Journal of Biomedical Informatics最新文献

Objective

Survival analysis is widely utilized in healthcare to predict the timing of disease onset. Traditional methods of survival analysis are usually based on Cox Proportional Hazards model and assume proportional risk for all subjects. However, this assumption is rarely true for most diseases, as the underlying factors have complex, non-linear, and time-varying relationships. This concern is especially relevant for pregnancy, where the risk for pregnancy-related complications, such as preeclampsia, varies across gestation. Recently, deep learning survival models have shown promise in addressing the limitations of classical models, as the novel models allow for non-proportional risk handling, capturing nonlinear relationships, and navigating complex temporal dynamics.

Methods

We present a methodology to model the temporal risk of preeclampsia during pregnancy and investigate the associated clinical risk factors. We utilized a retrospective dataset including 66,425 pregnant individuals who delivered in two tertiary care centers from 2015 to 2023. We modeled the preeclampsia risk by modifying DeepHit, a deep survival model, which leverages neural network architecture to capture time-varying relationships between covariates in pregnancy. We applied time series k-means clustering to DeepHit’s normalized output and investigated interpretability using Shapley values.

Results

We demonstrate that DeepHit can effectively handle high-dimensional data and evolving risk hazards over time with performance similar to the Cox Proportional Hazards model, achieving an area under the curve (AUC) of 0.78 for both models. The deep survival model outperformed traditional methodology by identifying time-varied risk trajectories for preeclampsia, providing insights for early and individualized intervention. K-means clustering resulted in patients delineating into low-risk, early-onset, and late-onset preeclampsia groups—notably, each of those has distinct risk factors.

Conclusion

This work demonstrates a novel application of deep survival analysis in time-varying prediction of preeclampsia risk. Our results highlight the advantage of deep survival models compared to Cox Proportional Hazards models in providing personalized risk trajectory and demonstrating the potential of deep survival models to generate interpretable and meaningful clinical applications in medicine.

Objective

The ability to apply results from a study to a broader population remains a primary objective in translational science. Distinct from intrinsic elements of scientific rigor, the extrinsic concept of generalization requires there be alignment between a study cohort and population in which results are expected to be applied. Widespread efforts have been made to quantify representativeness of study cohorts. These techniques, however, often consider the study and target cohorts as monolithic collections that can be directly compared. Overlooking known impacts to health from socio-demographic and environmental factors tied to individual’s geographical location, and potentially obfuscating misalignment in underrepresented population subgroups. This manuscript introduces several measures to account for geographic information in the assessment of cohort representation.

Methods

Metrics were defined across two themes: First, measures of recruitment, to assess if a study cohort is drawn at an expected rate and in an expected geographical pattern with respect to individuals in a reference cohort. Second, measures of individual characteristics, to assess if the individuals in the study cohort accurately reflect the sociodemographic, clinical, and geographic diversity observed across a reference cohort while accounting for the geospatial proximity of individuals.

Results

As an empirical demonstration, methods are applied to an active clinical study examining asthma in Black/African American patients at a US Midwestern pediatric hospital. Results illustrate how areas of over- and under-recruitment can be identified and contextualized in light of study recruitment patterns at an individual-level, highlighting the ability to identify a subset of features for which the study cohort closely resembled the broader population. In addition they provide an opportunity to dive deeper into misalignments, to identify study cohort members that are in some way distinct from the communities for which they are expected to represent.

Conclusion

Together, these metrics provide a comprehensive spatial assessment of a study cohort with respect to a broader target population. Such an approach offers researchers a toolset by which to target expected generalization of results derived from a given study.

Background

The increasing aging population presents a significant challenge, accompanied by a shortage of professional caregivers, adding to the therapeutic burden. Clinical decision support systems, utilizing computerized clinical guidelines, can improve healthcare quality, reduce expenses, save time, and boost caregiver efficiency.

Objectives

1) Develop and evaluate an automated quality assessment (QA) system for retrospective longitudinal care quality analysis, focusing on clinical staff adherence to evidence-based guidelines (GLs). 2) Assess the system’s technical feasibility and functional capability for senior nurse use in geriatric pressure-ulcer management.

Methods

A computational QA system using our Quality Assessment Temporal Patterns (QATP) methodology was designed and implemented. Our methodology transforms the GL’s procedural-knowledge into declarative-knowledge temporal-abstraction patterns representing the expected execution trace in the patient’s data for correct therapy application. Fuzzy temporal logic allows for partial compliance, reflecting individual and grouped action performance considering their values and temporal aspects. The system was tested using a pressure ulcer treatment GL and data from 100 geriatric patients’ Electronic Medical Records (EMR). After technical evaluation for accuracy and feasibility, an extensive functional evaluation was conducted by an experienced nurse, comparing QA scores with and without system support, and versus automated system scores. Time efficiency was also measured.

Results

QA scores from the geriatric nurse, with and without system’s support, did not significantly differ from those provided by the automated system (p < 0.05), demonstrating the effectiveness and reliability of both manual and automated methods. The system-supported manual QA process reduced scoring time by approximately two-thirds, from an average of 17.3 min per patient manually to about 5.9 min with the system’s assistance, highlighting the system’s efficiency potential in clinical practice.

Conclusion

The QA system based on QATP, produces scores consistent with an experienced nurse’s assessment for complex care over extended periods. It enables quick and accurate quality care evaluation for multiple patients after brief training. Such automated QA systems may empower nursing staff, enabling them to manage more patients, accurately and consistently, while reducing costs due to saved time and effort, and enhanced compliance with evidence-based guidelines.

The multimorbidity problem involves the identification and mitigation of adverse interactions that occur when multiple computer interpretable guidelines are applied concurrently to develop a treatment plan for a patient diagnosed with multiple diseases. Solving this problem requires decision support approaches which are difficult to comprehend for physicians. As such, the rationale for treatment plans generated by these approaches needs to be provided.

Objective:

To develop an explainability component for an automated planning-based approach to the multimorbidity problem, and to assess the fidelity and interpretability of generated explanations using a clinical case study.

Methods:

The explainability component leverages the task-network model for representing computer interpretable guidelines. It generates post-hoc explanations composed of three aspects that answer why specific clinical actions are in a treatment plan, why specific revisions were applied, and how factors like medication cost, patient’s adherence, etc. influence the selection of specific actions. The explainability component is implemented as part of MitPlan, where we revised our planning-based approach to support explainability. We developed an evaluation instrument based on the system causability scale and other vetted surveys to evaluate the fidelity and interpretability of its explanations using a two dimensional comparison study design.

Results:

The explainability component was implemented for MitPlan and tested in the context of a clinical case study. The fidelity and interpretability of the generated explanations were assessed using a physician-focused evaluation study involving 21 participants from two different specialties and two levels of experience. Results show that explanations provided by the explainability component in MitPlan are of acceptable fidelity and interpretability, and that the clinical justification of the actions in a treatment plan is important to physicians.

Conclusion:

We created an explainability component that enriches an automated planning-based approach to solving the multimorbidity problem with meaningful explanations for actions in a treatment plan. This component relies on the task-network model to represent computer interpretable guidelines and as such can be ported to other approaches that also use the task-network model representation. Our evaluation study demonstrated that explanations that support a physician’s understanding of the clinical reasons for the actions in a treatment plan are useful and important.

Objectives:

This study aims to enhance the analysis of healthcare processes by introducing Object-Centric Process Mining (OCPM). By offering a holistic perspective that accounts for the interactions among various objects, OCPM transcends the constraints of conventional patient-centric process mining approaches, ensuring a more detailed and inclusive understanding of healthcare dynamics.

Methods:

We develop a novel method to transform the Observational Medical Outcomes Partnership Common Data Models (OMOP CDM) into Object-Centric Event Logs (OCELs). First, an OMOP CDM4PM is created from the standard OMOP CDM, focusing on data relevant to generating OCEL and addressing healthcare data’s heterogeneity and standardization challenges. Second, this subset is transformed into OCEL based on specified healthcare criteria, including identifying various object types, clinical activities, and their relationships. The methodology is tested on the MIMIC-IV database to evaluate its effectiveness and utility.

Results:

Our proposed method effectively produces OCELs when applied to the MIMIC-IV dataset, allowing for the implementation of OCPM in the healthcare industry. We rigorously evaluate the comprehensiveness and level of abstraction to validate our approach’s effectiveness. Additionally, we create diverse object-centric process models intricately designed to navigate the complexities inherent in healthcare processes.

Conclusion:

Our approach introduces a novel perspective by integrating multiple viewpoints simultaneously. To the best of our knowledge, this is the inaugural application of OCPM within the healthcare sector, marking a significant advancement in the field.

Objective:

Linear and logistic regression are widely used statistical techniques in population genetics for analyzing genetic data and uncovering patterns and associations in large genetic datasets, such as identifying genetic variations linked to specific diseases or traits. However, obtaining statistically significant results from these studies requires large amounts of sensitive genotype and phenotype information from thousands of patients, which raises privacy concerns. Although cryptographic techniques such as homomorphic encryption offers a potential solution to the privacy concerns as it allows computations on encrypted data, previous methods leveraging homomorphic encryption have not addressed the confidentiality of shared models, which can leak information about the training data.

Methods:

In this work, we present a secure model evaluation method for linear and logistic regression using homomorphic encryption for six prediction tasks, where input genotypes, output phenotypes, and model parameters are all encrypted.

Results:

Our method ensures no private information leakage during inference and achieves high accuracy ($\ge 93\text{\%}$ for all outcomes) with each inference taking less than ten seconds for $\sim 200$ genomes.

Conclusion:

Our study demonstrates that it is possible to perform linear and logistic regression model evaluation while protecting patient confidentiality with theoretical security guarantees. Our implementation and test data are available at https://github.com/G2Lab/privateML/.

Objective

Comprehensive analysis of histopathology images and transcriptomics data enables the identification of candidate biomarkers and multimodal association patterns. Most existing multimodal data association studies are derived from extensions of the joint nonnegative matrix factorization model for identifying complex data associations, which can make full use of clinical prior information. However, the raw data were usually taken as the input without considering the underlying complex multi-subspace structure, influencing the subsequent integration analysis results.

Methods

This study proposed a deep-self reconstructed joint nonnegative matrix factorization (DSRJNMF) model to use self-expressive properties to reconstruct the raw data to characterize the similarity structure associated with clinical labels. Then, the sparsity, orthogonality, and regularization constraints constructed from prior information are added to the DSRJNMF model to determine the sparse set of biologically relevant features across modalities.

Results

The algorithm has been applied to identify the imaging genetic association of triple negative breast cancer (TNBC). Multilevel experimental results demonstrate that the proposed algorithm better estimates potential associations between pathological image features and miRNA-gene and identifies consistent multimodal imaging genetic biomarkers to guide the interpretation of TNBC.

Conclusion

The propose method provides a novel idea of data association analysis oriented to complex diseases.

Objective

Despite increased availability of methodologies to identify algorithmic bias, the operationalization of bias evaluation for healthcare predictive models is still limited. Therefore, this study proposes a process for bias evaluation through an empirical assessment of common hospital readmission models. The process includes selecting bias measures, interpretation, determining disparity impact and potential mitigations.

Methods

This retrospective analysis evaluated racial bias of four common models predicting 30-day unplanned readmission (i.e., LACE Index, HOSPITAL Score, and the CMS readmission measure applied as is and retrained). The models were assessed using 2.4 million adult inpatient discharges in Maryland from 2016 to 2019. Fairness metrics that are model-agnostic, easy to compute, and interpretable were implemented and apprised to select the most appropriate bias measures. The impact of changing model’s risk thresholds on these measures was further assessed to guide the selection of optimal thresholds to control and mitigate bias.

Results

Four bias measures were selected for the predictive task: zero-one-loss difference, false negative rate (FNR) parity, false positive rate (FPR) parity, and generalized entropy index. Based on these measures, the HOSPITAL score and the retrained CMS measure demonstrated the lowest racial bias. White patients showed a higher FNR while Black patients resulted in a higher FPR and zero-one-loss. As the models’ risk threshold changed, trade-offs between models’ fairness and overall performance were observed, and the assessment showed all models’ default thresholds were reasonable for balancing accuracy and bias.

Conclusions

This study proposes an Applied Framework to Assess Fairness of Predictive Models (AFAFPM) and demonstrates the process using 30-day hospital readmission model as the example. It suggests the feasibility of applying algorithmic bias assessment to determine optimized risk thresholds so that predictive models can be used more equitably and accurately. It is evident that a combination of qualitative and quantitative methods and a multidisciplinary team are necessary to identify, understand and respond to algorithm bias in real-world healthcare settings. Users should also apply multiple bias measures to ensure a more comprehensive, tailored, and balanced view. The results of bias measures, however, must be interpreted with caution and consider the larger operational, clinical, and policy context.

Objective

Failure to receive prompt blood transfusion leads to severe complications if massive bleeding occurs during surgery. For the timely preparation of blood products, predicting the possibility of massive transfusion (MT) is essential to decrease morbidity and mortality. This study aimed to develop a model for predicting MT 10 min in advance using non-invasive bio-signal waveforms that change in real-time.

Methods

In this retrospective study, we developed a deep learning-based algorithm (DLA) to predict intraoperative MT within 10 min. MT was defined as the transfusion of 3 or more units of red blood cells within an hour. The datasets consisted of 18,135 patients who underwent surgery at Seoul National University Hospital (SNUH) for model development and internal validation and 621 patients who underwent surgery at the Boramae Medical Center (BMC) for external validation. We constructed the DLA by using features extracted from plethysmography (collected at 500 Hz) and hematocrit measured during surgery.

Results

Among 18,135 patients in SNUH and 621 patients in BMC, 265 patients (1.46%) and 14 patients (2.25%) received MT during surgery, respectively. The area under the receiver operating characteristic curve (AUROC) of DLA predicting intraoperative MT before 10 min was 0.962 (95% confidence interval [CI], 0.948–0.974) in internal validation and 0.922 (95% CI, 0.882–0.959) in external validation, respectively.

Conclusion

The DLA can successfully predict intraoperative MT using non-invasive bio-signal waveforms.

Background

Studies confirm that significant biases exist in online recommendation platforms, exacerbating pre-existing disparities and leading to less-than-optimal outcomes for underrepresented demographics. We study issues of bias in inclusion and representativeness in the context of healthcare information disseminated via videos on the YouTube social media platform, a widely used online channel for multi-media rich information. With one in three US adults using the Internet to learn about a health concern, it is critical to assess inclusivity and representativeness regarding how health information is disseminated by digital platforms such as YouTube.

Methods

Leveraging methods from fair machine learning (ML), natural language processing and voice and facial recognition methods, we examine inclusivity and representativeness of video content presenters using a large corpus of videos and their metadata on a chronic condition (diabetes) extracted from the YouTube platform. Regression models are used to determine whether presenter demographics impact video popularity, measured by the video’s average daily view count. A video that generates a higher view count is considered to be more popular.

Results

The voice and facial recognition methods predicted the gender and race of the presenter with reasonable success. Gender is predicted through voice recognition (accuracy = 78%, AUC = 76%), while the gender and race predictions use facial recognition (accuracy = 93%, AUC = 92% and accuracy = 82%, AUC = 80%, respectively). The gender of the presenter is more significant for video views only when the face of the presenter is not visible while videos with male presenters with no face visibility have a positive relationship with view counts. Furthermore, videos with white and male presenters have a positive influence on view counts while videos with female and non − white group have high view counts.

Conclusion

Presenters’ demographics do have an influence on average daily view count of videos viewed on social media platforms as shown by advanced voice and facial recognition algorithms used for assessing inclusion and representativeness of the video content. Future research can explore short videos and those at the channel level because popularity of the channel name and the number of videos associated with that channel do have an influence on view counts.