2 = 0.42, MAPE = 8.8%). Salmonella prevalence in food; spatial features; and monitoring efforts in ready-to-eat milk, fruits and vegetables, and pig meat products contributed the most to the models' predictivity, reducing the variance by 90.5%. Conversely, the number of positive samples obtained for specific food matrices minimally influenced the predictions (2.9%). Spatiotemporal predictions for 2019 showed sensitivity and specificity levels of 46.5% (due to the lack of some infection hotspots) and 78.5%, respectively. This study demonstrates the added value of integrating data from human and veterinary health services to develop predictive models of human salmonellosis occurrence, providing early warnings useful for mitigating foodborne disease impacts on public health.

Can Machine Learning Correctly Predict Outcomes of Flexible Ureteroscopy with Laser Lithotripsy for Kidney Stone Disease? Results from a Large Endourology University Centre

Background and objective

The integration of machine learning (ML) in health care has garnered significant attention because of its unprecedented opportunities to enhance patient care and outcomes. In this study, we trained ML algorithms for automated prediction of outcomes of ureteroscopic laser lithotripsy (URSL) on the basis of preoperative characteristics.

Methods

Data were retrieved for patients treated with ureteroscopy for urolithiasis by a single experienced surgeon over a 7-yr period. Sixteen ML classification algorithms were trained to investigate correlation between preoperative characteristics and postoperative outcomes. The outcomes assessed were primary stone-free status (SFS, defined as the presence of only stone fragments <2 mm on endoscopic visualisation and at 3-mo imaging) and postoperative complications. An ensemble model was constructed from the best-performing algorithms for prediction of complications and for prediction of SFS. Simultaneous prediction of postoperative characteristics was then investigated using a multitask neural network, and explainable artificial intelligence (AI) was used to demonstrate the predictive power of the best models.

Key findings and limitations

An ensemble ML model achieved accuracy of 93% and precision of 87% for prediction of SFS. Complications were mainly associated with a preoperative positive urine culture (1.44). Logistic regression revealed that SFS was impacted by the total stone burden (0.34), the presence of a preoperative stent (0.106), a positive preoperative urine culture (0.14), and stone location (0.09). Explainable AI results emphasised the key features and their contributions to the output.

Conclusions and clinical implications

Technological advances are helping urologists to overcome the classic limits of ureteroscopy, namely stone size and the risk of complications. ML represents an excellent aid for correct prediction of outcomes after training on pre-existing data sets. Our ML model achieved accuracy of >90% for prediction of SFS and complications, and represents a basis for the development of an accessible predictive model for endourologists and patients in the URSL setting.

Patient summary

We tested the ability of artificial intelligence to predict treatment outcomes for patients with kidney stones. We trained 16 different machine learning tools with data before surgery, such as patient age and the stone characteristics. Our final model was >90% accurate in predicting stone-free status after surgery and the occurrence of complications.

Predicting intubation for intensive care units patients: A deep learning approach to improve patient management

OBJECTIVE

For patients in the Intensive Care Unit (ICU), the timing of intubation has a significant association with patients' outcomes. However, accurate prediction of the timing of intubation remains an unsolved challenge due to the noisy, sparse, heterogeneous, and unbalanced nature of ICU data. In this study, our objective is to develop a workflow for pre-processing ICU data and to develop a customized deep learning model to predict the need for intubation.

METHODS

To improve the prediction accuracy, we transform the intubation prediction task into a time series classification task. We carefully design a sequence of data pre-processing steps to handle the multimodal noisy data. Firstly, we discretize the sequential data and address missing data using interpolation. Next, we employ a sampling strategy to address data imbalance and standardize the data to facilitate faster model convergence. Furthermore, we employ the feature selection technique and propose an ensemble model to combine features learned by different deep learning models.

RESULTS

The performance is evaluated on Medical Information Mart for Intensive Care (MIMIC)-III, an ICU dataset. Our proposed Deep Feature Fusion method achieves an area under the curve (AUC) of the receiver operating curve (ROC) of 0.8953, surpassing the performance of other deep learning and traditional machine learning models.

CONCLUSION

Our proposed Deep Feature Fusion method proves to be a viable approach for predicting intubation and outperforms other deep learning and classical machine learning models. The study confirms that high-frequency time-varying indicators, particularly Mean Blood Pressure (MeanBP) and peripheral oxygen saturation (SpO2), are significant risk factors for predicting intubation.

The Role of Artificial Intelligence in Healthcare: Enhancing Coronary Computed Tomography Angiography for Coronary Artery Disease Management

This review aims to explore the potential of artificial intelligence (AI) in coronary CT angiography (CCTA), a key tool for diagnosing coronary artery disease (CAD). Because CAD is still a major cause of death worldwide, effective and accurate diagnostic methods are required to identify and manage the condition. CCTA certainly is a noninvasive alternative for diagnosing CAD, but it requires a large amount of data as input. We intend to discuss the idea of incorporating AI into CCTA, which enhances its diagnostic accuracy and operational efficiency. Using such AI technologies as machine learning (ML) and deep learning (DL) tools, CCTA images are automated to perfection and the analysis is significantly refined. It enables the characterization of a plaque, assesses the severity of the stenosis, and makes more accurate risk stratifications than traditional methods, with pinpoint accuracy. Automating routine tasks through AI-driven CCTA will reduce the radiologists' workload considerably, which is a standard benefit of such technologies. More importantly, it would enable radiologists to allocate more time and expertise to complex cases, thereby improving overall patient care. However, the field of AI in CCTA is not without its challenges, which include data protection, algorithm transparency, as well as criteria for standardization encoding. Despite such obstacles, it appears that the integration of AI technology into CCTA in the future holds great promise for keeping CAD itself in check, thereby aiding the fight against this disease and begetting better clinical outcomes and more optimized modes of healthcare. Future research on AI algorithms for CCTA, making ethical use of AI, and thereby overcoming the technical and clinical barriers to widespread adoption of this new tool, will hopefully pave the way for profound AI-driven transformations in healthcare.

Pathophysiological Features in Electronic Medical Records Sustain Model Performance under Temporal Dataset Shift

Access to real-world data streams like electronic medical records (EMRs) has accelerated the development of supervised machine learning (ML) models for clinical applications. However, few studies investigate the differential impact of particular features in the EMR on model performance under temporal dataset shift. To explain how features in the EMR impact models over time, this study aggregates features into feature groups by their source (e.g. medication orders, diagnosis codes and lab results) and feature categories based on their reflection of patient pathophysiology or healthcare processes. We adapt Shapley values to explain feature groups' and feature categories' marginal contribution to initial and sustained model performance. We investigate three standard clinical prediction tasks and find that while feature contributions to initial performance differ across tasks, pathophysiological features help mitigate temporal discrimination deterioration. These results provide interpretable insights on how specific feature groups contribute to model performance and robustness to temporal dataset shift.

Optimizing Acute Coronary Syndrome Patient Treatment: Leveraging Gated Transformer Models for Precise Risk Prediction and Management

BACKGROUND

Acute coronary syndrome (ACS) is a severe cardiovascular disease with globally rising incidence and mortality rates. Traditional risk assessment tools are widely used but are limited due to the complexity of the data.

METHODS

This study introduces a gated Transformer model utilizing machine learning to analyze electronic health records (EHRs) for an enhanced prediction of major adverse cardiovascular events (MACEs) in ACS patients. The model's efficacy was evaluated using metrics such as area under the curve (AUC), precision-recall (PR), and F1-scores. Additionally, a patient management platform was developed to facilitate personalized treatment strategies.

RESULTS

Incorporating a gating mechanism substantially improved the Transformer model's performance, especially in identifying true-positive cases. The TabTransformer+Gate model demonstrated an AUC of 0.836, a 14% increase in average precision (AP), and a 6.2% enhancement in accuracy, significantly outperforming other deep learning approaches. The patient management platform enabled healthcare professionals to effectively assess patient risks and tailor treatments, improving patient outcomes and quality of life.

CONCLUSION

The integration of a gating mechanism within the Transformer model markedly increases the accuracy of MACE risk predictions in ACS patients, optimizes personalized treatment, and presents a novel approach for advancing clinical practice and research.

A machine learning model to predict the risk of perinatal depression: Psychosocial and sleep-related factors in the Life-ON study cohort

Perinatal depression (PND) is a common complication of pregnancy associated with serious health consequences for both mothers and their babies. Identifying risk factors for PND is key to early detect women at increased risk of developing this condition. We applied a machine learning (ML) approach to data from a multicenter cohort study on sleep and mood changes during the perinatal period ("Life-ON") to derive models for PND risk prediction in a cross-validation setting. A wide range of sociodemographic variables, blood-based biomarkers, sleep, medical, and psychological data collected from 439 pregnant women, as well as polysomnographic parameters recorded from 353 women, were considered for model building. These covariates were correlated with the risk of future depression, as assessed by regularly administering the Edinburgh Postnatal Depression Scale across the perinatal period. The ML model indicated the mood status of pregnant women in the first trimester, previous depressive episodes and marital status, as the most important predictors of PND. Sleep quality, insomnia symptoms, age, previous miscarriages, and stressful life events also added to the model performance. Besides other predictors, sleep changes during early pregnancy should therefore assessed to identify women at higher risk of PND and support them with appropriate therapeutic strategies.

Medical history predicts phenome-wide disease onset and enables the rapid response to emerging health threats

The COVID-19 pandemic exposed a global deficiency of systematic, data-driven guidance to identify high-risk individuals. Here, we illustrate the utility of routinely recorded medical history to predict the risk for 1883 diseases across clinical specialties and support the rapid response to emerging health threats such as COVID-19. We developed a neural network to learn from health records of 502,460 UK Biobank. Importantly, we observed discriminative improvements over basic demographic predictors for 1774 (94.3%) endpoints. After transferring the unmodified risk models to the All of US cohort, we replicated these improvements for 1347 (89.8%) of 1500 investigated endpoints, demonstrating generalizability across healthcare systems and historically underrepresented groups. Ultimately, we showed how this approach could have been used to identify individuals vulnerable to severe COVID-19. Our study demonstrates the potential of medical history to support guidance for emerging pandemics by systematically estimating risk for thousands of diseases at once at minimal cost.

Artificial Intelligence in the Future Landscape of Pediatric Neuroradiology: Opportunities and Challenges

This paper will review how artificial intelligence (AI) will play an increasingly important role in pediatric neuroradiology in the future. A safe, transparent, and human-centric AI is needed to tackle the quadruple aim of improved health outcomes, enhanced patient and family experience, reduced costs, and improved well-being of the healthcare team in pediatric neuroradiology. Equity, diversity and inclusion, data safety, and access to care will need to always be considered. In the next decade, AI algorithms are expected to play an increasingly important role in access to care, workflow management, abnormality detection, classification, response prediction, prognostication, report generation, as well as in the patient and family experience in pediatric neuroradiology. Also, AI algorithms will likely play a role in recognizing and flagging rare diseases and in pattern recognition to identify previously unknown disorders. While AI algorithms will play an important role, humans will not only need to be in the loop, but in the center of pediatric neuroimaging. AI development and deployment will need to be closely watched and monitored by experts in the field. Patient and data safety need to be at the forefront, and the risks of a dependency on technology will need to be contained. The applications and implications of AI in pediatric neuroradiology will differ from adult neuroradiology.

Personalized anesthesia and precision medicine: a comprehensive review of genetic factors, artificial intelligence, and patient-specific factors

Precision medicine, characterized by the personalized integration of a patient's genetic blueprint and clinical history, represents a dynamic paradigm in healthcare evolution. The emerging field of personalized anesthesia is at the intersection of genetics and anesthesiology, where anesthetic care will be tailored to an individual's genetic make-up, comorbidities and patient-specific factors. Genomics and biomarkers can provide more accurate anesthetic protocols, while artificial intelligence can simplify anesthetic procedures and reduce anesthetic risks, and real-time monitoring tools can improve perioperative safety and efficacy. The aim of this paper is to present and summarize the applications of these related fields in anesthesiology by reviewing them, exploring the potential of advanced technologies in the implementation and development of personalized anesthesia, realizing the future integration of new technologies into clinical practice, and promoting multidisciplinary collaboration between anesthesiology and disciplines such as genomics and artificial intelligence.

Electronic health record signatures identify undiagnosed patients with common variable immunodeficiency disease

Human inborn errors of immunity include rare disorders entailing functional and quantitative antibody deficiencies due to impaired B cells called the common variable immunodeficiency (CVID) phenotype. Patients with CVID face delayed diagnoses and treatments for 5 to 15 years after symptom onset because the disorders are rare (prevalence of ~1/25,000), and there is extensive heterogeneity in CVID phenotypes, ranging from infections to autoimmunity to inflammatory conditions, overlapping with other more common disorders. The prolonged diagnostic odyssey drives excessive system-wide costs before diagnosis. Because there is no single causal mechanism, there are no genetic tests to definitively diagnose CVID. Here, we present PheNet, a machine learning algorithm that identifies patients with CVID from their electronic health records (EHRs). PheNet learns phenotypic patterns from verified CVID cases and uses this knowledge to rank patients by likelihood of having CVID. PheNet could have diagnosed more than half of our patients with CVID 1 or more years earlier than they had been diagnosed. When applied to a large EHR dataset, followed by blinded chart review of the top 100 patients ranked by PheNet, we found that 74% were highly probable to have CVID. We externally validated PheNet using >6 million records from disparate medical systems in California and Tennessee. As artificial intelligence and machine learning make their way into health care, we show that algorithms such as PheNet can offer clinical benefits by expediting the diagnosis of rare diseases.

Utilizing Big Data analytics and electronic health record data in HIV prevention, treatment, and care research: a literature review

Propelled by the transformative power of modern information and communication technologies, digitalization of data, and the increasing affordability of high-performance computing, Big Data science has brought forth revolutionary advancement in many areas of business, industry, health, and medicine. The HIV research and care service community is no exception to the benefits from the availability and utilization of Big Data analytics. Electronic health record (EHR) data (e.g., administrative and billing data, electronic medical records, or other digital records of information pertinent to individual or population health) are an essential source of health and disease outcome data because of the large amount of real-world, comprehensive, and often longitudinal data, which provide a good opportunity for leveraging advanced Big Data analytics in addressing challenges in HIV prevention, treatment, and care. This review focuses on studies that apply Big Data analytics to EHR data with aims to synthesize the HIV-related issues that EHR data studies can tackle, identify challenges in the utilization of EHR data in HIV research and practice, and discuss future needs and directions that can realize the promising potential role of Big Data in ending the HIV epidemic.

The Role of Artificial Intelligence in Cardiac Imaging

Artificial intelligence (AI) is having a significant impact in medical imaging, advancing almost every aspect of the field, from image acquisition and postprocessing to automated image analysis with outreach toward supporting decision making. Noninvasive cardiac imaging is one of the main and most exciting fields for AI development. The aim of this review is to describe the main applications of AI in cardiac imaging, including CT and MR imaging, and provide an overview of recent advancements and available clinical applications that can improve clinical workflow, disease detection, and prognostication in cardiac disease.

Overview and Clinical Applications of Artificial Intelligence and Machine Learning in Cardiac Anesthesiology

Artificial intelligence- (AI) and machine learning (ML)-based applications are becoming increasingly pervasive in the healthcare setting. This has in turn challenged clinicians, hospital administrators, and health policymakers to understand such technologies and develop frameworks for safe and sustained clinical implementation. Within cardiac anesthesiology, challenges and opportunities for AI/ML to support patient care are presented by the vast amounts of electronic health data, which are collected rapidly, interpreted, and acted upon within the periprocedural area. To address such challenges and opportunities, in this article, the authors review 3 recent applications relevant to cardiac anesthesiology, including depth of anesthesia monitoring, operating room resource optimization, and transthoracic/transesophageal echocardiography, as conceptual examples to explore strengths and limitations of AI/ML within healthcare, and characterize this evolving landscape. Through reviewing such applications, the authors introduce basic AI/ML concepts and methodologies, as well as practical considerations and ethical concerns for initiating and maintaining safe clinical implementation of AI/ML-based algorithms for cardiac anesthesia patient care.

Working memory load recognition with deep learning time series classification

Working memory load (WML) is one of the widely applied signals in the areas of human-machine interaction. The precise evaluation of the WML is crucial for this kind of application. This study aims to propose a deep learning (DL) time series classification (TSC) model for inter-subject WML decoding. We used fNIRS to record the hemodynamic signals of 27 participants during visual working memory tasks. Traditional machine learning and deep time series classification algorithms were respectively used for intra-subject and inter-subject WML decoding from the collected blood oxygen signals. The intra-subject classification accuracy of LDA and SVM were 94.6% and 79.1%. Our proposed TAResnet-BiLSTM model had the highest inter-subject WML decoding accuracy, reaching 92.4%. This study provides a new idea and method for the brain-computer interface application of fNIRS in real-time WML detection.

Differentiating viral and bacterial infections: A machine learning model based on routine blood test values

The growing threat of antibiotic resistance necessitates accurate differentiation between bacterial and viral infections for proper antibiotic administration. In this study, a Virus vs. Bacteria machine learning model was developed to distinguish between these infection types using 16 routine blood test results, C-reactive protein concentration (CRP), biological sex, and age. With a dataset of 44,120 cases from a single medical center, the model achieved an accuracy of 82.2 %, a sensitivity of 79.7 %, a specificity of 84.5 %, a Brier score of 0.129, and an area under the ROC curve (AUC) of 0.905, outperforming a CRP-based decision rule. Notably, the machine learning model enhanced accuracy within the CRP range of 10-40 mg/L, a range where CRP alone is less informative. These results highlight the advantage of integrating multiple blood parameters in diagnostics. The "Virus vs. Bacteria" model paves the way for advanced diagnostic tools, leveraging machine learning to optimize infection management.

Deep-Learning-Based Automated Anomaly Detection of EEGs in Intensive Care Units

An intensive care unit (ICU) is a special ward in the hospital for patients who require intensive care. It is equipped with many instruments monitoring patients' vital signs and supported by the medical staff. However, continuous monitoring demands a massive workload of medical care. To ease the burden, we aim to develop an automatic detection model to monitor when brain anomalies occur. In this study, we focus on electroencephalography (EEG), which monitors the brain electroactivity of patients continuously. It is mainly for the diagnosis of brain malfunction. We propose the gated-recurrent-unit-based (GRU-based) model for detecting brain anomalies; it predicts whether the spike or sharp wave happens within a short time window. Based on the banana montage setting, the proposed model exploits characteristics of multiple channels simultaneously to detect anomalies. It is trained, validated, and tested on separated EEG data and achieves more than 90% testing performance on sensitivity, specificity, and balanced accuracy. The proposed anomaly detection model detects the existence of a spike or sharp wave precisely; it will notify the ICU medical staff, who can provide immediate follow-up treatment. Consequently, it can reduce the medical workload in the ICU significantly.

Unmasking bias in artificial intelligence: a systematic review of bias detection and mitigation strategies in electronic health record-based models

OBJECTIVES

Leveraging artificial intelligence (AI) in conjunction with electronic health records (EHRs) holds transformative potential to improve healthcare. However, addressing bias in AI, which risks worsening healthcare disparities, cannot be overlooked. This study reviews methods to handle various biases in AI models developed using EHR data.

MATERIALS AND METHODS

We conducted a systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines, analyzing articles from PubMed, Web of Science, and IEEE published between January 01, 2010 and December 17, 2023. The review identified key biases, outlined strategies for detecting and mitigating bias throughout the AI model development, and analyzed metrics for bias assessment.

RESULTS

Of the 450 articles retrieved, 20 met our criteria, revealing 6 major bias types: algorithmic, confounding, implicit, measurement, selection, and temporal. The AI models were primarily developed for predictive tasks, yet none have been deployed in real-world healthcare settings. Five studies concentrated on the detection of implicit and algorithmic biases employing fairness metrics like statistical parity, equal opportunity, and predictive equity. Fifteen studies proposed strategies for mitigating biases, especially targeting implicit and selection biases. These strategies, evaluated through both performance and fairness metrics, predominantly involved data collection and preprocessing techniques like resampling and reweighting.

DISCUSSION

This review highlights evolving strategies to mitigate bias in EHR-based AI models, emphasizing the urgent need for both standardized and detailed reporting of the methodologies and systematic real-world testing and evaluation. Such measures are essential for gauging models' practical impact and fostering ethical AI that ensures fairness and equity in healthcare.

Serious Illness Communication in Cirrhosis Care: Tools to Improve Illness Understanding, Prognostic Understanding, and Care Planning

Patients with cirrhosis frequently experience an unpredictable illness trajectory, with frequent hospitalizations and complications. Along with the uncertain nature of the disease, the possibility of a lifesaving and curative transplant often makes prognostic discussions and future care decisions challenging. Serious illness communication (SIC) refers to supportive communication whereby clinicians assess patients' illness understanding, share prognostic information according to patients' preferences, explore patients' goals, and make recommendations for care that align with these goals. SIC includes 3 key components: (1) illness understanding; (2) prognostic understanding; and (3) care planning. In this piece, we explore current barriers to early implementation of SIC in cirrhosis care and share possible solutions, including adopting a multidisciplinary approach, delivering culturally competent care, and training clinicians in SIC core skills. By use of a case example, we aim to demonstrate SIC in action and to provide clinicians with tools and skills that can be used in practice.

Towards Transparent Healthcare: Advancing Local Explanation Methods in Explainable Artificial Intelligence

This paper focuses on the use of local Explainable Artificial Intelligence (XAI) methods, particularly the Local Rule-Based Explanations (LORE) technique, within healthcare and medical settings. It emphasizes the critical role of interpretability and transparency in AI systems for diagnosing diseases, predicting patient outcomes, and creating personalized treatment plans. While acknowledging the complexities and inherent trade-offs between interpretability and model performance, our work underscores the significance of local XAI methods in enhancing decision-making processes in healthcare. By providing granular, case-specific insights, local XAI methods like LORE enhance physicians' and patients' understanding of machine learning models and their outcome. Our paper reviews significant contributions to local XAI in healthcare, highlighting its potential to improve clinical decision making, ensure fairness, and comply with regulatory standards.

Predicting pressure injury risk in hospitalised patients using machine learning with electronic health records: a US multilevel cohort study

OBJECTIVE

To predict the risk of hospital-acquired pressure injury using machine learning compared with standard care.

DESIGN

We obtained electronic health records (EHRs) to structure a multilevel cohort of hospitalised patients at risk for pressure injury and then calibrate a machine learning model to predict future pressure injury risk. Optimisation methods combined with multilevel logistic regression were used to develop a predictive algorithm of patient-specific shifts in risk over time. Machine learning methods were tested, including random forests, to identify predictive features for the algorithm. We reported the results of the regression approach as well as the area under the receiver operating characteristics (ROC) curve for predictive models.

SETTING

Hospitalised inpatients.

PARTICIPANTS

EHRs of 35 001 hospitalisations over 5 years across 2 academic hospitals.

MAIN OUTCOME MEASURE

Longitudinal shifts in pressure injury risk.

RESULTS

The predictive algorithm with features generated by machine learning achieved significantly improved prediction of pressure injury risk (p<0.001) with an area under the ROC curve of 0.72; whereas standard care only achieved an area under the ROC curve of 0.52. At a specificity of 0.50, the predictive algorithm achieved a sensitivity of 0.75.

CONCLUSIONS

These data could help hospitals conserve resources within a critical period of patient vulnerability of hospital-acquired pressure injury which is not reimbursed by US Medicare; thus, conserving between 30 000 and 90 000 labour-hours per year in an average 500-bed hospital. Hospitals can use this predictive algorithm to initiate a quality improvement programme for pressure injury prevention and further customise the algorithm to patient-specific variation by facility.

Charting Tomorrow's Healthcare: A Traditional Literature Review for an Artificial Intelligence-Driven Future

Electronic health record (EHR) systems have developed over time in parallel with general advancements in mainstream technology. As artificially intelligent (AI) systems rapidly impact multiple societal sectors, it has become apparent that medicine is not immune from the influences of this powerful technology. Particularly appealing is how AI may aid in improving healthcare efficiency with note-writing automation. This literature review explores the current state of EHR technologies in healthcare, specifically focusing on possibilities for addressing EHR challenges through the automation of dictation and note-writing processes with AI integration. This review offers a broad understanding of existing capabilities and potential advancements, emphasizing innovations such as voice-to-text dictation, wearable devices, and AI-assisted procedure note dictation. The primary objective is to provide researchers with valuable insights, enabling them to generate new technologies and advancements within the healthcare landscape. By exploring the benefits, challenges, and future of AI integration, this review encourages the development of innovative solutions, with the goal of enhancing patient care and healthcare delivery efficiency.

Integrating clinical pharmacology and artificial intelligence: potential benefits, challenges, and role of clinical pharmacologists

INTRODUCTION

The integration of artificial intelligence (AI) into clinical pharmacology could be a potential approach for accelerating drug discovery and development, improving patient care, and streamlining medical research processes.

AREAS COVERED

We reviewed the current state of AI applications in clinical pharmacology, focusing on drug discovery and development, precision medicine, pharmacovigilance, and other ventures. Key AI applications in clinical pharmacology are examined, including machine learning, natural language processing, deep learning, and reinforcement learning etc. Additionally, the evolving role of clinical pharmacologists, ethical considerations, and challenges in implementing AI in clinical pharmacology are discussed.

EXPERT OPINION

The AI could be instrumental in accelerating drug discovery, predicting drug safety and efficacy, and optimizing clinical trial designs. It can play a vital role in precision medicine by helping in personalized drug dosing, treatment selection, and predicting drug response based on genetic, clinical, and environmental factors. The role of AI in pharmacovigilance, such as signal detection and adverse event prediction, is also promising. The collaboration between clinical pharmacologists and AI experts also poses certain ethical and practical challenges. Clinical pharmacologists can be instrumental in shaping the future of AI-driven clinical pharmacology and contribute to the improvement of healthcare systems.

Length of Stay Prediction With Standardized Hospital Data From Acute and Emergency Care Using a Deep Neural Network

OBJECTIVE

Length of stay (LOS) is an important metric for the organization and scheduling of care activities. This study sought to propose a LOS prediction method based on deep learning using widely available administrative data from acute and emergency care and compare it with other methods.

PATIENTS AND METHODS

All admissions between January 1, 2011 and December 31, 2019, at 6 university hospitals of the Hospices Civils de Lyon metropolis were included, leading to a cohort of 1,140,100 stays of 515,199 patients. Data included demographics, primary and associated diagnoses, medical procedures, the medical unit, the admission type, socio-economic factors, and temporal information. A model based on embeddings and a Feed-Forward Neural Network (FFNN) was developed to provide fine-grained LOS predictions per hospitalization step. Performances were compared with random forest and logistic regression, with the accuracy, Cohen kappa, and a Bland-Altman plot, through a 5-fold cross-validation.

RESULTS

The FFNN achieved an accuracy of 0.944 (CI: 0.937, 0.950) and a kappa of 0.943 (CI: 0.935, 0.950). For the same metrics, random forest yielded 0.574 (CI: 0.573, 0.575) and 0.602 (CI: 0.601, 0.603), respectively, and 0.352 (CI: 0.346, 0.358) and 0.414 (CI: 0.408, 0.422) for the logistic regression. The FFNN had a limit of agreement ranging from -2.73 to 2.67, which was better than random forest (-6.72 to 6.83) or logistic regression (-7.60 to 9.20).

CONCLUSION

The FFNN was better at predicting LOS than random forest or logistic regression. Implementing the FFNN model for routine acute care could be useful for improving the quality of patients' care.

Surveying biomedical relation extraction: a critical examination of current datasets and the proposal of a new resource

Natural language processing (NLP) has become an essential technique in various fields, offering a wide range of possibilities for analyzing data and developing diverse NLP tasks. In the biomedical domain, understanding the complex relationships between compounds and proteins is critical, especially in the context of signal transduction and biochemical pathways. Among these relationships, protein-protein interactions (PPIs) are of particular interest, given their potential to trigger a variety of biological reactions. To improve the ability to predict PPI events, we propose the protein event detection dataset (PEDD), which comprises 6823 abstracts, 39 488 sentences and 182 937 gene pairs. Our PEDD dataset has been utilized in the AI CUP Biomedical Paper Analysis competition, where systems are challenged to predict 12 different relation types. In this paper, we review the state-of-the-art relation extraction research and provide an overview of the PEDD's compilation process. Furthermore, we present the results of the PPI extraction competition and evaluate several language models' performances on the PEDD. This paper's outcomes will provide a valuable roadmap for future studies on protein event detection in NLP. By addressing this critical challenge, we hope to enable breakthroughs in drug discovery and enhance our understanding of the molecular mechanisms underlying various diseases.

Health Care AI and Patient Privacy-Dinerstein v Google

This Viewpoint summarizes a recent lawsuit alleging that a hospital violated patients’ privacy by sharing electronic health record (EHR) data with Google for development of medical artificial intelligence (AI) and discusses how the federal court’s decision in the case provides key insights for hospitals planning to share EHR data with for-profit companies developing medical AI.

Artificial intelligence in positive mental health: a narrative review

The paper reviews the entire spectrum of Artificial Intelligence (AI) in mental health and its positive role in mental health. AI has a huge number of promises to offer mental health care and this paper looks at multiple facets of the same. The paper first defines AI and its scope in the area of mental health. It then looks at various facets of AI like machine learning, supervised machine learning and unsupervised machine learning and other facets of AI. The role of AI in various psychiatric disorders like neurodegenerative disorders, intellectual disability and seizures are discussed along with the role of AI in awareness, diagnosis and intervention in mental health disorders. The role of AI in positive emotional regulation and its impact in schizophrenia, autism spectrum disorders and mood disorders is also highlighted. The article also discusses the limitations of AI based approaches and the need for AI based approaches in mental health to be culturally aware, with structured flexible algorithms and an awareness of biases that can arise in AI. The ethical issues that may arise with the use of AI in mental health are also visited.

Multimodal risk prediction with physiological signals, medical images and clinical notes

The broad adoption of electronic health record (EHR) systems brings us a tremendous amount of clinical data and thus provides opportunities to conduct data-based healthcare research to solve various clinical problems in the medical domain. Machine learning and deep learning methods are widely used in the medical informatics and healthcare domain due to their power to mine insights from raw data. When adapting deep learning models for EHR data, it is essential to consider its heterogeneous nature: EHR contains patient records from various sources including medical tests (e.g. blood test, microbiology test), medical imaging, diagnosis, medications, procedures, clinical notes, etc. Those modalities together provide a holistic view of patient health status and complement each other. Therefore, combining data from multiple modalities that are intrinsically different is challenging but intuitively promising in deep learning for EHR. To assess the expectations of multimodal data, we introduce a comprehensive fusion framework designed to integrate temporal variables, medical images, and clinical notes in EHR for enhanced performance in clinical risk prediction. Early, joint, and late fusion strategies are employed to combine data from various modalities effectively. We test the model with three predictive tasks: in-hospital mortality, long length of stay, and 30-day readmission. Experimental results show that multimodal models outperform uni-modal models in the tasks involved. Additionally, by training models with different input modality combinations, we calculate the Shapley value for each modality to quantify their contribution to multimodal performance. It is shown that temporal variables tend to be more helpful than CXR images and clinical notes in the three explored predictive tasks.

Development and validation of 'Patient Optimizer' (POP) algorithms for predicting surgical risk with machine learning

BACKGROUND

Pre-operative risk assessment can help clinicians prepare patients for surgery, reducing the risk of perioperative complications, length of hospital stay, readmission and mortality. Further, it can facilitate collaborative decision-making and operational planning.

OBJECTIVE

To develop effective pre-operative risk assessment algorithms (referred to as Patient Optimizer or POP) using Machine Learning (ML) that predict the development of post-operative complications and provide pilot data to inform the design of a larger prospective study.

METHODS

After institutional ethics approval, we developed a base model that encapsulates the standard manual approach of combining patient-risk and procedure-risk. In an automated process, additional variables were included and tested with 10-fold cross-validation, and the best performing features were selected. The models were evaluated and confidence intervals calculated using bootstrapping. Clinical expertise was used to restrict the cardinality of categorical variables (e.g. pathology results) by including the most clinically relevant values. The models were created with logistic regression (LR) and extreme gradient-boosted trees using XGBoost (Chen and Guestrin, 2016). We evaluated performance using the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). Data was obtained from a metropolitan university teaching hospital from January 2015 to July 2020. Data collection was restricted to adult patients undergoing elective surgery.

RESULTS

A total of 11,475 adult admissions were included. The performance of XGBoost and LR was very similar across endpoints and metrics. For predicting the risk of any post-operative complication, kidney failure and length-of-stay (LOS), POP with XGBoost achieved an AUROC (95%CI) of 0.755 (0.744, 0.767), 0.869 (0.846, 0.891) and 0.841 (0.833, 0.847) respectively and AUPRC of 0.651 (0.632, 0.669), 0.336 (0.282, 0.390) and 0.741 (0.729, 0.753) respectively. For 30-day readmission and in-patient mortality, POP with XGBoost achieved an AUROC (95%CI) of 0.610 (0.587, 0.635) and 0.866 (0.777, 0.943) respectively and AUPRC of 0.116 (0.104, 0.132) and 0.031 (0.015, 0.072) respectively.

CONCLUSION

The POP algorithms effectively predicted any post-operative complication, kidney failure and LOS in the sample population. A larger study is justified to improve the algorithm to better predict complications and length of hospital stay. A larger dataset may also improve the prediction of additional specific complications, readmission and mortality.

Real-time prognostic biomarkers for predicting in-hospital mortality and cardiac complications in COVID-19 patients

Hospitalized patients with Coronavirus disease 2019 (COVID-19) are highly susceptible to in-hospital mortality and cardiac complications such as atrial arrhythmias (AA). However, the utilization of biomarkers such as potassium, B-type natriuretic peptide, albumin, and others for diagnosis or the prediction of in-hospital mortality and cardiac complications has not been well established. The study aims to investigate whether biomarkers can be utilized to predict mortality and cardiac complications among hospitalized COVID-19 patients. Data were collected from 6,927 hospitalized COVID-19 patients from March 1, 2020, to March 31, 2021 at one quaternary (Henry Ford Health) and five community hospital registries (Trinity Health Systems). A multivariable logistic regression prediction model was derived using a random sample of 70% for derivation and 30% for validation. Serum values, demographic variables, and comorbidities were used as input predictors. The primary outcome was in-hospital mortality, and the secondary outcome was onset of AA. The associations between predictor variables and outcomes are presented as odds ratio (OR) with 95% confidence intervals (CIs). Discrimination was assessed using area under ROC curve (AUC). Calibration was assessed using Brier score. The model predicted in-hospital mortality with an AUC of 90% [95% CI: 88%, 92%]. In addition, potassium showed promise as an independent prognostic biomarker that predicted both in-hospital mortality, with an AUC of 71.51% [95% Cl: 69.51%, 73.50%], and AA with AUC of 63.6% [95% Cl: 58.86%, 68.34%]. Within the test cohort, an increase of 1 mEq/L potassium was associated with an in-hospital mortality risk of 1.40 [95% CI: 1.14, 1.73] and a risk of new onset of AA of 1.55 [95% CI: 1.25, 1.93]. This cross-sectional study suggests that biomarkers can be used as prognostic variables for in-hospital mortality and onset of AA among hospitalized COVID-19 patients.

A clinically actionable and explainable real-time risk assessment framework for stroke-associated pneumonia

The current medical practice is more responsive rather than proactive, despite the widely recognized value of early disease detection, including improving the quality of care and reducing medical costs. One of the cornerstones of early disease detection is clinically actionable predictions, where predictions are expected to be accurate, stable, real-time and interpretable. As an example, we used stroke-associated pneumonia (SAP), setting up a transformer-encoder-based model that analyzes highly heterogeneous electronic health records in real-time. The model was proven accurate and stable on an independent test set. In addition, it issued at least one warning for 98.6 % of SAP patients, and on average, its alerts were ahead of physician diagnoses by 2.71 days. We applied Integrated Gradient to glean the model's reasoning process. Supplementing the risk scores, the model highlighted critical historical events on patients' trajectories, which were shown to have high clinical relevance.

Understanding New Machine Learning Architectures: Practical Generative Artificial Intelligence for Anesthesiologists

Recent advances in neural networks have given rise to generative artificial intelligence, systems able to produce fluent responses to natural questions or attractive and even photorealistic images from text prompts. These systems were developed through new network architectures that permit massive computational resources to be applied efficiently to enormous data sets. First, this review examines autoencoder architecture and its derivatives the variational autoencoder and the U-Net in annotating and manipulating images and extracting salience. This architecture will be important for applications like automated x-ray interpretation or real-time highlighting of anatomy in ultrasound images. Second, this article examines the transformer architecture in the interpretation and generation of natural language, as it will be useful in producing automated summarization of medical records or performing initial patient screening. The author also applies the GPT-3.5 algorithm to example questions from the American Board of Anesthesiologists Basic Examination and find that, under surprisingly reasonable conditions, it correctly answers more than half the questions.

A novel deep learning-based method for automatic stereology of microglia cells from low magnification images

Microglial cells mediate diverse homeostatic, inflammatory, and immune processes during normal development and in response to cytotoxic challenges. During these functional activities, microglial cells undergo distinct numerical and morphological changes in different tissue volumes in both rodent and human brains. However, it remains unclear how these cytostructural changes in microglia correlate with region-specific neurochemical functions. To better understand these relationships, neuroscientists need accurate, reproducible, and efficient methods for quantifying microglial cell number and morphologies in histological sections. To address this deficit, we developed a novel deep learning (DL)-based classification, stereology approach that links the appearance of Iba1 immunostained microglial cells at low magnification (20×) with the total number of cells in the same brain region based on unbiased stereology counts as ground truth. Once DL models are trained, total microglial cell numbers in specific regions of interest can be estimated and treatment groups predicted in a high-throughput manner (<1 min) using only low-power images from test cases, without the need for time and labor-intensive stereology counts or morphology ratings in test cases. Results for this DL-based automatic stereology approach on two datasets (total 39 mouse brains) showed >90% accuracy, 100% percent repeatability (Test-Retest) and 60× greater efficiency than manual stereology (<1 min vs. ∼ 60 min) using the same tissue sections. Ongoing and future work includes use of this DL-based approach to establish clear neurodegeneration profiles in age-related human neurological diseases and related animal models.

Temporal self-attention for risk prediction from electronic health records using non-stationary kernel approximation

Effective modeling of patient representation from electronic health records (EHRs) is increasingly becoming a vital research topic. Yet, modeling the non-stationarity in EHR data has received less attention. Most existing studies follow a strong assumption of stationarity in patient representation from EHRs. However, in practice, a patient's visits are irregularly spaced over a relatively long period of time, and disease progression patterns exhibit non-stationarity. Furthermore, the time gaps between patient visits often encapsulate significant domain knowledge, potentially revealing undiscovered patterns that characterize specific medical conditions. To address these challenges, we introduce a new method which combines the self-attention mechanism with non-stationary kernel approximation to capture both contextual information and temporal relationships between patient visits in EHRs. To assess the effectiveness of our proposed approach, we use two real-world EHR datasets, comprising a total of 76,925 patients, for the task of predicting the next diagnosis code for a patient, given their EHR history. The first dataset is a general EHR cohort and consists of 11,451 patients with a total of 3,485 unique diagnosis codes. The second dataset is a disease-specific cohort that includes 65,474 pregnant patients and encompasses a total of 9,782 unique diagnosis codes. Our experimental evaluation involved nine prediction models, categorized into three distinct groups. Group 1 comprises the baselines: original self-attention with positional encoding model, RETAIN model, and LSTM model. Group 2 includes models employing self-attention with stationary kernel approximations, specifically incorporating three variations of Bochner's feature maps. Lastly, Group 3 consists of models utilizing self-attention with non-stationary kernel approximations, including quadratic, cubic, and bi-quadratic polynomials. The experimental results demonstrate that non-stationary kernels significantly outperformed baseline methods for NDCG@10 and Hit@10 metrics in both datasets. The performance boost was more substantial in dataset 1 for the NDCG@10 metric. On the other hand, stationary Kernels showed significant but smaller gains over baselines and were nearly as effective as Non-stationary Kernels for Hit@10 in dataset 2. These findings robustly validate the efficacy of employing non-stationary kernels for temporal modeling of EHR data, and emphasize the importance of modeling non-stationary temporal information in healthcare prediction tasks.

PROPOSE. Development and validation of a prediction model for shared decision making for patients with lumbar spinal stenosis

Background

Decompression for lumbar spinal stenosis (LSS) is the most frequently performed spine surgery in Denmark. According to the Danish spine registry DaneSpine, at 1 year after surgery, about 75% of patients experiences considerable pain relief and around 66% improvement in quality of life. However, 25% do not improve very much. We have developed a predictive decision support tool, PROPOSE. It is intended to be used in the clinical conversation between healthcare providers and LSS patients as a shared decision-making aid presenting pros and cons of surgical intervention. This study presents the development and evaluation of PROPOSE in a clinical setting.

Methods

For model development, 6.357 LSS patients enrolled in DaneSpine were identified. For model validation, predictor response and predicted outcome was collected via PROPOSE from 228 patients. Observed outcome at 1 year was retrieved from DaneSpine. All participants were treated at 3 Danish spine centers. The outcome measures presented are improvement in walking distance, the Oswestry Disability Index, EQ-5D-3L and leg/back pain on the Visual Analog Scale. Outcome variables were dichotomized into success (1) and failure (0). With the exception of walking distance, a success was defined as reaching minimal clinically important difference at 1-year follow-up. Models were trained using Multivariate Adaptive Regression Splines. Performance was assessed by inspecting confusion matrix, ROC curves and comparing GCV (generalized cross-validation) errors. Final performance of the models was evaluated on independent test data.

Results

The walking distance model demonstrated excellent performance with an AUC of 0.88 and a Brier score of 0.14. The VAS leg pain model had the lowest discriminatory performance with an AUC of 0.67 and a Brier score of 0.22.

Conclusions

PROPOSE works in a real-world clinical setting as a proof of concept and demonstrates acceptable performance. It may have the potential of aiding shared decision making.

Exploring the application and future outlook of Artificial intelligence in pancreatic cancer

Pancreatic cancer, an exceptionally malignant tumor of the digestive system, presents a challenge due to its lack of typical early symptoms and highly invasive nature. The majority of pancreatic cancer patients are diagnosed when curative surgical resection is no longer possible, resulting in a poor overall prognosis. In recent years, the rapid progress of Artificial intelligence (AI) in the medical field has led to the extensive utilization of machine learning and deep learning as the prevailing approaches. Various models based on AI technology have been employed in the early screening, diagnosis, treatment, and prognostic prediction of pancreatic cancer patients. Furthermore, the development and application of three-dimensional visualization and augmented reality navigation techniques have also found their way into pancreatic cancer surgery. This article provides a concise summary of the current state of AI technology in pancreatic cancer and offers a promising outlook for its future applications.

The Experimentalist's Guide to Machine Learning for Small Molecule Design

Initially part of the field of artificial intelligence, machine learning (ML) has become a booming research area since branching out into its own field in the 1990s. After three decades of refinement, ML algorithms have accelerated scientific developments across a variety of research topics. The field of small molecule design is no exception, and an increasing number of researchers are applying ML techniques in their pursuit of discovering, generating, and optimizing small molecule compounds. The goal of this review is to provide simple, yet descriptive, explanations of some of the most commonly utilized ML algorithms in the field of small molecule design along with those that are highly applicable to an experimentally focused audience. The algorithms discussed here span across three ML paradigms: supervised learning, unsupervised learning, and ensemble methods. Examples from the published literature will be provided for each algorithm. Some common pitfalls of applying ML to biological and chemical data sets will also be explained, alongside a brief summary of a few more advanced paradigms, including reinforcement learning and semi-supervised learning.