Dataset Information

Predicting Hepatocellular Carcinoma With Minimal Features From Electronic Health Records: Development of a Deep Learning Model.

ABSTRACT:

Background

Hepatocellular carcinoma (HCC), usually known as hepatoma, is the third leading cause of cancer mortality globally. Early detection of HCC helps in its treatment and increases survival rates.

Objective

The aim of this study is to develop a deep learning model, using the trend and severity of each medical event from the electronic health record to accurately predict the patients who will be diagnosed with HCC in 1 year.

Methods

Patients with HCC were screened out from the National Health Insurance Research Database of Taiwan between 1999 and 2013. To be included, the patients with HCC had to register as patients with cancer in the catastrophic illness file and had to be diagnosed as a patient with HCC in an inpatient admission. The control cases (non-HCC patients) were randomly sampled from the same database. We used age, gender, diagnosis code, drug code, and time information as the input variables of a convolution neural network model to predict those patients with HCC. We also inspected the highly weighted variables in the model and compared them to their odds ratio at HCC to understand how the predictive model works.

Results

We included 47,945 individuals, 9553 of whom were patients with HCC. The area under the receiver operating curve (AUROC) of the model for predicting HCC risk 1 year in advance was 0.94 (95% CI 0.937-0.943), with a sensitivity of 0.869 and a specificity 0.865. The AUROC for predicting HCC patients 7 days, 6 months, 1 year, 2 years, and 3 years early were 0.96, 0.94, 0.94, 0.91, and 0.91, respectively.

Conclusions

The findings of this study show that the convolutional neural network model has immense potential to predict the risk of HCC 1 year in advance with minimal features available in the electronic health records.

SUBMITTER: Liang CW

PROVIDER: S-EPMC8587326 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Abstract Objective Hypertension has long been recognized as one of the most important predisposing factors for cardiovascular diseases and mortality. In recent years, machine learning methods have shown potential in diagnostic and predictive approaches in chronic diseases. Electronic health records (EHRs) have emerged as a reliable source of longitudinal data. The aim of this study is to predict the onset of hypertension using modern deep learning (DL) architectures, specifically long short-term memory (LSTM) networks, and longitudinal EHRs. Materials and Methods We compare this approach to the best performing models reported from previous works, particularly XGboost, applied to aggregated features. Our work is based on data from 233 895 adult patients from a large health system in the United States. We divided our population into 2 distinct longitudinal datasets based on the diagnosis date. To ensure generalization to unseen data, we trained our models on the first dataset (dataset A “train and validation”) using cross-validation, and then applied the models to a second dataset (dataset B “test”) to assess their performance. We also experimented with 2 different time-windows before the onset of hypertension and evaluated the impact on model performance. Results With the LSTM network, we were able to achieve an area under the receiver operating characteristic curve value of 0.98 in the “train and validation” dataset A and 0.94 in the “test” dataset B for a prediction time window of 1 year. Lipid disorders, type 2 diabetes, and renal disorders are found to be associated with incident hypertension. Conclusion These findings show that DL models based on temporal EHR data can improve the identification of patients at high risk of hypertension and corresponding driving factors. In the long term, this work may support identifying individuals who are at high risk for developing hypertension and facilitate earlier intervention to prevent the future development of hypertension.

Project description:PurposeAdvances in artificial intelligence have produced a few predictive models in glaucoma, including a logistic regression model predicting glaucoma progression to surgery. However, uncertainty exists regarding how to integrate the wealth of information in free-text clinical notes. The purpose of this study was to predict glaucoma progression requiring surgery using deep learning (DL) approaches on data from electronic health records (EHRs), including features from structured clinical data and from natural language processing of clinical free-text notes.DesignDevelopment of DL predictive model in an observational cohort.ParticipantsAdult patients with glaucoma at a single center treated from 2008 through 2020.MethodsOphthalmology clinical notes of patients with glaucoma were identified from EHRs. Available structured data included patient demographic information, diagnosis codes, prior surgeries, and clinical information including intraocular pressure, visual acuity, and central corneal thickness. In addition, words from patients' first 120 days of notes were mapped to ophthalmology domain-specific neural word embeddings trained on PubMed ophthalmology abstracts. Word embeddings and structured clinical data were used as inputs to DL models to predict subsequent glaucoma surgery.Main outcome measuresEvaluation metrics included area under the receiver operating characteristic curve (AUC) and F1 score, the harmonic mean of positive predictive value, and sensitivity on a held-out test set.ResultsSeven hundred forty-eight of 4512 patients with glaucoma underwent surgery. The model that incorporated both structured clinical features as well as input features from clinical notes achieved an AUC of 73% and F1 of 40%, compared with only structured clinical features, (AUC, 66%; F1, 34%) and only clinical free-text features (AUC, 70%; F1, 42%). All models outperformed predictions from a glaucoma specialist's review of clinical notes (F1, 29.5%).ConclusionsWe can successfully predict which patients with glaucoma will need surgery using DL models on EHRs unstructured text. Models incorporating free-text data outperformed those using only structured inputs. Future predictive models using EHRs should make use of information from within clinical free-text notes to improve predictive performance. Additional research is needed to investigate optimal methods of incorporating imaging data into future predictive models as well.

Project description:Determining the discrepancy between chronological and physiological age of patients is central to preventative and personalized care. Electronic medical records (EMR) provide rich information about the patient physiological state, but it is unclear whether such information can be predictive of chronological age. Here we present a deep learning model that uses vital signs and lab tests contained within the EMR of Mount Sinai Health System (MSHS) to predict chronological age. The model is trained on 377,686 EMR from patients of ages 18-85 years old. The discrepancy between the predicted and real chronological age is then used as a proxy to estimate physiological age. Overall, the model can predict the chronological age of patients with a standard deviation error of ∼7 years. The ages of the youngest and oldest patients were more accurately predicted, while patients of ages ranging between 40 and 60 years were the least accurately predicted. Patients with the largest discrepancy between their physiological and chronological age were further inspected. The patients predicted to be significantly older than their chronological age have higher systolic blood pressure, higher cholesterol, damaged liver, and anemia. In contrast, patients predicted to be younger than their chronological age have lower blood pressure and shorter stature among other indicators; both groups display lower weight than the population average. Using information from ∼10,000 patients from the entire cohort who have been also profiled with SNP arrays, genome-wide association study (GWAS) uncovers several novel genetic variants associated with aging. In particular, significant variants were mapped to genes known to be associated with inflammation, hypertension, lipid metabolism, height, and increased lifespan in mice. Several genes with missense mutations were identified as novel candidate aging genes. In conclusion, we demonstrate how EMR data can be used to assess overall health via a scale that is based on deviation from the patient's predicted chronological age.

Project description:BackgroundDeep learning models showed great success and potential when applied to many biomedical problems. However, the accuracy of deep learning models for many disease prediction problems is affected by time-varying covariates, rare incidence, and covariate imbalance when using structured electronic health records data. The situation is further exasperated when predicting the risk of one disease on condition of another disease, such as the hepatocellular carcinoma risk among patients with nonalcoholic fatty liver disease due to slow, chronic progression, the scarce of data with both disease conditions and the sex bias of the diseases.ObjectiveThe goal of this study is to investigate the extent to which time-varying covariates, rare incidence, and covariate imbalance influence deep learning performance, and then devised strategies to tackle these challenges. These strategies were applied to improve hepatocellular carcinoma risk prediction among patients with nonalcoholic fatty liver disease.MethodsWe evaluated two representative deep learning models in the task of predicting the occurrence of hepatocellular carcinoma in a cohort of patients with nonalcoholic fatty liver disease (n = 220,838) from a national EHR database. The disease prediction task was carefully formulated as a classification problem while taking censorship and the length of follow-up into consideration.ResultsWe developed a novel backward masking scheme to evaluate how the length of longitudinal information after the index date affects disease prediction. We observed that modeling time-varying covariates improved the performance of the algorithms and transfer learning mitigated reduced performance caused by the lack of data. In addition, covariate imbalance, such as sex bias in data impaired performance. Deep learning models trained on one sex and evaluated in the other sex showed reduced performance, indicating the importance of assessing covariate imbalance while preparing data for model training.ConclusionsDevising proper strategies to address challenges from time-varying covariates, lack of data, and covariate imbalance can be key to counteracting data bias and accurately predicting disease occurrence using deep learning models. The novel strategies developed in this work can significantly improve the performance of hepatocellular carcinoma risk prediction among patients with nonalcoholic fatty liver disease. Furthermore, our novel strategies can be generalized to apply to other disease risk predictions using structured electronic health records, especially for disease risks on condition of another disease.

Project description:BackgroundEmergency admissions are a major source of healthcare spending. We aimed to derive, validate, and compare conventional and machine learning models for prediction of the first emergency admission. Machine learning methods are capable of capturing complex interactions that are likely to be present when predicting less specific outcomes, such as this one.Methods and findingsWe used longitudinal data from linked electronic health records of 4.6 million patients aged 18-100 years from 389 practices across England between 1985 to 2015. The population was divided into a derivation cohort (80%, 3.75 million patients from 300 general practices) and a validation cohort (20%, 0.88 million patients from 89 general practices) from geographically distinct regions with different risk levels. We first replicated a previously reported Cox proportional hazards (CPH) model for prediction of the risk of the first emergency admission up to 24 months after baseline. This reference model was then compared with 2 machine learning models, random forest (RF) and gradient boosting classifier (GBC). The initial set of predictors for all models included 43 variables, including patient demographics, lifestyle factors, laboratory tests, currently prescribed medications, selected morbidities, and previous emergency admissions. We then added 13 more variables (marital status, prior general practice visits, and 11 additional morbidities), and also enriched all variables by incorporating temporal information whenever possible (e.g., time since first diagnosis). We also varied the prediction windows to 12, 36, 48, and 60 months after baseline and compared model performances. For internal validation, we used 5-fold cross-validation. When the initial set of variables was used, GBC outperformed RF and CPH, with an area under the receiver operating characteristic curve (AUC) of 0.779 (95% CI 0.777, 0.781), compared to 0.752 (95% CI 0.751, 0.753) and 0.740 (95% CI 0.739, 0.741), respectively. In external validation, we observed an AUC of 0.796, 0.736, and 0.736 for GBC, RF, and CPH, respectively. The addition of temporal information improved AUC across all models. In internal validation, the AUC rose to 0.848 (95% CI 0.847, 0.849), 0.825 (95% CI 0.824, 0.826), and 0.805 (95% CI 0.804, 0.806) for GBC, RF, and CPH, respectively, while the AUC in external validation rose to 0.826, 0.810, and 0.788, respectively. This enhancement also resulted in robust predictions for longer time horizons, with AUC values remaining at similar levels across all models. Overall, compared to the baseline reference CPH model, the final GBC model showed a 10.8% higher AUC (0.848 compared to 0.740) for prediction of risk of emergency admission within 24 months. GBC also showed the best calibration throughout the risk spectrum. Despite the wide range of variables included in models, our study was still limited by the number of variables included; inclusion of more variables could have further improved model performances.ConclusionsThe use of machine learning and addition of temporal information led to substantially improved discrimination and calibration for predicting the risk of emergency admission. Model performance remained stable across a range of prediction time windows and when externally validated. These findings support the potential of incorporating machine learning models into electronic health records to inform care and service planning.

Project description:OBJECTIVES:Predictive models for hepatocellular carcinoma (HCC) have been limited by modest accuracy and lack of validation. Machine-learning algorithms offer a novel methodology, which may improve HCC risk prognostication among patients with cirrhosis. Our study's aim was to develop and compare predictive models for HCC development among cirrhotic patients, using conventional regression analysis and machine-learning algorithms. METHODS:We enrolled 442 patients with Child A or B cirrhosis at the University of Michigan between January 2004 and September 2006 (UM cohort) and prospectively followed them until HCC development, liver transplantation, death, or study termination. Regression analysis and machine-learning algorithms were used to construct predictive models for HCC development, which were tested on an independent validation cohort from the Hepatitis C Antiviral Long-term Treatment against Cirrhosis (HALT-C) Trial. Both models were also compared with the previously published HALT-C model. Discrimination was assessed using receiver operating characteristic curve analysis, and diagnostic accuracy was assessed with net reclassification improvement and integrated discrimination improvement statistics. RESULTS:After a median follow-up of 3.5 years, 41 patients developed HCC. The UM regression model had a c-statistic of 0.61 (95% confidence interval (CI) 0.56-0.67), whereas the machine-learning algorithm had a c-statistic of 0.64 (95% CI 0.60-0.69) in the validation cohort. The HALT-C model had a c-statistic of 0.60 (95% CI 0.50-0.70) in the validation cohort and was outperformed by the machine-learning algorithm. The machine-learning algorithm had significantly better diagnostic accuracy as assessed by net reclassification improvement (P<0.001) and integrated discrimination improvement (P=0.04). CONCLUSIONS:Machine-learning algorithms improve the accuracy of risk stratifying patients with cirrhosis and can be used to accurately identify patients at high-risk for developing HCC.

Project description:IntroductionPredictive models have been used to aid early diagnosis of PCOS, though existing models are based on small sample sizes and limited to fertility clinic populations. We built a predictive model using machine learning algorithms based on an outpatient population at risk for PCOS to predict risk and facilitate earlier diagnosis, particularly among those who meet diagnostic criteria but have not received a diagnosis.MethodsThis is a retrospective cohort study from a SafetyNet hospital's electronic health records (EHR) from 2003-2016. The study population included 30,601 women aged 18-45 years without concurrent endocrinopathy who had any visit to Boston Medical Center for primary care, obstetrics and gynecology, endocrinology, family medicine, or general internal medicine. Four prediction outcomes were assessed for PCOS. The first outcome was PCOS ICD-9 diagnosis with additional model outcomes of algorithm-defined PCOS. The latter was based on Rotterdam criteria and merging laboratory values, radiographic imaging, and ICD data from the EHR to define irregular menstruation, hyperandrogenism, and polycystic ovarian morphology on ultrasound.ResultsWe developed predictive models using four machine learning methods: logistic regression, supported vector machine, gradient boosted trees, and random forests. Hormone values (follicle-stimulating hormone, luteinizing hormone, estradiol, and sex hormone binding globulin) were combined to create a multilayer perceptron score using a neural network classifier. Prediction of PCOS prior to clinical diagnosis in an out-of-sample test set of patients achieved an average AUC of 85%, 81%, 80%, and 82%, respectively in Models I, II, III and IV. Significant positive predictors of PCOS diagnosis across models included hormone levels and obesity; negative predictors included gravidity and positive bHCG.ConclusionMachine learning algorithms were used to predict PCOS based on a large at-risk population. This approach may guide early detection of PCOS within EHR-interfaced populations to facilitate counseling and interventions that may reduce long-term health consequences. Our model illustrates the potential benefits of an artificial intelligence-enabled provider assistance tool that can be integrated into the EHR to reduce delays in diagnosis. However, model validation in other hospital-based populations is necessary.