Machine Learning Models for Pancreatic Cancer Risk Prediction Using Electronic Health Record Data-A Systematic Review and Assessment.

INTRODUCTION: Accurate risk prediction can facilitate screening and early detection of pancreatic cancer (PC). We conducted a systematic review to critically evaluate effectiveness of machine learning (ML) and artificial intelligence (AI) techniques applied to electronic health records (EHR) for PC risk prediction. METHODS: Ovid MEDLINE(R), Ovid EMBASE, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, Scopus, and Web of Science were searched for articles that utilized ML/AI techniques to predict PC, published between January 1, 2012, and February 1, 2024. Study selection and data extraction were conducted by 2 independent reviewers. Critical appraisal and data extraction were performed using the CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies checklist. Risk of bias and applicability were examined using prediction model risk of bias assessment tool. RESULTS: Thirty studies including 169,149 PC cases were identified. Logistic regression was the most frequent modeling method. Twenty studies utilized a curated set of known PC risk predictors or those identified by clinical experts. ML model discrimination performance (C-index) ranged from 0.57 to 1.0. Missing data were underreported, and most studies did not implement explainable-AI techniques or report exclusion time intervals. DISCUSSION: AI/ML models for PC risk prediction using known risk factors perform reasonably well and may have near-term applications in identifying cohorts for targeted PC screening if validated in real-world data sets. The combined use of structured and unstructured EHR data using emerging AI models while incorporating explainable-AI techniques has the potential to identify novel PC risk factors, and this approach merits further study.

Regional assessment of in vivo myocardial stiffness using 3D magnetic resonance elastography in a porcine model of myocardial infarction.

PURPOSE: The stiffness of a myocardial infarct affects the left ventricular pump function and remodeling. Magnetic resonance elastography (MRE) is a noninvasive imaging technique for measuring soft-tissue stiffness in vivo. The purpose of this study was to investigate the feasibility of assessing in vivo regional myocardial stiffness with high-frequency 3D cardiac MRE in a porcine model of myocardial infarction, and compare the results with ex vivo uniaxial tensile testing. METHODS: Myocardial infarct was induced in a porcine model by embolizing the left circumflex artery. Fourteen days postinfarction, MRE imaging was performed in diastole using an echocardiogram-gated spin-echo echo-planar-imaging sequence with 140-Hz vibrations and 3D MRE processing. The MRE stiffness and tensile modulus from uniaxial testing were compared between the remote and infarcted myocardium. RESULTS: Myocardial infarcts showed increased in vivo MRE stiffness compared with remote myocardium (4.6 ± 0.7 kPa versus 3.0 ± 0.6 kPa, P = 0.02) within the same pig. Ex vivo uniaxial mechanical testing confirmed the in vivo MRE results, showing that myocardial infarcts were stiffer than remote myocardium (650 ± 80 kPa versus 110 ± 20 kPa, P = 0.01). CONCLUSIONS: These results demonstrate the feasibility of assessing in vivo regional myocardial stiffness with high-frequency 3D cardiac MRE. Magn Reson Med 79:361-369, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Cardiac MR elastography for quantitative assessment of elevated myocardial stiffness in cardiac amyloidosis.

PURPOSE: To evaluate if cardiac magnetic resonance elastography (MRE) can measure increased stiffness in patients with cardiac amyloidosis. Myocardial tissue stiffness plays an important role in cardiac function. A noninvasive quantitative imaging technique capable of measuring myocardial stiffness could aid in disease diagnosis, therapy monitoring, and disease prognostic strategies. We recently developed a high-frequency cardiac MRE technique capable of making noninvasive stiffness measurements. MATERIALS AND METHODS: In all, 16 volunteers and 22 patients with cardiac amyloidosis were enrolled in this study after Institutional Review Board approval and obtaining formal written consent. All subjects were imaged head-first in the supine position in a 1.5T closed-bore MR imager. 3D MRE was performed using 5 mm isotropic resolution oblique short-axis slices and a vibration frequency of 140 Hz to obtain global quantitative in vivo left ventricular stiffness measurements. The median stiffness was compared between the two cohorts. An octahedral shear strain signal-to-noise ratio (OSS-SNR) threshold of 1.17 was used to exclude exams with insufficient motion amplitude. RESULTS: Five volunteers and six patients had to be excluded from the study because they fell below the 1.17 OSS-SNR threshold. The myocardial stiffness of cardiac amyloid patients (median: 11.4 kPa, min: 9.2, max: 15.7) was significantly higher (P = 0.0008) than normal controls (median: 8.2 kPa, min: 7.2, max: 11.8). CONCLUSION: This study demonstrates the feasibility of 3D high-frequency cardiac MRE as a contrast-agent-free diagnostic imaging technique for cardiac amyloidosis. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2017;46:1361-1367.