The electrocardiogram (ECG) is one of the most common diagnostic tools available to assess cardio-vascular health. The advent of advanced computational techniques such as deep learning has ...dramatically expanded the breadth of clinical problems that can be addressed using ECG data, leading to increasing popularity of ECG deep-learning models aimed at predicting clinical endpoints.
The purpose of this study was to define the current landscape of clinically relevant ECG deep-learning models and examine practices in the scientific reporting of these studies.
We performed a systematic review of PubMed and EMBASE databases to identify clinically relevant ECG deep-learning models published through July 1, 2022.
We identified 44 manuscripts including 53 unique, clinically relevant ECG deep-learning models. The rate of publication of ECG deep-learning models is increasing rapidly. The most common clinical applications of ECG deep learning were identification of cardiomyopathy (14/53 26%), followed by arrhythmia detection (9/53 17%). Methodologic reporting varied; while 33/44 (75%) publications included model architecture diagrams, complete information required to reproduce these models was provided in only 10/44 (23%). Saliency analysis was performed in 20/44 (46%) of publications. Only 18/53 (34%) models were tested within external validation cohorts. Model code or resources allowing for model implementation by external groups were available for only 5/44 (11%) publications.
While ECG deep-learning models are increasingly clinically relevant, their reporting is highly variable, and few publications provide sufficient detail for methodologic reproduction or model validation by external groups. The field of ECG deep learning would benefit from adherence to a set of standardized scientific reporting guidelines.
To conduct a systematic review to assess drug exposure handling in diabetic retinopathy (DR) risk prediction models, a network-meta-analysis to identify drugs associated with DR and a meta-analysis ...to determine which drugs contributed to enhanced model performance.
Systematic review and meta-analysis.
We included studies presenting DR models incorporating drug exposure as a predictor. We searched EMBASE, MEDLINE and SCOPUS from inception to December 2023. We evaluated the quality of studies using the Prediction model Risk of Bias Assessment Tool and certainty using GRADE. We conducted network meta-analysis and meta-analysis to estimate the odds ratio (OR) and pooled C-statistic, respectively, and 95% confidence intervals (CI) (PROSPERO: CRD42022349764).
Of 5,653 records identified, we included 28 studies of 678,837 type 1 or 2 diabetes participants, of which 38,579 (5.7%) had DR. A total of 19, 3 and 7 studies were at high, unclear, and low risk of bias, respectively. Drugs included in models as predictors were: insulin (n=24), antihypertensives (n=5), oral antidiabetics (n=12), lipid-lowering drugs (n=7), antiplatelets (n=2). Drug exposure was modelled primarily as a categorical variable (n=23 studies). Two studies handled drug exposure as time-varying covariates, and one as a time-dependent covariate. Insulin was associated with an increased risk of DR (OR= 2.50; 95%-CI: 1.61-3.86). Models that included insulin (n=9) had a higher pooled C-statistic (C-statistic=0.84, CI: 0.80-0.88), compared to models (n=9) that incorporated a combination of drugs alongside insulin (C-statistic= 0.79, CI:0.74-0.84), as well as models (n=3) not including insulin (C-statistic =0.70, CI: 0.64-0.75). Limitations include the high risk of bias and significant heterogeneity in reviewed studies.
This is the first review assessing drug exposure handling in DR prediction models. Drug exposure was primarily modelled as a categorical variable, with insulin associated with improved model performance. However, due to suboptimal drug handling, associations between other drugs and model performance may have been overlooked. This review proposes the following for future DR prediction models: 1) evaluation of drug exposure as a variable, 2) use of time-varying methodologies, and 3) consideration of drug regimen details. Improving drug exposure handling could potentially unveil novel variables capable of significantly enhancing the predictive capability of prediction models.
Risk prediction tools for colorectal cancer (CRC) have potential to improve the efficiency of population-based screening by facilitating risk-adapted strategies. However, such an applicable tool has ...yet to be established in the Chinese population. In this study, a risk score was created using data from the China Kadoorie Biobank (CKB), a nationwide cohort study of 409,854 eligible participants. Diagnostic performance of the risk score was evaluated in an independent CRC screening programme, which included 91,575 participants who accepted colonoscopy at designed hospitals in Zhejiang Province, China. Over a median follow-up of 11.1 years, 3136 CRC cases were documented in the CKB. A risk score was created based on nine questionnaire-derived variables, showing moderate discrimination for 10-year CRC risk (C-statistic = 0.68, 95 % CI: 0.67–0.69). In the CRC screening programme, the detection rates of CRC were 0.25 %, 0.82 %, and 1.93 % in low-risk (score <6), intermediate-risk (score: 6–19), and high-risk (score >19) groups, respectively. The newly developed score exhibited a C-statistic of 0.65 (95 % CI: 0.63–0.66), surpassing the widely adopted tools such as the Asia-Pacific Colorectal Screening (APCS), modified APCS, and Korean Colorectal Screening scores (all C-statistics = 0.60). In conclusion, we developed a novel risk prediction tool that is useful to identify individuals at high risk of CRC. A user-friendly online calculator was also constructed to encourage broader adoption of the tool.
•We developed a new risk score that effectively stratified individuals based on their CRC risk.•The novel risk score outperformed commonly used models in identifying individuals at risk for CRC.•We made the tool easily accessible online, allowing the general population to actively engage in CRC prevention.
In the past decade, precision genomics based medicine has emerged to provide tailored and effective healthcare for patients depending upon their genetic features. Genome Wide Association Studies have ...also identified population based risk genetic variants for common and complex diseases. In order to meet the full promise of precision medicine, research is attempting to leverage our increasing genomic understanding and further develop personalized medical healthcare through ever more accurate disease risk prediction models. Polygenic risk scoring and machine learning are two primary approaches for disease risk prediction. Despite recent improvements, the results of polygenic risk scoring remain limited due to the approaches that are currently used. By contrast, machine learning algorithms have increased predictive abilities for complex disease risk. This increase in predictive abilities results from the ability of machine learning algorithms to handle multi-dimensional data. Here, we provide an overview of polygenic risk scoring and machine learning in complex disease risk prediction. We highlight recent machine learning application developments and describe how machine learning approaches can lead to improved complex disease prediction, which will help to incorporate genetic features into future personalized healthcare. Finally, we discuss how the future application of machine learning prediction models might help manage complex disease by providing tissue-specific targets for customized, preventive interventions.
Risk prediction plays an important role in clinical cardiology research. Traditionally, most risk models have been based on regression models. While useful and robust, these statistical methods are ...limited to using a small number of predictors which operate in the same way on everyone, and uniformly throughout their range. The purpose of this review is to illustrate the use of machine-learning methods for development of risk prediction models. Typically presented as black box approaches, most machine-learning methods are aimed at solving particular challenges that arise in data analysis that are not well addressed by typical regression approaches. To illustrate these challenges, as well as how different methods can address them, we consider trying to predicting mortality after diagnosis of acute myocardial infarction. We use data derived from our institution's electronic health record and abstract data on 13 regularly measured laboratory markers. We walk through different challenges that arise in modelling these data and then introduce different machine-learning approaches. Finally, we discuss general issues in the application of machine-learning methods including tuning parameters, loss functions, variable importance, and missing data. Overall, this review serves as an introduction for those working on risk modelling to approach the diffuse field of machine learning.
The influence of the stability of the insurance industry on the strength of the financial market cannot be ignored, and the healthy and strong growth of the insurance industry plays an indispensable ...role in maintaining the stability of the whole society. To fully analyze the influence of the stability of the insurance industry on the strength of the financial market, this paper proposes a prediction model of systemic financial risk. The insurance industry’s systemic financial risk prediction indicators are determined based on the principle of selecting systemic risk factors and indicators. After preprocessing the collected relevant research data, relevant research variables are determined, a Logit prediction model for the systemic risk of the insurance industry is constructed, and a case study is conducted. The data show that the average value of GDP during these 21 years is 379642.64, and it reaches about 2.4 times the average value in 2023. The insurance industry’s economy grows steadily and rapidly. It reached the maximum response in the 10th period (0.04616), i.e., the development of China’s life insurance industry, which has a more obvious role in promoting the development of the economy. It is predicted that China’s financial risk in the insurance industry in the second half of 2024 will have a great probability of being in a low-risk state. Its predicted probability value is close to 1. The insurance industry’s healthy development has a positive impact on economic development, and the insurance industry’s stable development is of great importance to economic development, as evidenced by this study.
Breast cancer (BC) risk prediction allows systematic identification of individuals at highest and lowest risk. We extend the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation ...Algorithm (BOADICEA) risk model to incorporate the effects of polygenic risk scores (PRS) and other risk factors (RFs).
BOADICEA incorporates the effects of truncating variants in BRCA1, BRCA2, PALB2, CHEK2, and ATM; a PRS based on 313 single-nucleotide polymorphisms (SNPs) explaining 20% of BC polygenic variance; a residual polygenic component accounting for other genetic/familial effects; known lifestyle/hormonal/reproductive RFs; and mammographic density, while allowing for missing information.
Among all factors considered, the predicted UK BC risk distribution is widest for the PRS, followed by mammographic density. The highest BC risk stratification is achieved when all genetic and lifestyle/hormonal/reproductive/anthropomorphic factors are considered jointly. With all factors, the predicted lifetime risks for women in the UK population vary from 2.8% for the 1st percentile to 30.6% for the 99th percentile, with 14.7% of women predicted to have a lifetime risk of ≥17-<30% (moderate risk according to National Institute for Health and Care Excellence NICE guidelines) and 1.1% a lifetime risk of ≥30% (high risk).
This comprehensive model should enable high levels of BC risk stratification in the general population and women with family history, and facilitate individualized, informed decision-making on prevention therapies and screening.
Abstract Objective Calibrated risk models are vital for valid decision support. We define four levels of calibration and describe implications for model development and external validation of ...predictions. Study Design and Setting We present results based on simulated data sets. Results A common definition of calibration is “having an event rate of R % among patients with a predicted risk of R %,” which we refer to as “moderate calibration.” Weaker forms of calibration only require the average predicted risk (mean calibration) or the average prediction effects (weak calibration) to be correct. “Strong calibration” requires that the event rate equals the predicted risk for every covariate pattern. This implies that the model is fully correct for the validation setting. We argue that this is unrealistic: the model type may be incorrect, the linear predictor is only asymptotically unbiased, and all nonlinear and interaction effects should be correctly modeled. In addition, we prove that moderate calibration guarantees nonharmful decision making. Finally, results indicate that a flexible assessment of calibration in small validation data sets is problematic. Conclusion Strong calibration is desirable for individualized decision support but unrealistic and counter productive by stimulating the development of overly complex models. Model development and external validation should focus on moderate calibration.
The aim was to study the prognostic value of plasma ceramides (Cer) as cardiovascular death (CV death) markers in three independent coronary artery disease (CAD) cohorts.
Corogene study is a ...prospective Finnish cohort including stable CAD patients (n = 160). Multiple lipid biomarkers and C-reactive protein were measured in addition to plasma Cer(d18:1/16:0), Cer(d18:1/18:0), Cer(d18:1/24:0), and Cer(d18:1/24:1). Subsequently, the association between high-risk ceramides and CV mortality was investigated in the prospective Special Program University Medicine-Inflammation in Acute Coronary Syndromes (SPUM-ACS) cohort (n = 1637), conducted in four Swiss university hospitals. Finally, the results were validated in Bergen Coronary Angiography Cohort (BECAC), a prospective Norwegian cohort study of stable CAD patients. Ceramides, especially when used in ratios, were significantly associated with CV death in all studies, independent of other lipid markers and C-reactive protein. Adjusted odds ratios per standard deviation for the Cer(d18:1/16:0)/Cer(d18:1/24:0) ratio were 4.49 (95% CI, 2.24-8.98), 1.64 (1.29-2.08), and 1.77 (1.41-2.23) in the Corogene, SPUM-ACS, and BECAC studies, respectively. The Cer(d18:1/16:0)/Cer(d18:1/24:0) ratio improved the predictive value of the GRACE score (net reclassification improvement, NRI = 0.17 and ΔAUC = 0.09) in ACS and the predictive value of the Marschner score in stable CAD (NRI = 0.15 and ΔAUC = 0.02).
Distinct plasma ceramide ratios are significant predictors of CV death both in patients with stable CAD and ACS, over and above currently used lipid markers. This may improve the identification of high-risk patients in need of more aggressive therapeutic interventions.