ABSTRACT
Background: Machine learning has shown to be an effective method for early prediction and intervention of Gestational diabetes mellitus (GDM), which greatly decreases GDM incidence, reduces maternal and infant complications and improves the prognosis. However, there is still much room for improvement in data quality, feature dimension, and accuracy. The contributions and mechanism explanations of clinical data at different pregnancy stages to the prediction accuracy are still lacking. More importantly, current models still face notable obstacles in practical applications due to the complex and diverse input features and difficulties in redeployment. As a result, a simple, practical but accurate enough model is urgently needed. Design and methods: In this study, 2309 samples from two public hospitals in Shenzhen, China were collected for analysis. Different algorithms were systematically compared to build a robust and stepwise prediction system (level A to C) based on advanced machine learning, and models under different levels were interpreted. Results: XGBoost reported the best performance with ACC of 0.922, 0.859 and 0.850, AUC of 0.974, 0.924 and 0.913 for the selected level A to C models in the test set, respectively. Tree-based feature importance and SHAP method successfully identified the commonly recognized risk factors, while indicated new inconsistent impact trends for GDM in different stages of pregnancy. Conclusion: A stepwise prediction system was successfully established. A practical tool that enables a quick prediction of GDM was released at https://github.com/ifyoungnet/MedGDM.This study is expected to provide a more detailed profiling of GDM risk and lay the foundation for the application of the model in practice.
ABSTRACT
Cross-disease genome-wide association studies (GWASs) unveil pleiotropic loci, mostly situated within the non-coding genome, each of which exerts pleiotropic effects across multiple diseases. However, the challenge "W-H-W" (namely, whether, how, and in which specific diseases pleiotropy can inform clinical therapeutics) calls for effective and integrative approaches and tools. We here introduce a pleiotropy-driven approach specifically designed for therapeutic target prioritization and evaluation from cross-disease GWAS summary data, with its validity demonstrated through applications to two systems of disorders (neuropsychiatric and inflammatory). We illustrate its improved performance in recovering clinical proof-of-concept therapeutic targets. Importantly, it identifies specific diseases where pleiotropy informs clinical therapeutics. Furthermore, we illustrate its versatility in accomplishing advanced tasks, including pathway crosstalk identification and downstream crosstalk-based analyses. To conclude, our integrated solution helps bridge the gap between pleiotropy studies and therapeutics discovery.
Subject(s)
Genetic Pleiotropy , Genome-Wide Association Study , Humans , Genetic Predisposition to Disease , Polymorphism, Single NucleotideABSTRACT
OBJECTIVE: Root causes of disease intuitively correspond to root vertices of a causal model that increase the likelihood of a diagnosis. This description of a root cause nevertheless lacks the rigorous mathematical formulation needed for the development of computer algorithms designed to automatically detect root causes from data. We seek a definition of patient-specific root causes of disease that models the intuitive procedure routinely utilized by physicians to uncover root causes in the clinic. METHODS: We use structural equation models, interventional counterfactuals and the recently developed mathematical formalization of backtracking counterfactuals to propose a counterfactual formulation of patient-specific root causes of disease matching clinical intuition. RESULTS: We introduce a definition of patient-specific root causes of disease that climbs to the third rung of Pearl's Ladder of Causation and matches clinical intuition given factual patient data and a working causal model. We then show how to assign a root causal contribution score to each variable using Shapley values from explainable artificial intelligence. CONCLUSION: The proposed counterfactual formulation of patient-specific root causes of disease accounts for noisy labels, adapts to disease prevalence and admits fast computation without the need for counterfactual simulation.
Subject(s)
Artificial Intelligence , Models, Theoretical , Humans , Computer SimulationABSTRACT
With the growing amount of COVID-19 cases, especially in developing countries with limited medical resources, it is essential to accurately and efficiently diagnose COVID-19. Due to characteristic ground-glass opacities (GGOs) and other types of lesions being present in both COVID-19 and other acute lung diseases, misdiagnosis occurs often - 26.6% of the time in manual interpretations of CT scans. Current deep-learning models can identify COVID-19 but cannot distinguish it from other common lung diseases like bacterial pneumonia. Concretely, COVision is a deep-learning model that can differentiate COVID-19 from other common lung diseases, with high specificity using CT scans and other clinical factors. COVision was designed to minimize overfitting and complexity by decreasing the number of hidden layers and trainable parameters while still achieving superior performance. Our model consists of two parts: the CNN which analyzes CT scans and the CFNN (clinical factors neural network) which analyzes clinical factors such as age, gender, etc. Using federated averaging, we ensembled our CNN with the CFNN to create a comprehensive diagnostic tool. After training, our CNN achieved an accuracy of 95.8% and our CFNN achieved an accuracy of 88.75% on a validation set. We found a statistical significance that COVision performs better than three independent radiologists with at least 10 years of experience, especially in differentiating COVID-19 from pneumonia. We analyzed our CNN's activation maps through Grad-CAMs and found that lesions in COVID-19 presented peripherally, closer to the pleura, whereas pneumonia lesions presented centrally.
Subject(s)
COVID-19 , Pneumonia , Humans , COVID-19/diagnostic imaging , Neural Networks, Computer , Lung/diagnostic imaging , Tomography, X-Ray Computed , Pneumonia/diagnosisABSTRACT
Objective: Healthcare systems require transformation to meet societal challenges and projected health demands. Digital and computational tools and approaches are fundamental to this transformation, and hospitals have a key role to play in their development and implementation. This paper reports on a study with the objective of exploring the challenges encountered by hospital leaders and innovators as they implement a strategy to become a data-driven hospital organisation. In doing so, this paper provides guidance to future leaders and innovators seeking to build computational and digital capabilities in complex clinical settings. Methods: Interviews were undertaken with 42 participants associated with a large public hospital organisation within England's National Health Service. Using the concept of institutional readiness as an analytical framework, the paper explores participants' perspectives on the organisation's capacity to support the development of, and benefit from, digital and computational approaches. Results: Participants' accounts reveal a range of specific institutional readiness criteria relating to organisational vision, technical capability, organisational agility, and talent and skills that, when met, enhance the organisations' capacity to support the development and implementation of digital and computational tools. Participant accounts also reveal challenges relating to these criteria, such as unrealistic expectations and the necessary prioritisation of clinical work in resource-constrained settings. Conclusions: The paper identifies a general set of institutional readiness criteria that can guide future hospital leaders and innovators aiming to improve their organisation's digital and computational capability. The paper also illustrates the challenges of pursuing digital and computational innovation in resource-constrained hospital environments.
ABSTRACT
Asthma is a chronic disease that is caused by a combination of genetic risks and environmental triggers and can affect both adults and children. Genome-wide association studies have revealed partly distinct genetic architectures for its two age-of-onset subtypes (namely, adult-onset and childhood-onset). We reason that identifying shared and distinct drug targets between these subtypes may inform the development of subtype-specific therapeutic strategies. In attempting this, we here introduce Priority Index for Asthma or PIA, a genetics-led and network-driven drug target prioritisation tool for asthma. We demonstrate the validity of the tool in improving drug target prioritisation for asthma compared to the status quo methods, as well as in capturing the underlying etiology and existing therapeutics for the disease. We also illustrate how PIA can be used to prioritise drug targets for adult- and childhood-onset asthma, as well as to identify shared and distinct pathway crosstalk genes. Shared crosstalk genes are mostly involved in JAK-STAT signaling, with clinical evidence supporting that targeting this pathway may be a promising drug repurposing opportunity for both subtypes. Crosstalk genes specific to childhood-onset asthma are enriched for PI3K-AKT-mTOR signaling, and we identify genes that are already targeted by licensed medications as repurposed drug candidates for this subtype. We make all our results accessible and reproducible at http://www.genetictargets.com/PIA. Collectively, our study has significant implications for asthma computational medicine research and can guide the future development of subtype-specific therapeutic strategies for the disease.
Subject(s)
Asthma , Genome-Wide Association Study , Humans , Child , Adult , Phosphatidylinositol 3-Kinases/genetics , Asthma/drug therapy , Asthma/genetics , Risk Factors , Polymorphism, Single NucleotideABSTRACT
BACKGROUND: Cardiac arrest is a leading cause of mortality prior to discharge for children admitted to the pediatric intensive care unit. To address this problem, we used machine learning to predict cardiac arrest up to three hours in advance. METHODS: Our data consists of 240 Hz ECG waveform data, 0.5 Hz physiological time series data, medications, and demographics from 1,145 patients in the pediatric intensive care unit at the Johns Hopkins Hospital, 15 of whom experienced a cardiac arrest. The data were divided into training, validating, and testing sets, and features were generated every five minutes. 23 heart rate variability (HRV) metrics were determined from ECG waveforms. 96 summary statistics were calculated for 12 vital signs, such as respiratory rate and blood pressure. Medications were classified into 42 therapeutic drug classes. Binary features were generated to indicate the administration of these different drugs. Next, six machine learning models were evaluated: logistic regression, support vector machine, random forest, XGBoost, LightGBM, and a soft voting ensemble. RESULTS: XGBoost performed the best, with 0.971 auROC, 0.797 auPRC, 99.5% sensitivity, and 69.6% specificity on an independent test set. CONCLUSION: We have created high-performing models that identify signatures of in-hospital cardiac arrest (IHCA) that may not be evident to clinicians. These signatures include a combination of heart rate variability metrics, vital signs data, and therapeutic drug classes. These machine learning models can predict IHCA up to three hours prior to onset with high performance, allowing clinicians to intervene earlier, improving patient outcomes.
Subject(s)
Heart Arrest , Child , Humans , Pilot Projects , Intensive Care Units, Pediatric , Vital Signs , Machine Learning , Intensive Care UnitsABSTRACT
Alzheimer's disease (AD) is a common cognitive disorder. Recently, many computer-aided diagnostic techniques have been used for AD prediction utilizing deep learning technology, among which graph neural networks have received increasing attention owing to their ability to model sample relationships on large population graphs. Most of the existing graph-based methods predict diseases according to a single model, which makes it difficult to select an appropriate node embedding algorithm for a certain classification task. Moreover, integrating data from different patterns into a unified model to improve the quality of disease diagnosis remains a challenge. Hence, in this study, we aimed to develop a multi-model fusion framework for AD prediction. A spectral graph attention model was used to aggregate intra- and inter-cluster node embeddings of normal and diseased populations, whereafter, a bilinear aggregation model was applied as an auxiliary model to enhance the abnormality degree in different categories of populations, and finally, an adaptive fusion module was designed to dynamically fuse the results of both models and enhance AD prediction. Compared to other comparison methods, the model proposed in this study provides the best results.
Subject(s)
Alzheimer Disease , Cognitive Dysfunction , Humans , Alzheimer Disease/diagnosis , Algorithms , Neural Networks, ComputerABSTRACT
This article uses the concept of sociotechnical imaginaries to explore how public hospitals are being reimagined and reconfigured by promissory digital health. Drawing on interviews with 42 senior leaders and staff from a large NHS hospital organisation, the article describes the imaginary of a data-driven hospital and the tensions of its operationalisation. These relate to data quality, data curation and data access, and reflect a discord between the organisation's commitment to immediate patient care and its research aspirations. These tensions, however, serve to invigorate, rather than undermine, the sociotechnical imaginary of a data-driven hospital, as they prompt the translation of a general data-driven imaginary into specific sociotechnical arrangements. The article argues that the potency of the data-driven hospital imaginary must be understood in terms of its enchanting qualities: it has the capacity to excite hospital staff and to align distinct and potentially diverging hopes and expectations regarding the societal role of public hospitals. The article concludes by suggesting that the entrenchment of the data-driven imaginary can be partly explained by its strategic utility for severely resource-constrained healthcare organisations: it provides a means for organisations to position themselves towards a viable future in an otherwise dire health-care context.
Subject(s)
Delivery of Health Care , Hospitals , HumansABSTRACT
Genes functionally associated with SARS-CoV-2 infection and genes functionally related to the COVID-19 disease can be different, whose distinction will become the first essential step for successfully fighting against the COVID-19 pandemic. Unfortunately, this first step has not been completed in all biological and medical research. Using a newly developed max-competing logistic classifier, two genes, ATP6V1B2 and IFI27, stand out to be critical in the transcriptional response to SARS-CoV-2 infection with differential expressions derived from NP/OP swab PCR. This finding is evidenced by combining these two genes with another gene in predicting disease status to achieve better-indicating accuracy than existing classifiers with the same number of genes. In addition, combining these two genes with three other genes to form a five-gene classifier outperforms existing classifiers with ten or more genes. These two genes can be critical in fighting against the COVID-19 pandemic as a new focus and direction with their exceptional predicting accuracy. Comparing the functional effects of these genes with a five-gene classifier with 100% accuracy identified and tested from blood samples in our earlier work, the genes and their transcriptional response and functional effects on SARS-CoV-2 infection, and the genes and their functional signature patterns on COVID-19 antibodies, are significantly different. We will use a total of fourteen cohort studies (including breakthrough infections and omicron variants) with 1481 samples to justify our results. Such significant findings can help explore the causal and pathological links between SARS-CoV-2 infection and the COVID-19 disease, and fight against the disease with more targeted genes, vaccines, antiviral drugs, and therapies.
ABSTRACT
Machine learning models are increasingly adopted for facilitating clinical decision-making. However, recent research has shown that machine learning techniques may result in potential biases when making decisions for people in different subgroups, which can lead to detrimental effects on the health and well-being of specific demographic groups such as vulnerable ethnic minorities. This problem, termed algorithmic bias, has been extensively studied in theoretical machine learning recently. However, the impact of algorithmic bias on medicine and methods to mitigate this bias remain topics of active discussion. This paper presents a comprehensive review of algorithmic fairness in the context of computational medicine, which aims at improving medicine with computational approaches. Specifically, we overview the different types of algorithmic bias, fairness quantification metrics, and bias mitigation methods, and summarize popular software libraries and tools for bias evaluation and mitigation, with the goal of providing reference and insights to researchers and practitioners in computational medicine.
Subject(s)
Clinical Decision-Making , Machine Learning , Bias , Decision Making , HumansABSTRACT
Cancers are complex dynamic ecosystems. Reductionist approaches to science are inadequate in characterizing their self-organized patterns and collective emergent behaviors. Since current approaches to single-cell analysis in cancer systems rely primarily on single time-point multiomics, many of the temporal features and causal adaptive behaviors in cancer dynamics are vastly ignored. As such, tools and concepts from the interdisciplinary paradigm of complex systems theory are introduced herein to decode the cellular cybernetics of cancer differentiation dynamics and behavioral patterns. An intuition for the attractors and complex networks underlying cancer processes such as cell fate decision-making, multiscale pattern formation systems, and epigenetic state-transitions is developed. The applications of complex systems physics in paving targeted therapies and causal pattern discovery in precision oncology are discussed. Pediatric high-grade gliomas are discussed as a model-system to demonstrate that cancers are complex adaptive systems, in which the emergence and selection of heterogeneous cellular states and phenotypic plasticity are driven by complex multiscale network dynamics. In specific, pediatric glioblastoma (GBM) is used as a proof-of-concept model to illustrate the applications of the complex systems framework in understanding GBM cell fate decisions and decoding their adaptive cellular dynamics. The scope of these tools in forecasting cancer cell fate dynamics in the emerging field of computational oncology and patient-centered systems medicine is highlighted.
Subject(s)
Cybernetics , Glioblastoma , Child , Ecosystem , Humans , Models, Biological , Precision MedicineABSTRACT
Computational medicine is an emerging discipline that uses computer models and complex software to simulate the development and treatment of diseases. Advances in computer hardware and software technology, especially the development of algorithms and graphics processing units (GPUs), have led to the broader application of computers in the medical field. Computer vision based on mathematical biological modelling will revolutionize clinical research and diagnosis, and promote the innovative development of Chinese medicine, some biological models have begun to play a practical role in various types of research. This paper introduces the concepts and characteristics of computational medicine and then reviews the developmental history of the field, including Digital Human in Chinese medicine. Additionally, this study introduces research progress in computational medicine around the world, lists some specific clinical applications of computational medicine, discusses the key problems and limitations of the research and the development and application of computational medicine, and ultimately looks forward to the developmental prospects, especially in the field of computational Chinese medicine.
Subject(s)
Algorithms , Computer Simulation , HumansABSTRACT
Computational medicine is an emerging discipline that uses computer models and complex software to simulate the development and treatment of diseases. Advances in computer hardware and software technology, especially the development of algorithms and graphics processing units (GPUs), have led to the broader application of computers in the medical field. Computer vision based on mathematical biological modelling will revolutionize clinical research and diagnosis, and promote the innovative development of Chinese medicine, some biological models have begun to play a practical role in various types of research. This paper introduces the concepts and characteristics of computational medicine and then reviews the developmental history of the field, including Digital Human in Chinese medicine. Additionally, this study introduces research progress in computational medicine around the world, lists some specific clinical applications of computational medicine, discusses the key problems and limitations of the research and the development and application of computational medicine, and ultimately looks forward to the developmental prospects, especially in the field of computational Chinese medicine.
Subject(s)
Humans , Algorithms , Computer SimulationABSTRACT
Artificial intelligence (AI) has been developed through repeated new discoveries since around 1960. The use of AI is now becoming widespread within society and our daily lives. AI is also being introduced into healthcare, such as medicine and drug development; however, it is currently biased towards specific domains. The present review traces the history of the development of various AI-based applications in healthcare and compares AI-based healthcare with conventional healthcare to show the future prospects for this type of care. Knowledge of the past and present development of AI-based applications would be useful for the future utilization of novel AI approaches in healthcare.
ABSTRACT
Digital health can drive patient-centric innovation in neuromodulation by leveraging current tools to identify response predictors and digital biomarkers. Iterative technological evolution has led us to an ideal point to integrate digital health with neuromodulation. Here, we provide an overview of the digital health building-blocks, the status of advanced neuromodulation technologies, and future applications for neuromodulation with digital health integration.
ABSTRACT
Computational fluid dynamics (CFD) is a tool that has been used by engineers for over 50 years to analyse heat transfer and fluid flow phenomena. In recent years, there have been rapid developments in biomedical and health research applications of CFD. It has been used to evaluate drug delivery systems, analyse physiological flows (e.g. laryngeal jet flow), facilitate surgical planning (e.g. management of intracranial aneurysms), and develop medical devices (e.g. vascular stents and valve prostheses). Due to the complexity of these fluid flows, it demands an interdisciplinary approach consisting of engineers, computer scientists, and mathematicians to develop the computer programs and software used to solve the mathematical equations. Advances in technology and decreases in computational cost are allowing CFD to be more widely accessible and therefore used in more varied contexts. Cardiovascular medicine is the most common area of biomedical research in which CFD is currently being used, followed closely by upper and lower respiratory tract medicine. CFD is also being used in research investigating cerebrospinal fluid, synovial joints, and intracellular fluid. Although CFD can provide meaningful and aesthetically pleasing outputs, interpretation of the data can be challenging for those without a strong understanding of mathematical and engineering principles. Future development and evolution of computational medicine will therefore require close collaboration between experts in engineering, computer science, and biomedical research. This chapter aims to introduce computational fluid dynamics and present the reader with the basics of biological fluid properties, the CFD method, and its applications within biomedical research through published examples, in hope of bridging knowledge gaps in this rapidly emerging method of biomedical analysis.