Creation of mortality risk calculator using a I-123 mIBG-based machine learning model: differential prediction of arrhythmic death and heart-failure death

Funding Acknowledgements Type of funding sources: None. Background Although I-123 meta-iodobenzylguanidine (mIBG) has been applied to patients with chronic heart failure (CHF), a diagnostic tool for differential prediction of fatal arrhythmic events (ArE) and heart-failure death (HFD) has been pursued. Purpose The aim of this study was to create a calculator of mortality risk for differentiating mode of cardiac death using a machine learning (ML) method, and to test the accuracy in a new cohort of patients with CHF. Methods A total of 529 patients with CHF was used as the training database for ML. The ArE group consisted of patients with arrhythmic death, sudden cardiac death and appropriate therapy by implantable cardioverter defibrillator. A heart-to-mediastinum ratio (H/M) standardized to the medium-energy collimator condition was calculated with a planar anterior mIBG scintigram. The best classifier models for predicting HFD and ArE were determined by four-fold cross validation. Input variables included age, sex, New York Heart Association (NYHA) functional class, left ventricular ejection fraction, ischemic etiology, mIBG H/M and washout rate, and b-type natriuretic peptide (BNP) or NT Pro BNP, estimated glomerular filtration rate, hemoglobin, and complications such as diabetes and hypertension. After creating the ML-based model, the constructed classifier functions for ArE, HFD, and survival were exported for subsequent use. A new cohort of patients (n = 312, age 67 ± 13 years, 2015 or later) was used to test the ML-based model. Results The training database included 141 events (27%) with ArE (7%) and HFD (20%). Receiver-operating characteristic analysis by four-fold validation showed area under the curve value of 0.90 for HFD and 0.73 for ArE. Among various ML methods, the logistic regression method demonstrated the most stable calculation of the probability of ArE followed by random forest and gradient boosted tree methods. Therefore, the logistic-regression method was used for calculating both HFD and ArE probabilities. In the test cohort, patients with a high HFD probability >8% resulted in 6.3-fold higher HFD than those with low probability (≤ 8%). Patients with high ArE probability >8% showed 2.5-fold higher ArE than those with low probability (≤ 8%). Conclusion The ML-based mortality risk calculator could be used for stratifying patients at high and low risks, which might be useful for estimating appropriate treatment strategy.

Current approach to perinatal risk stratification

Objective. To develop a new approach to perinatal risk stratification based on the determination of prognostic criteria for antepartum and intrapartum fetal death, and early neonatal death to improve the efficiency of predicting adverse perinatal outcomes. Patients and methods. A retrospective case-control study with the participation of patients with antepartum (n = 134) and intrapartum (n = 41) fetal death, early neonatal death (n = 61), and favorable perinatal outcome (n = 50) was carried out. The Bayes–Wald–Gubler method was used to determine prognostic criteria. Comparative evaluation of the efficacy of the proposed prognostic method and the generally accepted determination of perinatal risk was performed retrospectively in patients with perinatal loss (n = 102) and favorable perinatal outcome (n = 100); ROC analysis was performed. Results. Forty-two risk factors were identified and divided into three groups: universal for all types of perinatal loss, common for two of them, and specific for each of them (antepartum and intrapartum fetal death, early neonatal death). The prognostic value of factors in their presence and absence was determined. It was found that universal and common risk factors for each type of perinatal loss had different prognostic value. The method of differential prediction of perinatal loss was presented. The sensitivity of the new and generally accepted prognostic methods was 95.1 and 69.6%, the specificity was 80 and 53%, and the accuracy of predicting adverse outcomes was 87.6 and 61.4%, respectively. Conclusion. The conducted study allowed to suggest a new approach to perinatal risk stratification based on differential prediction of perinatal loss, which is superior to the traditional risk assessment methods in terms of the effectiveness of predicting adverse perinatal outcomes. Key words: antepartum fetal death, intrapartum fetal death, perinatal outcomes, prediction, early neonatal death

Detection for Frying Times of Various Edible Oils Based on Near-Infrared Spectroscopy

Taking a variety of edible oils as the research object, including soybean oil, peanut oil, rapeseed oil, a method based on Near-Infrared Spectroscopy (NIRS) to identify the frying times is proposed to evaluate the quality of frying oil. Ten rounds of frying experiments are carried out for each of the three oils. The spectra of the first eight rounds are used to build the model, and the last two are used for model testing. First, all the original spectra are preprocessed using the first derivative (1D). Then, the correlation coefficient between the sequence of frying times and absorbance is calculated, and the characteristic wavelengths with a high correlation coefficient are extracted. Finally, a differential prediction model is established based on the characteristic wavelengths. The results show that the differential prediction model accurately predicts the frying times of various edible oils and provides a new method for quality inspection of frying oil, and the predicted accuracy of the frying times of three frying oils is 100% within the allowable range of error.

Predictive Properties and Factor Structure of the VRS-SO in an Austrian Sample

We examined the structural and predictive properties of the Violence Risk Scale-Sexual Offense (VRS-SO) version in an Austrian sample of N = 666 men incarcerated for sexual offenses; 353 of whom were followed up an average of 11 years post-release. Results of a confirmatory factor analysis of dynamic item scores supported a three-factor model (Sexual Deviance, Criminality, and Treatment Responsivity) consistent with prior research. VRS-SO static, dynamic, and total scores showed good properties of discrimination for sexual (area under the receiver operating curve [AUC] = .68–.80) and violent (AUC = .65–.68) recidivism, while the factor scores showed differential prediction of these outcomes. Calibration analyses demonstrated lower estimated rates of 5-year sexual reoffense associated with VRS-SO score bands in the present sample compared to observed rates from the normative sample, with closest correspondence observed for the highest risk band (E/O index = 1.01). Implications for the psychometric properties and application of the VRS-SO in international settings are discussed.

The approximate formulas predicting personal somatic telomere length using patient blood test data

Biological aging underlies lifestyle-related diseases. It can be assessed by measuring personal somatic cell telomere length. However, measuring the telomere length is laborious, and its clinical surrogate parameters have not been developed. This study analyzed the correlation between telomere length in peripheral leukocytes and laboratory data to select test items relating closely to biological aging. We established formulas from these clinical data to predict the personal telomere length. The subjects were patients having visited Kyushu University Beppu Hospital from 2012 to 2015. Two hundred and thirty-two patients were enrolled. The blood data were collected and telomere lengths were measured by Southern blotting method. The patients showed significant correlations between the telomere length and several blood test data with a sex-related difference. Candidate formulas are as follows: Predicted telomere length (kb) in men = 8.59 − 0.037 × Age (years) + 0.024 × Hemoglobin (g/dL); Predicted telomere length (kb) in women = 4.83 − 0.019 × Age (years) + 0.23 × Albumin (g/dL) + 0.0001 × White blood cells (/mm3) + 0.0020 × Red blood cells (× 104/mm3) + 0.0032 × Total cholesterol (mg/dL). Thus, the derived formulas allow for the accurate differential prediction of telomeric length in male and female patients.

The GRE and Students With Disabilities: A Validity Study at 10 Universities

We evaluated the validity of using the GRE General Test to assist with graduate school admissions for individuals with disabilities. We studied a sample of 16,239 graduate students from 10 U.S. research universities in three groups: students without any reported disabilities, students who reported disabilities and took the computer-delivered GRE with accommodations, and students who reported disabilities but took the computer-delivered GRE without accommodations. We examined differential prediction using multilevel modeling and residual analyses. The results revealed that the first-year graduate grade point average of students with disabilities was neither over- nor underpredicted by more than one tenth of a point on the 0- to 4-grading scale. However, variations on the magnitude and direction of differential prediction existed among students with different types of disabilities. We discuss data collection needs and research on students with disabilities attending graduate and professional schools.

A National Study of the Differential Impact of Novice Teacher Certification on Teacher Traits and Race-Based Mathematics Achievement

In this study, differential prediction of student outcomes by race and teacher traits relative to the certification levels of novice teachers was assessed. Overall, algebra achievement was higher for students who were taught by teachers with standard certificates relative to students who were taught by novice teachers with nonstandard certificates. The most conservative estimates show that findings are equivalent to approximately 8 months of additional instruction for students who were taught by teachers with standard certificates. However, the benefits of being taught by a teacher with standard certification did not translate to underrepresented racial groups. Overall, there were several differences in dispositions across certification conditions. With respect to underrepresented racial groups, only one disposition was significantly different across conditions. Teachers with nonstandard certificates reported higher emphasis on increasing mathematics interests. For underrepresented racial groups, relationships between both certification conditions and achievement were underwhelming. Recommendations to improve teacher effectiveness are provided.