scholarly journals Early Detection of Sepsis using Machine Learning

2020 ◽  
Vol 9 (1) ◽  
pp. 489-493
Keyword(s):  
Cross Validation ◽  
Splitting Method ◽  
Recursive Feature Elimination ◽  
Or Efficiency ◽  
Clinical Sepsis ◽  
Sepsis Diagnosis ◽  
Life Saving ◽  
Pre Treatment ◽  
Logistic Regression Algorithm ◽  
Normal Immune Response

Sepsis is a global cause of the death due to infection and subsequent overreaction of the immune system. Mortality rates are highest in developed and developing countries in cases of septic shock. Sepsis is a clinical condition with an emergency referral that can be avoided by advance warning. SIRS is a normal immune response to any infection, and hospitals or antibiotics are not required by most people. Early sepsis prediction is possibly life-saving, and we are planning to predict sepsis 6 hours before clinical sepsis diagnosis. We used two types of Standard Scalar and Min Max Scalar pre-treatment methods to pre-treat the data. The Recursive Feature Elimination (RFE) method is used to pick the features most closely related to the predictive or efficiency factor. The ML algorithms used are the Logistic Regression algorithm and XGboost. Cross-validation 10 times with the train splitting method is used to validate attributes. we selected the best model from these two trained models.

scholarly journals NIMG-37. PREDICTING SEIZURE IN GLIOMA PATIENTS USING A RANDOM FOREST CLASSIFIER TRAINED ON SEX-SPECIFIC AND MIXED COHORTS

2019 ◽  
Vol 21 (Supplement_6) ◽  
pp. vi169-vi169
Author(s):  
Aditya Khurana ◽  
Sandra Johnston ◽  
Paula Whitmire ◽  
Sara Ranjbar ◽  
Akanksha Sharma ◽  
...  
Keyword(s):  
Random Forest ◽  
Mixed Model ◽  
Cross Validation ◽  
Class Imbalance ◽  
Random Forest Classifier ◽  
Co Morbidity ◽  
Pre Treatment ◽  
Male Model ◽  
Sensitivity Specificity ◽  
Testing Set

Abstract PURPOSE Brain tumor related epilepsy (BTE) is a major co-morbidity in patients with glioma. It is difficult to determine whether the use of anti-epileptic drugs is necessary. We attempted to build a machine-learning model to predict the probability of seizure presentation (SP) with glioma. METHODS We trained a random forest classifier using the following variables: volumetric data of pre-treatment MR images (T1Gd and T2-FLAIR sequences), patient demographics (age; sex), and measurements of tumor proliferation (log(ρ)), invasiveness (log(D)) and their relative ratio (log(ρ/D)). Our cohort consisted of 221 patients total. Using an 80-20 ratio, we used 176 patients (76 SP, 100 nSP) for training and the remaining 45 patients (19 SP, 26 nSP) were used for testing. We also trained on male-only and female-only cohorts to evaluate any sex differences in prediction. For training, 108 males (53 SP, 55 nSP) were used and 28 for testing (14 SP, 14 nSP). We used 72 females (21 SP, 49 nSP) for training and 15 (7 SP, 8 nSP) for testing. We corrected for class imbalance in the female cohort before training. Using 10-fold cross-validation and a separate testing set, we measured performance by ROC curve (AUC), accuracy, sensitivity, and specificity of predictions (average of folds in cross validation). RESULTS The female model achieved the highest AUC (0.853) followed by the mixed model (0.726) and the male model (0.651). In the validation set, the accuracy/sensitivity/specificity of the three cohorts were as follows: mixed (0.726/0.696/0.750), female (0.853/0.830/0.875), and male (0.651/0.577/0.722). The performance of the testing set, in terms of accuracy/sensitivity/specificity were: mixed (0.733/0.74/0.73), female (0.8/0.57/1), and male (0.714/0.64/0.79). CONCLUSION We found a negative correlation between seizure probability and size and invasiveness of tumors. Our model shows promising performance on testing set data. Further cohort studies and training is warranted.

scholarly journals Rational design of a nanoparticle platform for oral prophylactic immunotherapy to prevent immunogenicity of therapeutic proteins

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Nhan H. Nguyen ◽  
Fiona Y. Glassman ◽  
Robert K. Dingman ◽  
Gautam N. Shenoy ◽  
Elizabeth A. Wohlfert ◽  
...  
Keyword(s):  
Immune Response ◽  
Rational Design ◽  
Therapeutic Proteins ◽  
Cost Effective ◽  
Therapeutic Protein ◽  
Safety And Efficacy ◽  
Life Saving ◽  
Pre Treatment ◽  
Design And Testing

AbstractThe safety and efficacy of several life-saving therapeutic proteins are compromised due to their immunogenicity. Once a sustained immune response against a protein-based therapy is established, clinical options that are safe and cost-effective become limited. Prevention of immunogenicity of therapeutic proteins prior to their initial use is critical as it is often difficult to reverse an established immune response. Here, we discuss a rational design and testing of a phosphatidylserine-containing nanoparticle platform for novel oral prophylactic reverse vaccination approach, i.e., pre-treatment of a therapeutic protein in the presence of nanoparticles to prevent immunogenicity of protein therapies.

An Objective Trial of Alternative Methods of Heparin Administration

1979 ◽  
Author(s):  
P.G. Bentley ◽  
V.P. Ward ◽  
V.V. Kakkar
Keyword(s):  
Infusion Pump ◽  
Alternative Methods ◽  
Constant Infusion ◽  
Bleeding Complications ◽  
Vein Thrombosis ◽  
Heparin Administration ◽  
Life Saving ◽  
The Difference ◽  
Pre Treatment ◽  
Exact Size

Although it ts generally agreed that heparin is truly a life saving drug, the best method of administration remains to be determined. 100 patients with calf vein thrombosis were randomly allocated to receive either subcutaneous (SC) or intravenous (IV) heparin for seven days. Venography was performed in each patient to confirm the exact size and site of thrombus and was repeated at the end of heparin treatment. IV heparin was administered using a constant infusion pump. Platelet count, haemoglobin and haematocrit estimations were performed at the beginning, mid point, and end of hep rin treatment. The dose of heparin to be administered was determined by dally estimation of KCCT. In SC group, thrombi increased in size in 2%. remained unchanged in 50% and decreased in 48%. In IV group, they increased in size in 20%, were unchanged in 62% and decreased in 18%. The difference in decreases in size was significant (P < 0.01). Pre-treatment lung scans corfirmed the presence of emboli in 10 patients in SC and 11 in IV group; post treatmel scan showed improvement in 60% in SC compared to 91% in IV group. The mean daily heparin requirement (±SD) was 36,998 ± 7. 12IU in the SC group and 36,814 ± 6.63 IU in the IV group. The average KCCT fluctuation index in the SC group was 59.08 seconds compared to 64.4 seconds in the IV group. Major bleeding complications occurred in 2 patients on SC heparin and in 4 treated with IV heprin. The clinical signficance of these will be discussed

scholarly journals Magnetic Bead-Based Electrochemical Immunoassays On-Drop and On-Chip for Procalcitonin Determination: Disposable Tools for Clinical Sepsis Diagnosis

Biosensors ◽  
2020 ◽  
Vol 10 (6) ◽  
pp. 66
Author(s):  
Águeda Molinero-Fernández ◽  
María Moreno-Guzmán ◽  
Miguel Ángel López ◽  
Alberto Escarpa
Keyword(s):  
Limit Of Detection ◽  
Magnetic Bead ◽  
Clinical Analysis ◽  
Good Precision ◽  
Microfluidic Chips ◽  
Reliable Determination ◽  
Clinical Sepsis ◽  
Sepsis Diagnosis ◽  
On Chip ◽  
Electrochemical Immunoassays

Procalcitonin (PCT) is a known protein biomarker clinically used for the early stages of sepsis diagnosis and therapy guidance. For its reliable determination, sandwich format magnetic bead-based immunoassays with two different electrochemical detection approaches are described: (i) disposable screen-printed carbon electrodes (SPE-C, on-drop detection); (ii) electro-kinetically driven microfluidic chips with integrated Au electrodes (EMC-Au, on-chip detection). Both approaches exhibited enough sensitivity (limit of detection (LOD) of 0.1 and 0.04 ng mL−1 for SPE-C and EMC-Au, respectively; cutoff 0.5 ng mL−1), an adequate working range for the clinically relevant concentrations (0.5–1000 and 0.1–20 ng mL−1 for SPE-C and EMC-Au, respectively), and good precision (RSD < 9%), using low sample volumes (25 µL) with total assay times less than 20 min. The suitability of both approaches was successfully demonstrated by the analysis of human serum and plasma samples, for which good recoveries were obtained (89–120%). Furthermore, the EMC-Au approach enabled the easy automation of the process, constituting a reliable alternative diagnostic tool for on-site/bed-site clinical analysis.

scholarly journals The Proportion for Splitting Data into Training and Test Set for the Bootstrap in Classification Problems

2021 ◽  
Vol 12 (1) ◽  
pp. 228-242
Author(s):  
Borislava Vrigazova
Keyword(s):  
Logistic Regression ◽  
Decision Tree ◽  
Cross Validation ◽  
Computing Time ◽  
Splitting Method ◽  
Classification Methods ◽  
Classification Problems ◽  
Test Set ◽  
Resampling Methods ◽  
Tenfold Cross Validation

Abstract Background: The bootstrap can be alternative to cross-validation as a training/test set splitting method since it minimizes the computing time in classification problems in comparison to the tenfold cross-validation. Objectives: Тhis research investigates what proportion should be used to split the dataset into the training and the testing set so that the bootstrap might be competitive in terms of accuracy to other resampling methods. Methods/Approach: Different train/test split proportions are used with the following resampling methods: the bootstrap, the leave-one-out cross-validation, the tenfold cross-validation, and the random repeated train/test split to test their performance on several classification methods. The classification methods used include the logistic regression, the decision tree, and the k-nearest neighbours. Results: The findings suggest that using a different structure of the test set (e.g. 30/70, 20/80) can further optimize the performance of the bootstrap when applied to the logistic regression and the decision tree. For the k-nearest neighbour, the tenfold cross-validation with a 70/30 train/test splitting ratio is recommended. Conclusions: Depending on the characteristics and the preliminary transformations of the variables, the bootstrap can improve the accuracy of the classification problem.

scholarly journals DIAGNOSTIC UTILITY OF PRESEPSIN AND PROCALCITONIN IN CRITICALLY ILL PATIENTS WITH SUSPECTED SEPSIS: COMPARISON WITH RECENT CLINICAL CRITERIA OF SEPSIS-3

2019 ◽  
Vol 3 (6) ◽  
Author(s):  
Babak Alikiaie ◽  
Sarah Mousavi ◽  
Mohammad Nasri
Keyword(s):  
Critically Ill ◽  
Critically Ill Patients ◽  
Diagnostic Value ◽  
Suspected Sepsis ◽  
Clinical Sepsis ◽  
Clinical Criteria ◽  
Decision Tools ◽  
Sepsis Diagnosis ◽  
Previous Definition ◽  
Definition Of

Background: In this study, we aimed to investigate the diagnostic and prognostic utilities of presepsin and procalcitonin (PCT) in critically ill patients with suspected sepsis, for whom sepsis was diagnosed clinically based on the Survival Sepsis Campaign (SSC) criteria and to compare it with recent criteria of Sepsis-3. Methods: Blood samples for biomarker measurements of presepsin and PCT were drawn on days 1, 3 and 7 of ICU admission in a total of 26 patients. All patients were followed-up until death or discharge. All studied biomarkers were analyzed according to the diagnosis and severity of sepsis and for prognosis (all-cause mortality) at days 1, 3 and 7. Agreement between the diagnosis of clinical sepsis and presepsin or PCT-based sepsis was assessed using Cohen’s kappa Results: Clinical sepsis (based on Sepsis-3) and presepsin or PCT-based sepsis showed poor agreement (Kappa<0.4). Presepsin levels at day1 correlated significantly with mortality (r=0.45, P; 0.02). The diagnostic value of both presepsin and PCT to diagnose sepsis was weak (Area under curve (AUC) <0.75). The overall agreement in sepsis diagnosis was fair to good based on the both clinical criteria (P<0.05, Kappa: 0.5-0.75). More than 80% of patients (N=21) had sepsis based on presepsin upon admission. Both clinical criteria predicated that less than 20% of patients (N=5) had sepsis upon admission. Conclusion: Based on our findings, the overall agreement between the diagnosis of clinical sepsis and presepsin or PCT-based sepsis was poor. Also, our results show that the new Sepsis-3 definitions were accurate and equal to the previous definition of SSC guideline. Although, availability of diagnostic assays is variable in Iran, but, it seems that addition of developing decision tools that utilize biomarkers to help aid the rapid diagnosis of sepsis is necessary and may  improve patient outcomes. Keywords: Presepsin, Pro Calcitonin, Sepsis, Diagnosis

scholarly journals A Machine-Learning-Based Prediction Method for Hypertension Outcomes Based on Medical Data

Diagnostics ◽  
2019 ◽  
Vol 9 (4) ◽  
pp. 178 ◽  
Author(s):  
Wenbing Chang ◽  
Yinglai Liu ◽  
Yiyong Xiao ◽  
Xinglong Yuan ◽  
Xingxing Xu ◽  
...  
Keyword(s):  
Physical Examination ◽  
Operating Characteristic ◽  
Cross Validation ◽  
Characteristic Curve ◽  
Prediction Method ◽  
Prediction Performance ◽  
Recursive Feature Elimination ◽  
Second Step ◽  
Operating Characteristic Curve ◽  
Key Features

The outcomes of hypertension refer to the death or serious complications (such as myocardial infarction or stroke) that may occur in patients with hypertension. The outcomes of hypertension are very concerning for patients and doctors, and are ideally avoided. However, there is no satisfactory method for predicting the outcomes of hypertension. Therefore, this paper proposes a prediction method for outcomes based on physical examination indicators of hypertension patients. In this work, we divide the patients’ outcome prediction into two steps. The first step is to extract the key features from the patients’ many physical examination indicators. The second step is to use the key features extracted from the first step to predict the patients’ outcomes. To this end, we propose a model combining recursive feature elimination with a cross-validation method and classification algorithm. In the first step, we use the recursive feature elimination algorithm to rank the importance of all features, and then extract the optimal features subset using cross-validation. In the second step, we use four classification algorithms (support vector machine (SVM), C4.5 decision tree, random forest (RF), and extreme gradient boosting (XGBoost)) to accurately predict patient outcomes by using their optimal features subset. The selected model prediction performance evaluation metrics are accuracy, F1 measure, and area under receiver operating characteristic curve. The 10-fold cross-validation shows that C4.5, RF, and XGBoost can achieve very good prediction results with a small number of features, and the classifier after recursive feature elimination with cross-validation feature selection has better prediction performance. Among the four classifiers, XGBoost has the best prediction performance, and its accuracy, F1, and area under receiver operating characteristic curve (AUC) values are 94.36%, 0.875, and 0.927, respectively, using the optimal features subset. This article’s prediction of hypertension outcomes contributes to the in-depth study of hypertension complications and has strong practical significance.

scholarly journals Serial measurement of pancreatic stone protein for the early detection of sepsis in intensive care unit patients: a prospective multicentric study

Critical Care ◽  
2021 ◽  
Vol 25 (1) ◽  
Author(s):  
Jérôme Pugin ◽  
Thomas Daix ◽  
Jean-Luc Pagani ◽  
Davide Morri ◽  
Angelo Giacomucci ◽  
...  
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Early Recognition ◽  
Clinical Signs ◽  
Pancreatic Stone Protein ◽  
Multicentric Study ◽  
Pancreatic Stone ◽  
Clinical Sepsis ◽  
Sepsis Diagnosis ◽  
Nosocomial Sepsis

Abstract Background The early recognition and management of sepsis improves outcomes. Biomarkers may help in identifying earlier sub-clinical signs of sepsis. We explored the potential of serial measurements of C-reactive protein (CRP), procalcitonin (PCT) and pancreatic stone protein (PSP) for the early recognition of sepsis in patients hospitalized in the intensive care unit (ICU). Methods This was a multicentric international prospective observational clinical study conducted in 14 ICUs in France, Switzerland, Italy, and the United Kingdom. Adult ICU patients at risk of nosocomial sepsis were included. A biomarker-blinded adjudication committee identified sepsis events and the days on which they began. The association of clinical sepsis diagnoses with the trajectories of PSP, CRP, and PCT in the 3 days preceding these diagnoses of sepsis were tested for markers of early sepsis detection. The performance of the biomarkers in sepsis diagnosis was assessed by receiver operating characteristic (ROC) analysis. Results Of the 243 patients included, 53 developed nosocomial sepsis after a median of 6 days (interquartile range, 3–8 days). Clinical sepsis diagnosis was associated with an increase in biomarkers value over the 3 days preceding this diagnosis [PSP (p = 0.003), PCT (p = 0.025) and CRP (p = 0.009)]. PSP started to increase 5 days before the clinical diagnosis of sepsis, PCT 3 and CRP 2 days, respectively. The area under the ROC curve at the time of clinical sepsis was similar for all markers (PSP, 0.75; CRP, 0.77; PCT, 0.75). Conclusions While the diagnostic accuracy of PSP, CRP and PCT for sepsis were similar in this cohort, serial PSP measurement demonstrated an increase of this marker the days preceding the onset of signs necessary to clinical diagnose sepsis. This observation justifies further evaluation of the potential clinical benefit of serial PSP measurement in the management of critically ill patients developing nosocomial sepsis. Trial registration The study has been registered at ClinicalTrials.gov (no. NCT03474809), on March 16, 2018. https://www.clinicaltrials.gov/ct2/show/NCT03474809?term=NCT03474809&draw=2&rank=1.

scholarly journals Calcium ion chelation preserves platelet function during cold storage

2020 ◽  
Author(s):  
Binggang Xiang ◽  
Guoying Zhang ◽  
Yan Zhang ◽  
Congqing Wu ◽  
Smita Joshi ◽  
...  
Keyword(s):  
Platelet Function ◽  
Cold Storage ◽  
Calcium Ion ◽  
Platelet Transfusion ◽  
Underlying Mechanism ◽  
Integrin Activation ◽  
Life Saving ◽  
A Cell ◽  
Rapid Clearance ◽  
Pre Treatment

AbstractObjectivePlatelet transfusion is a life-saving therapy to prevent or treat bleeding in patients with thrombocytopenia or platelet dysfunction. However, for more than six decades, safe and effective strategies for platelet storage have been an impediment to widespread use of platelet transfusion. Refrigerated platelets are cleared rapidly from circulation, precluding cold storage of platelets for transfusion. Consequently, platelets are stored at room temperature (RT) with an upper limit of 5 days due to risks of bacterial contamination and loss of platelet function. This practice severely limits platelet availability for transfusion. This study is to identify the mechanism of platelet clearance after cold storage and develop a method for platelet cold storage.Approach and ResultsWe found that rapid clearance of cold-stored platelets was largely due to integrin activation and apoptosis. Deficiency of integrin β3 or caspase-3 prolonged cold-stored platelets in circulation. Pre-treatment of platelets with EGTA, a cell impermeable calcium ion chelator, reversely inhibited cold storage-induced platelet activation and consequently prolonged circulation of cold-stored platelets. Moreover, transfusion of EGTA-treated, cold-stored platelets, but not RT-stored platelets, into the mice deficient in glycoprotein Ibα significantly shortened tail-bleeding times and diminished blood loss.ConclusionIntegrin activation and apoptosis is the underlying mechanism of rapid clearance of platelets after cold storage. Addition of a cell impermeable calcium ion chelator to platelet products is potentially a simple and effective method to enable cold storage of platelets for transfusion.

scholarly journals Serial Measurement of Pancreatic Stone Protein for the Early Detection of Sepsis in Intensive Care Unit Patients: A Prospective Multicentric Study

2021 ◽  
Author(s):  
Jérôme Pugin ◽  
Thomas Daix ◽  
Jean-Luc Pagani ◽  
Davide Morri ◽  
Angelo Giacomucci ◽  
...  
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Early Recognition ◽  
Clinical Signs ◽  
Pancreatic Stone Protein ◽  
Multicentric Study ◽  
Pancreatic Stone ◽  
Clinical Sepsis ◽  
Sepsis Diagnosis ◽  
Nosocomial Sepsis

Abstract BackgroundThe early recognition and management of sepsis improves outcomes. Biomarkers may help in identifying earlier sub-clinical signs of sepsis. We explored the potential of serial measurements of C-reactive protein (CRP), procalcitonin (PCT) and pancreatic stone protein (PSP) for the early recognition of sepsis in patients hospitalized in the intensive care unit (ICU).MethodsThis was a multicentric international prospective observational clinical study conducted in 14 ICUs in France, Switzerland, Italy, and the United Kingdom. Adult ICU patients at risk of nosocomial sepsis were included. A biomarker-blinded adjudication committee identified sepsis events and the days on which they began. The associations of clinical sepsis diagnoses with the trajectories of PSP, CRP, and PCT in the 3 days preceding these diagnoses of sepsis were tested for markers of early sepsis detection. The performance of the biomarkers in sepsis diagnosis was assessed by receiver operating characteristic (ROC) analysis.ResultsOf the 243 patients included, 53 developed nosocomial sepsis after a median of 6 days (interquartile range, 3–8 days). The association of clinical sepsis diagnosis with an increase in a biomarker value in the 3 days preceding this diagnosis was stronger for PSP (p = 0.003) than for PCT (p = 0.025) and CRP (p = 0.009). The area under the ROC curve at the time of clinical sepsis was similar for all markers (PSP, 0.75; CRP, 0.77; PCT, 0.75).ConclusionsWhile the diagnostic accuracy for sepsis of PSP, CRP and PCT were similar in this cohort, serial PSP measurement demonstrated an increase of this marker the days preceding the onset of signs necessary for a clinical diagnosis the sepsis. This observation justifies further evaluation of the potential clinical benefit of serial PSP measurement in the management of critically ill patients developing nosocomial sepsis.Trial registrationThe study has been registered at ClinicalTrials.gov (no. NCT03474809), on March 16, 2018.https://www.clinicaltrials.gov/ct2/show/NCT03474809?term=NCT03474809&draw=2&rank=1

Sign in / Sign up

Export Citation Format

Share Document