ROC Area Research Articles

P1472Predicting significant ventricular arrhythmias in STEMI patients in middle-range and preserved EF

Abstract Background Predicting malignant ventricular and sudden cardiac death (SCD) in STEMI patients with middle-range and preserved EF is challenge. Aim To identify the best parameters to predict composite end-point defined as secondary VF, sustained/non-sustained VT and sudden death, 48h after and during the first year of follow up after STEMI in patients with middle-range and preserved EF. Methods In the 192 consecutive STEMI patients (pts) 57.8±10.4yr, 69.9% males, in PREDICT-VT study (NCT03263949) treated with pPCI, with EF ≥40%, early echo (5±2 days) was done including conventional parameters and comprehensive speckle tracking left ventricle (LV) deformation analysis with longitudinal (L), circumferential (C) strain (S; %) strain rate (SR, 1/sec), index mechanical dispersion (IMD) and rotational LV mechanics. Results Thirteen patients (8.3%) reached the end-point. Classical parameters of LV systolic function, including LVEF, wall motion score index and parameters of diastolic dysfunction were not significant predictors of the malignant arrhythmias. IMD of late rotation rate (63.7 vs. 40.7ms, p=0.055) and late diastolic untwisting rate (−48.85 vs. −63.18°/s, p=0.059) had trend to become the significant predictors. CS in papillary muscle level in endo and mid layers predicted the primary end-point (endo: −20.5±11.8 vs. −24.9±4.6, mid: −14.6±3.9 vs. −17.0±2.1, epi: −10.1±3.3 vs. −11.8±1.8) (table). Parameter of circumferencial mechanics ROC area 95% CI p Cutt-off Sens Spec PM endo (%) 0.302 0.146–0.458 0.038 −22.75 70 67 PM mid (%) 0268 0.153–0.383 0.015 −15.65 80 62 Conclusion Myocardial deformation imaging offers deeper insight into complex mechanical abnormalities during LV contraction and relaxation in STEMI patients with middle-range and preserved EF that predicts serious arrhythmic events.

Electrostatic-field and surface-shape similarity for virtual screening and pose prediction

We introduce a new method for rapid computation of 3D molecular similarity that combines electrostatic field comparison with comparison of molecular surface-shape and directional hydrogen-bonding preferences (called “eSim”). Rather than employing heuristic “colors” or user-defined molecular feature types to represent conformation-dependent molecular electrostatics, eSim calculates the similarity of the electrostatic fields of two molecules (in addition to shape and hydrogen-bonding). We present detailed virtual screening performance data on the standard 102 target DUD-E set. In its moderately fast screening mode, eSim running on a single computing core is capable of processing over 60 molecules per second. In this mode, eSim performed significantly better than all alternate methods for which full DUD-E data were available (mean ROC area of 0.74, p < 10^{-9}, by paired t-test, compared with the best performing alternate method). In addition, for 92 targets of the DUD-E set where multiple ligand-bound crystal structures were available, screening performance was assessed using alternate ligands or sets thereof (in their bound poses) as similarity targets. Using the joint alignment of five ligands for each protein target, mean ROC area exceeded 0.82 for the 92 targets. Design-focused application of ligand similarity methods depends on accurate predictions of geometric molecular relationships. We comprehensively assessed pose prediction accuracy by curating nearly 400,000 bound ligand pose pairs across the DUD-E targets. Overall, beginning from agnostic initial poses, we observed an 80% success rate for RMSD le 2.0 Å among the top 20 predicted eSim poses. These examples were split roughly 50/50 into cases with high direct atomic overlap (where a shared scaffold exists between a pair) and low direct atomic overlap (where where a ligand pair has dissimilar scaffolds but largely occupies the same space). Within the high direct atomic overlap subset, the pose prediction success rate was 93%. For the more challenging subset (where dissimilar scaffolds are to be aligned), the success rate was 70%. The eSim approach enables both large-scale screening and rational design of ligands and is rooted in physically meaningful, non-heuristic, molecular comparisons.

P1772Diagnostic accuracy of a novel electrocardiographic criterion for the diagnosis of left ventricular hypertrophy in hypertrophic cardiomyopathy

Abstract Introduction The 12-lead electrocardiogram (ECG) is a fundamental initial diagnostic modality for the early evaluation of a patient suspected of having hypertrophic cardiomyopathy (HCM). ECG criteria for the diagnosis of left ventricular hypertrophy (LVH) typically have low sensitivity and high specificity. Recently, a novel ECG criterion (Peguero-Lo Presti, PLP) with higher sensitivity (62%) and similar specificity (90%) was developed in a cohort of hypertensive patients, but its accuracy in patients with HCM has not been tested. We hypothesized that Peguero-Lo Presti criterion would improve upon the sensitivity of other criteria, while maintaining high specificity, for the diagnosis of LVH in patients for with HCM. Methods We retrospectively analyzed 215 consecutive patients who underwent cardiac magnetic resonance (CMR) between 2010 and 2018 for suspected HCM. All patients aged 18 years or older, who had CMR-confirmed HCM and an ECG without confounders (complete left or right bundle brunch block or paced ventricular rhythm) were included for analysis (n=88). Left ventricular mass (LVM) index and maximum wall thickness were derived from CMR analysis. The PLP criteria was defined as the sum of the deepest S wave (SD) in any lead and the S wave amplitude of lead V4 (SV4). Cornell voltage (CL) and Sokolow-Lyon (SL) were used for comparison. We randomly selected 88 gender-matched patients who performed an ECG and CMR for other clinical reasons and who had no structural heart disease or LVH for use as controls. The DeLong and McNemar's test were used to compare ROC area under the curve (AUC) and sensitivity and specificity, respectively, between the three criteria. Results 88 patients with HCM (63% male, mean age 56.7±15 years) were analyzed. The mean maximum wall thickness was 19.9±4.4mm and mean indexed LVM was 89.7±27g/m2. 34 patients (38.6%) had increased indexed LVM and 77 (87.5%) had at least one segment with late gadolinium enhancement (LGE). Discrimination by AUC was highest for PLP (0.85 [95% CI 0.8–0.9]), compared to CL (0.79, p=0.03) and SL (0.73, p=0.02). Using literature cut-offs, the sensitivity of PLP (60% [95% CI 50–70%]) was significantly higher compared to CL (40% [95% CI 30–50%, p&lt;0,001) and SL (41%, [95% CI 31–51%], p=0.01), whilst maintaining high specificity (PLP 96%; CL 98%; SL 94%). After adjusting for LVM, the amount of LGE had a positive correlation with PLP amplitude (Spearman's rho=0.6, coef=2.4, p=0.01), but not Cornell or Sokolow. The sensitivity of PLP was significantly higher than CL and Sokolow in patients with LGE (61% vs 44% vs 43%, p&lt;0.05). Conclusion The Peguero-Lo Presti criteria demonstrated higher sensitivity and similar specificity when compared to the Cornell and Sokolow-Lyon criteria for the diagnosis of LVH in a cohort of patients with hypertrophic cardiomyopathy. Therefore, they could become the standard ECG diagnostic criteria in patients suspected of having LVH and HCM.

P4748HASBLED score, frailty index or comorbidity index for bleeding risk assessment in patients with atrial fibrillation?

Abstract Background Charlson comorbidity index (CCI) is a tool to measure comorbid disease status and a strong estimator of mortality. The quantifiable frailty phenotype has also been validated as predictive of mortality and disability. Claims data can be used to classify individuals as frail and non-frail using the Claims-based Frailty Index (CFI). We evaluated whether these tools may help to predict the risk of bleeding in patients with atrial fibrillation (AF). Methods All patients with AF seen in an academic institution were identified and followed up for mortality, stroke and bleeding events. HAS-BLED, HEMORR2HAGES, ATRIA and ORBIT scores, CCI and CFI were calculated for each patient. Hazard ratios were calculated and predictive abilities of the scores were compared using the c-statistic in the whole population and then separately in elderly patients (&gt;75 yo). Results Among 8962 patients with AF, 274 major bleeding events were recorded during a follow-up of 874±1054 days. Bleeding occurred more commonly in patients with higher bleeding risk scores, CCI and CFI. The 4 bleeding risk scores significantly had lower c-statistics than CCI and CFI for predicting major bleeding (table). Results were similar whether patients were treated with OAC or no OAC. In elderly patients, the c-statistics were all lower and were not significantly different for the 4 scores, CCI and CFI scores (0.594, 0,572, 0.595, 0.594, 0.616 and 0.591 for HAS-BLED, HEMORR2HAGES, ATRIA, ORBIT, CCI and CFI, respectively). Predictive values for major bleeding ROC Area 95% Conf. Interval P value vs CCI/CFI HASBLED 0.588 0.555–0.621 0.002/0.003 HEMORR2HAGES 0.564 0.531–0.598 &lt;0.0001/&lt;0.0001 ATRIA 0.559 0.522–0.595 &lt;0.0001/&lt;0.0001 ORBIT 0.577 0.542–0.612 0.0002/0.0003 Charlson, CCI 0.652 0.619–0.684 –/0.58 Frailty index, CFI 0.648 0.615–0.681 0.58/– Conclusion Comorbidities and frailty, respectively assessed with CCI and CFI, demonstrated statistically better performances in predicting major bleeding than the 4 established bleeding risk scores in AF, although all c-indexes were broadly similar. The 4 bleeding risk scores, CCI and CFI showed lower performance in predicting bleeding within elderly patients in whom they all performed equally to predict bleeding events. Given their simplicity and similar performances, the user-friendly bleeding risk scores remain attractive tools for the estimation of bleeding risk in elderly patients with AF.

P5977Predicting significant ventricular arrhythmias in STEMI patients: never-ending challenge, still more place for myocardial deformation imaging?

Abstract Background Malignant ventricular arrhythmias in STEMI patients carry ominous prognosis including sudden cardiac death (SCD). According to the current guidelines only EF&lt;35%, 40 days after STEMI, is indication for ICD implantation. Recently, index of myocardial dispersion (IMD) estimated by myocardial deformation imaging (speckle tracking echocardiography) was documented to provide better risk stratification. Aim To define whether quantification of myocardial mechanics early after pPCI using modern echocardiography offers information more to predict malignant arrhythmias during the first year after STEMI. Methods In the 226 consecutive STEMI patients (pts) 57.8±10.4yr, 71.7% males, in PREDICT-VT study (NCT03263949) treated with pPCI early echo (5±2 days) was done including conventional parameters and comprehensive speckle tracking LV deformation analysis with longitudinal (L), circumferential (C) strain (S;%) and strain rate (SR, 1/sec) and rotational LV mechanics. ROC analysis was performed to identify the best parameters to predict composite end-point defined as secondary VF, sustained/non-sustained VT and SCD, 48h after pPCI and during the first year of follow up. Results Twenty two patients (9.7%) reached the end-point. Classical parameters of LV systolic function, including LVEF, wall motion score index; global, systolic LS, CS and parameters of diastolic dysfunction were not significant predictors of the malignant arrhythmias. Early L SR, systolic C SR, IMD of global rotation and late rotation rate predicted the primary end-point (table). Parameter ROC area 95% CI p Cutt-off Sens Spec Longirudinal mechanics SR E (1/sec) 0.687 0.577–0.796 0.019 0.69 64 65 IMD S (ms) 0.752 0.666–0.838 0.002 66.1 71 72 Circumferencial mechanics SR S (1/sec) 0.732 0.613–0.852 0.002 −1.22 71 67 Rotational mechanics Global IMD (ms) 0.329 0.177–0.481 0.036 82.9 63 62 Late rotation rate IMD (ms) 0.318 0.196–0.442 0.026 41.1 65 64 Conclusion Myocardial deformation imaging offers deeper insight into complex mechanical abnormalities during LV contraction and relaxation in longitudinal, circumferential and rotational directions (impaired and asynchronous deformations) in STEMI patients and predicts serious arrhythmic events.

Monocyte/lymphocyte ratio is associated with carotid stenosis in ischemic stroke: A retrospective analysis.

BackgroundCarotid artery stenosis, mainly caused by carotid atherosclerosis, is related to ischemic stroke. This study was to investigate whether monocyte/lymphocyte ratio (MLR) was associated with increased severity of carotid stenosis in patients with ischemic stroke.MethodsA total of 395 participants with ischemic stroke were retrospectively analyzed. The severity of carotid stenosis was evaluated by ultrasound examination. Patients were divided into two groups: nonsevere stenosis group and severe stenosis group. Multivariate logistic analysis was used to evaluate risk factors.ResultsA significant correlation was found between MLR and the severity of carotid stenosis in patients with ischemic stroke. MLR was the independent risk factor of carotid stenosis (OR: 9.74, 95% CI: 1.16–81.54). In the ROC curves analysis, a cutoff value of 0.20 for MLR predicted the severity of carotid stenosis with a sensitivity of 80.40% and specificity of 26.40% (ROC area under the curve: 0.598, 95% CI: 0.53–0.67, p = .004).ConclusionMonocyte/lymphocyte ratio plays important roles in carotid stenosis in patients with ischemic stroke and is an independent risk factor of the severity of carotid stenosis. Therefore, MLR might be considered a potential index in the diagnosis of carotid stenosis in patients with ischemic stroke.

Patient-Derived Organoids (PDO) As the Potential Model to Predict Treatment Outcome of Rectal Cancer Patients Underwent Neo-Adjuvant Chemoradiotherapy

Recent researches showed patient-derived organoids (PDO) could recapitulate main features of the primary tumors, and could be used as in vitro clinical model to predictive outcome in clinical setting. Here, we have launched the observational investigation to prove whether the PDO radiosensitivity could predict the clinical outcome in locally advanced rectal cancer (LARC) patients (pts) underwent neo-adjuvant chemoradiotherapy (nCRT) in locally advanced rectal cancer (LARC). Primary tumor tissues of LARC pts for PDO culture were collected by rectal biopsy before any treatment. Rectal cancer cells were isolated and cultured following established protocols for PDO culture and expansion. PDOs were exposed to 8Gy X-rays at 246 cGy/min, 5-fluorouracil (10μM) and CPT-11 (10μM), respectively, and PDO images were captured every 3 days in the following 24 days. PDO radiosensitivity and drugs sensitivity was determined by ROC area changes. Tumor response to neoadjuvant therapy was evaluated according to the pathological tumor regression grading (TRG). Finally, PDO radiosensitivity and drug sensitivity data compared with clinical patients outcome. We established 34 rectal adenocarcinoma organoids from 40 pts with 85% success rate. RCO data showed 23, 25 and 17 pts’ organoids were resistant, while 11, 9 and 13 pts’ organoids were sensitive to irradiation, 5-fluorouracil and CPT-11 treatment, respectively. All of the 34 pts received surgery (total mesorectal excision) and got the pathologic TRG data, or achieved clinical complete response (cCR). Clinical data showed 19 pts achieved good response (TRG=0 or 1, or cCR), and other 15 pts had poor response (TRG= 2 or 3). Through comparative analysis, RCO radiosensitivity and drug sensitivity (5-fluorouracil and CPT-11) were concordant with clinical outcome in 76.5% (26/34), 70.6% (24/34) and 70 % (21/30) pts, respectively. In consideration of combined treatment in clinic, chemoradiation responses in patients are highly matched to RCO responses, with 86.70 % sensitivity, 94.70% specificity, 92.90 % positive predict value, 90.0 % negative predict value and 91.20 % accuracy. The patient-derived organoids’ radiosensitivity correlates to clinical outcome and indicated that organoid model may be used to predict TRG and prognosis in LARC patients receiving nCRT. PDO may be potentially used as a biomarker to for treatment stratification and worth study further.

Temel Bileşen Analizi Yöntemleri Kullanarak Parkinson Hastalığının Otomatik Teşhisi

Parkinson's disease is a sneaky brain disorder that progresses very slowly. The diagnostic methods of this disease include the analysis of individual voices. The earliest detection of Parkinson's by voice analysis is made possible by various methods. In this study, the results obtained from the application of Tunable Q-factor Wavelet Transformation (TQWT) method to the recorded audio signals of 188 Parkinson's patients and 64 healthy individuals were used. The principal component analysis (PCA) and its types (kernel PCA (KPCA) and Probabilistic PCA (PPCA)) which are dimension reduction methods have been applied to the TQWT features. Afterwards, k-fold cross validation method was applied to the new data groups and the training-test data were obtained. In the next step, the data were separately classified by random forest (RF) algorithm to investigate the effectiveness of the dimension reduction methods and the results were also interpreted by statistical criteria. In terms of classification results, OTBA was the most successful in size reduction methods with 87.56% accuracy rate. In addition, as a result of this method, ROC and PRC area values reached a band of about 0.95, proving that patient and healthy class decomposition approached perfection. The performance results of this study, which is suitable for real-life applications, were compared with the single study in the literature in which the same data was used, and this study showed that higher statistical ratios were obtained in comparison to the other study. Moreover, the high number of people in whom the data was recorded compared to other studies in the literature increases the importance of the study in this field.

Malicious Intrusion Detection Using Machine Learning Schemes

Wireless networks are continuously facing challenges in the field of Information Security. This leads to major researches in the area of Intrusion detection. The working of Intrusion detection is performed mainly by signature based detection and anomaly based detection. Anomaly based detection is based on the behavior of the network. One of the major challenge in this domain is to identify and detect the malicious node in wireless networks. The intrusion detection mechanism has to analyse the behavior of the node in the network by means of the several features possessed by each node. Intelligent schemes are the need of the hour in such scenario. This paper has taken a standard dataset for studying the features of the wireless node and reduced the features by applying the most efficient Correlation Attribute feature selection method. The machine learning algorithms are applied to obtain an effective training model which is then applied on the testing dataset to validate the model. The accuracy of the model is determined by the performance parameters such as true positive rate, false positive rate and ROC area. Neural network, bagging and decision tree algorithm RepTree are giving promising results in comparison with other classification algorithms.

The effects of long-term exposure to silica dust on serum CC16 and KL-6 levels

Objective: To investigate the effects of long-term exposure to silica dust on serum CC16 and KL-6 levels. Methods: The patients with stage I silicosis who were hospitalized in our hospital from April 2016 to April 2017 were treated as silicosis group. The silica dust exposed workers without silicosis who were taken the physical examination in our hospital were taken as a dust-exposed group. The healthy control group comes from in the same period of community physical examination did not touch the dust. The levels of CC16 and KL-6 in serum of all subjects were determined by enzyme-linked immunosorbent assay (ELISA) , and the levels of CC16 and KL-6 in serum were compared in three groups. Results: Compared with the control group, the serum levels of CC16 in the silicosis group (P<0.01) and the dust-exposed group (P<0.01) were significantly lower. Compared with the control group, the level of serum KL-6 in the silicosis group was significantly decreased (P<0.01) compared with the control group, while the level of KL-6 in the serum of the dust-exposed group was significantly increased (P<0.01) . The ROC area of CC16 for diagnosis of silicosis was 0.92 (P<0.01) , with a sensitivity of 81.37%, specificity of 92.63% and Kappa value of 0.74. Conclusion: Long-term exposure to silica dust may lead to a decrease in serum CC16 levels. Reduced serum CC16 levels may be useful in identifying the diagnosis of silicosis.

Combination of novel and tradition biomarkers to enhance diagnostic sensitivity and specificity for early diagnosis of rheumatoid arthritis and seronegative rheumatoid arthritis

Background and objective: Rheumatoid arthritis is an autoimmune disease. This study investigated the potential value of combining cartilage oligometric matrix protein, anti-cyclic citrullinated protein, and 14-3-3 eta protein with traditional biomarkers to reduce the diagnostic gap. Methods: This case-control study included 46 male and female patients and 42 age- and gender-matched adults as control group. The biomarkers were measured using ELISA technique. Results: Tests for anti-cyclic citrullinated protein and cartilage oligometric matrix protein are excellent tools to diagnose rheumatoid arthritis because anti-cyclic citrullinated protein and cartilage oligometric matrix protein are associated with the highest ROC area. The validity of the test for 14-3-3 eta protein, which is a good test to predict rheumatoid arthritis, ranks second. The optimum cut-off values for high cartilage oligometric matrix protein, high anti-cyclic citrullinated protein, and high 14-3-3 eta protein were ≥0.242µg/L, ≥0.566ng/L, and ≥0.145ng/L, respectively. 14-3-3η protein, cartilage oligometric matrix protein status as a parallel combination which is considered as a wonderful combination in classification criteria for rheumatoid arthritis. Parallel combination, both criteria two tests are positive, namely “high cartilage oligometric matrix protein (≥0.242) + high 14-3-3η protein (≥0.145)” was associated with a perfect test, that the patients have rheumatoid arthritis (sensitivity 100%, specificity 100%, accuracy 100%, positive predictive value at pre-test probability 50% and 90% = 100%). A positive test using this combination is 100% diagnostic and establishes a possible diagnosis of rheumatoid arthritis with 100%, while a negative test would exclude a possible diagnosis of rheumatoid arthritis with 100% confidence. Conclusion: Results confirmed that high serum level of cartilage oligometric matrix protein, anti-cyclic citrullinated protein, and 14-3-3 eta protein are significantly associated with increased risk for rheumatoid arthritis, demonstrating the potential value of combining these new biomarkers with traditional biomarkers to enhance diagnostic sensitivity and specificity and ultimately reduce the diagnostic gap.

Natural killer cell activity and prostate cancer risk in veteran men undergoing prostate biopsy

BackgroundA previous pilot study found that men with a positive prostate biopsy had low numbers of circulating natural killer (NK) cells, compared to biopsy negative men. MethodsTo confirm these data, we analyzed differences in NK cells from 94 men undergoing prostate biopsy to determine whether NK cells could predict for a positive biopsy. NK cells activity (NKA) was measured by an in vitro diagnostic system, with a pre-defined cut-off value for NKA at 200 pg/mL. Logistic regression and receiver operator characteristics (Area Under the Curve (AUC)) analyses were used to test the diagnostic value of NKA. ResultsThe NKA test performance showed specificity of 88%, positive predictive value of 84%, sensitivity of 34%, and a negative predictive value of 41%. Among the 94 men analyzed, NKA was not significantly linked with age, race, digital rectal examination (DRE), prostate volume, PSA or biopsy grade group (all P ≥ 0.14). In multivariable logistic regression analysis, the odds ratio (OR) of low NKA (<200 pg/mL) for the detection of PC was 4.89, 95%CI 1.34–17.8, with a ROC area under the curve of 0.79 in all participants and increasing to 0.83 and 0.85 for the detection of PC and high-grade PC, respectively, among men with a normal DRE. ConclusionsMen with a low NKA value had five-times higher odds of PC at biopsy. The implementation of this NKA assay in the clinic together with PSA may help to advise patients with the highest risk of PC whether, or not, to undergo a prostate biopsy.

Bedside Ultrasound in Fluid Assessment: Correlation of Internal Jugular and Subclavian Vein Diameter Variation with Invasive Right Heart Catheterization

Introduction Accurate estimation of fluid status is of paramount importance in patients with heart failure. Frequent invasive right heart catheterization is not feasible and accurate inferior vena cava imaging requires training. Internal jugular vein (IJV) and subclavian vein (SCV) could provide easier alternatives. Methods Patients scheduled to undergo right heart catheterization were prospectively enrolled in the study. Body surface area (BSA) was noted. Patient was positioned flat supine. IJV was imaged at the apex of the right sternocleidomastoid muscle and SCV was imaged at the lateral third of the right clavicle. Using M-mode on a portable ultrasound machine, the maximum (Dmax) and minimum (Dmin) anteroposterior diameters were noted during normal breathing, without applying any external pressure. Respiratory collapsibility was assessed with sniff. Respiratory variation in diameter (RVD) was calculated as: [(Dmax - Dmin)/ Dmax] and expressed as percent. Diameter variation index (DVI) was calculated as: (Dmax - Dmin)/ BSA. Results Total 72 patients enrolled with mean age 61 ± 14 years, mean BSA 2 ± 0.3m2, left ventricular (LV) ejection fraction 45 ± 20%. SCV could not be imaged in 7 patients. Elevated right atrial [RA] pressure (>=10mmHg) patients showed less RVD with resting respiration (14 vs. 41% for IJV and 24% vs. 46% for SCV, p= 0.01) and less likely to show total collapsibility with sniff (table 1). Similarly, DVI was lower with high RA pressure. Dmax alone did not correlate for SCV (table 1). For RA pressure >=10mmHg, lack of IJV collapsibility with sniff had a sensitivity of 84% and ROC area 0.75. Dmax >1cm and RVD =10mmHg. Conclusion There is a positive correlation between fluid status with IJV and SCV diameters (absolute and respiratory variation). The diagnostic performance of these veins was similar to inferior vena cava, but easier to perform.

Random plasma glucose predicts the diagnosis of diabetes.

Aims/HypothesisEarly recognition of those at high risk for diabetes as well as diabetes itself can permit preventive management, but many Americans with diabetes are undiagnosed. We sought to determine whether routinely available outpatient random plasma glucose (RPG) would be useful to facilitate the diagnosis of diabetes.MethodsRetrospective cohort study of 942,446 U.S. Veterans without diagnosed diabetes, ≥3 RPG in a baseline year, and ≥1 primary care visit/year during 5-year follow-up. The primary outcome was incident diabetes (defined by diagnostic codes and outpatient prescription of a diabetes drug).ResultsOver 5 years, 94,599 were diagnosed with diabetes [DIAB] while 847,847 were not [NONDIAB]. Baseline demographics of DIAB and NONDIAB were clinically similar, except DIAB had higher BMI (32 vs. 28 kg/m2) and RPG (150 vs. 107 mg/dl), and were more likely to have Black race (18% vs. 15%), all p<0.001. ROC area for prediction of DIAB diagnosis within 1 year by demographic factors was 0.701, and 0.708 with addition of SBP, non-HDL cholesterol, and smoking. These were significantly less than that for prediction by baseline RPG alone (≥2 RPGs at/above a given level, ROC 0.878, p<0.001), which improved slightly when other factors were added (ROC 0.900, p<0.001). Having ≥2 RPGs ≥115 mg/dl had specificity 77% and sensitivity 87%, and ≥2 RPGs ≥130 mg/dl had specificity 93% and sensitivity 59%. For predicting diagnosis within 3 and 5 years by RPG alone, ROC was reduced but remained substantial (ROC 0.839 and 0.803, respectively).ConclusionsRPG levels below the diabetes “diagnostic” range (≥200 mg/dl) provide good discrimination for follow-up diagnosis. Use of such levels–obtained opportunistically, during outpatient visits–could signal the need for further testing, allow preventive intervention in high risk individuals before onset of disease, and lead to earlier identification of diabetes.

Serum heart-type fatty acid-binding protein as a predictor for the development of sepsis-associated acute kidney injury

ABSTRACTBackground: We analyzed the correlation between heart-type fatty acid-binding protein (HFABP) and the development of acute kidney injury (AKI) in patients with sepsis and estimated the predictive capacity of HFABP for sepsis-associated acute kidney injury (SAKI).Methods: In this retrospective observational study, we screened 2,452 patients who received the HFABP test in the emergency department. 442 admitted patients with sepsis were finally enrolled. Based on the diagnostic criteria for AKI in Kidney Disease: Improving Global Outcomes (KDIGO) guidelines, patients were divided into the no-AKI group (n = 317) and AKI group (n = 125). We analyzed the correlation between HFABP and SAKI occurrence by logistic regression analysis and evaluated the predictive ability of HFABP to SAKI using c-index, net reclassification improvement index (NRI) and integrated discrimination improvement index (IDI).Results: Patients in the AKI group with significantly higher the level of HFABP and in-hospital mortality. HFABP concentration is an independent risk factor for SAKI (OR: 11.398; 95% CI: 6.218–20.891, P < 0.001), but not for in-hospital mortality (OR: 1.189, 95%CI: 0.954–2.607, P = 0.076). The addition of HFABP to the prediction model significantly improved the ROC area (0.867 vs 0.755, P < 0.001), NRI 25.03% (95% CI 9.72–38.51%) and IDI 14.33 (95% CI 11.04–17.62).Conclusion: Serum HFABP is correlated with SAKI development and could become a potential predictive biomarker.

The Malawi trauma score: A model for predicting trauma-associated mortality in a resource-poor setting

BackgroundGlobally, traumatic injury is a leading cause of morbidity and mortality in low-income countries. Current tools for predicting trauma-associated mortality are often not applicable in low-resource environments due to a lack of diagnostic adjuncts. This study sought to derive and validate a model for predicting mortality that requires only a history and physical exam. MethodsWe conducted a retrospective analysis of all patients recorded in the Kamuzu Central Hospital trauma surveillance registry in Lilongwe, Malawi from 2011 through 2014. Using statistical randomization, 80% of patients were used for derivation and 20% were used for validation. Logistic regression modeling was used to derive factors associated with mortality and the Malawi Trauma Score (MTS) was constructed. The model fitness was tested. Results62,354 patients are included. Patients are young (mean age 23.0, SD 15.9 years) with a male preponderance (72%). Overall mortality is 1.8%. The MTS is tabulated based on initial mental status (alert, responds to voice, responds only to pain or worse), anatomical location of the most severe injury, the presence or absence of a radial pulse on examination, age, and sex. The score range is 2–32. A mental status exam of only responding to pain or worse, head injury, the absence of a radial pulse, extremes of age, and male sex all conferred a higher probability of mortality. The ROC area under the curve for the derivation cohort and validation cohort were 0.83 (95% CI 0.78, 0.87) and 0.83 (95% CI 0.75, 0.92), respectively. A MTS of 25 confers a 50% probability of death. ConclusionsThe MTS provides a reliable tool for trauma triage in sub-Saharan Africa and helps risk stratify patient populations. Unlike other models previously developed, its strength is its utility in virtually any environment, while reliably predicting injury- associated mortality.

Machine learning classifiers can predict Gleason pattern 4 prostate cancer with greater accuracy than experienced radiologists

ObjectiveThe purpose of this study was: To test whether machine learning classifiers for transition zone (TZ) and peripheral zone (PZ) can correctly classify prostate tumors into those with/without a Gleason 4 component, and to compare the performance of the best performing classifiers against the opinion of three board-certified radiologists.MethodsA retrospective analysis of prospectively acquired data was performed at a single center between 2012 and 2015. Inclusion criteria were (i) 3-T mp-MRI compliant with international guidelines, (ii) Likert ≥ 3/5 lesion, (iii) transperineal template ± targeted index lesion biopsy confirming cancer ≥ Gleason 3 + 3. Index lesions from 164 men were analyzed (119 PZ, 45 TZ). Quantitative MRI and clinical features were used and zone-specific machine learning classifiers were constructed. Models were validated using a fivefold cross-validation and a temporally separated patient cohort. Classifier performance was compared against the opinion of three board-certified radiologists.ResultsThe best PZ classifier trained with prostate-specific antigen density, apparent diffusion coefficient (ADC), and maximum enhancement (ME) on DCE-MRI obtained a ROC area under the curve (AUC) of 0.83 following fivefold cross-validation. Diagnostic sensitivity at 50% threshold of specificity was higher for the best PZ model (0.93) when compared with the mean sensitivity of the three radiologists (0.72). The best TZ model used ADC and ME to obtain an AUC of 0.75 following fivefold cross-validation. This achieved higher diagnostic sensitivity at 50% threshold of specificity (0.88) than the mean sensitivity of the three radiologists (0.82).ConclusionsMachine learning classifiers predict Gleason pattern 4 in prostate tumors better than radiologists.Key Points• Predictive models developed from quantitative multiparametric magnetic resonance imaging regarding the characterization of prostate cancer grade should be zone-specific.• Classifiers trained differently for peripheral and transition zone can predict a Gleason 4 component with a higher performance than the subjective opinion of experienced radiologists.• Classifiers would be particularly useful in the context of active surveillance, whereby decisions regarding whether to biopsy are necessitated.

Novel Prognostic Markers in Triple-Negative Breast Cancer Discovered by MALDI-Mass Spectrometry Imaging

There are no widely-accepted prognostic markers currently available to predict outcomes in patients with triple-negative breast cancer (TNBC), and no targeted therapies with confirmed benefit. We have used MALDI mass spectrometry imaging (MSI) of tryptic peptides to compare regions of cancer and benign tissue in 10 formalin-fixed, paraffin-embedded sections of TNBC tumors. Proteins were identified by reference to a peptide library constructed by LC-MALDI-MS/MS analyses of the same tissues. The prognostic significance of proteins that distinguished between cancer and benign regions was estimated by Kaplan-Meier analysis of their gene expression from public databases. Among peptides that distinguished between cancer and benign tissue in at least 3 tissues with a ROC area under the curve >0.7, 14 represented proteins identified from the reference library, including proteins not previously associated with breast cancer. Initial network analysis using the STRING database showed no obvious functional relationships except among collagen subunits COL1A1, COL1A2, and COL63A, but manual curation, including the addition of EGFR to the analysis, revealed a unique network connecting 10 of the 14 proteins. Kaplan-Meier survival analysis to examine the relationship between tumor expression of genes encoding the 14 proteins, and recurrence-free survival (RFS) in patients with basal-like TNBC showed that, compared to low expression, high expression of nine of the genes was associated with significantly worse RFS, most with hazard ratios >2. In contrast, in estrogen receptor-positive tumors, high expression of these genes showed only low, or no, association with worse RFS. These proteins are proposed as putative markers of RFS in TNBC, and some may also be considered as possible targets for future therapies.

Label-free serum proteomics and multivariate data analysis identifies biomarkers and expression trends that differentiate Intraductal papillary mucinous neoplasia from pancreatic adenocarcinoma and healthy controls

BackgroundIntraductal Papillary Mucinous Neoplasia (IPMN) are potentially malignant cystic tumors of the pancreas. IPMN can progress from low to moderate to high grade dysplasia and further to IPMN associated carcinoma. Often the difference between benign and malignant nature of the IPMN is not clear preoperatively. We aim to elucidate molecular expression patterns of various grades of IPMN and pancreatic carcinoma. Additionally we suggest potential novel biomarkers to differentiate IPMN from healthy individuals and pancreatic carcinoma to enable early detection as well as help in differential diagnosis in future.MethodsWe have performed retrospective label-free proteomic analysis of the serum samples from 44 patients with various grades of benign IPMN or IPMN associated carcinoma and 11 healthy controls. Proteomic data was further analyzed by various multivariate statistical methods. Four groups of samples (low-grade, high-grade IPMN, pancreatic carcinoma and age- and sex-matched healthy controls) were compared with ANOVA. Orthogonal projections to latent structures-discriminant analysis (OPLS-DA) modeling gave S-plot for feature selection. Stringently selected potential markers were further evaluated with ROC curve analysis and area under the curve was calculated. Differentially expressed proteins were used for pathway analysis. Linear trend analysis (Mann Kendall test) was used for identifying significant increasing or decreasing trends from healthy-low grade-high grade IPMN-pancreatic carcinoma.ResultsBased on protein expression (436 proteins quantified), PCA separated most sample groups from each other. S-Plot selected biomarker panels with moderate to very high AUC values for differentiating controls from Low-, High-Grade IPMN and carcinoma. Linear trend analysis identified 12 proteins which were consistently increasing or decreasing trend among the groups. We found potential biomarkers to differentiate healthy controls from different degrees of dysplasia and pancreatic carcinoma. These biomarkers can classify IPMN, carcinoma and healthy controls from each other which is an unmet clinical need. Data are available via ProteomeXchange with identifier PXD009139.ConclusionKininogen-1 was able to differentiate healthy persons from low and high-grade IPMN. Retinol binding protein-4 could classify the low-grade IPMN from pancreatic carcinoma. Twelve proteins including apolipoproteins and complement proteins had significantly increasing or decreasing trends from healthy to low to high-grade IPMN to pancreatic carcinoma.

A neural network approach to classify carotid disorders from Heart Rate Variability analysis

BackgroundAtherosclerosis is a progressive process responsible for most heart diseases and ischemic stroke. It constitutes, in fact, the most common cause of stroke in middle-aged people. To avoid or, at least, limit the disabling deficits that may derive from a carotid disease, a prompt and early diagnosis is necessary. The diagnostic technique used to detect a carotid disease is the eco-color Doppler. Unfortunately, this method is not free from errors, due to manufacturer mistakes or its operator dependence. MethodsIn this study, we propose an automated methodology capable of identifying the presence of a carotid disease from the Heart Rate Variability analysis of electrocardiographic signals. A Correlation-based Feature Selector for data reduction and Artificial Neural Networks are used to distinguish between pathological and healthy subjects. ResultsA series of tests has been realized to evaluate the proposed approach by using electrocardiographic signals selected from an available database in order to analyse the classification ability in comparison with other algorithms existing in literature. The results obtained show that the proposed approach provides values of accuracy, sensitivity, specificity, precision, F-measure and ROC area, respectively equal to 90.5%, 97.7%, 72.9%, 89.7%, 93.5% and 0.957, better than those achieved by other algorithms. ConclusionsConsidering the achieved accuracy, our methodology is more effective than any of the main algorithm existing in literature. It is important to note that this approach is proposed as a support for the diagnosis of a carotid disorder through a non-invasive approach.