In vivo Hounsfield Unit (HU) values have traditionally been determined using direct CT image measurements. These measurements are dependent on the window/level used to examine the CT image and the ...individual conducting the fat tissue tracing.
Using an indirect method, a new reference interval (RI) is proposed. A total of 4000 samples of fat tissues were collected from routine abdominal CT examinations. A linear regression equation was then calculated using the linear part of the cumulative frequency plot of their average values.
The regression function for total abdominal fat was determined to be y = 35.376*x - 123.48, and a 95% confidence RI of -123 to -89 was computed. A significant difference of 3.82 was observed between the average fat HU values of visceral and subcutaneous areas.
Using statistical methods and the in vivo measurements of patient data, a series of RIs were determined for fat HU that is consistent with theoretical values.
: The importance of applying the normality tests is underlined by the way of continuing the statistical protocol for numerical data within inferential statistics, respectively by the parametric or ...non-parametric tests that we will apply further on.
: To check the calculation mode, we used sets of random values and we performed the normality assessment using statistical calculation programs. We took non-Gaussian data (n = 30, n = 50, n = 100, n = 500) and Gaussian data (n = 30, n = 50, n = 100, n = 500) for which we checked the normality of the data. Data chosen for this study were most representative for each batch (n).
: The application of normality tests to the data under study confirms that the data are non-Gaussian for the first data set. For the Gaussian data sample, the verification of normality is confirmed by the results.
: For data up to 50 subjects, it is recommended to apply the Shapiro-Wilk test, but also to apply graphical methods to confirm the accuracy of the result. If the data samples have more than 50 values, the D’Agostino & Pearson omnibus normality test should be applied and if the statistical program does not contain this test, the Shapiro-Wilk test can be applied (in the case of SPSS). Graphical methods, although they require some experience, are useful for identifying the normality of distributions with a small number of data.
Occupational stress amongst doctors has been intensively studied as doctors are exposed to several stress factors daily. The purpose of this study was to investigate if there are associations between ...personality dimensions and the factors that generate stress at work. We conducted a cross-sectional study of 280 medical doctors from Romania between February 2021 and September 2021 who were evaluated using the DECAS and ASSET Scales. Our results showed that the agreeableness and emotional stability dimensions of personality, according to the Big Five model, were statistically associated with work relationships (A p < 0.0001; ES p = 0.0005), work-life balance (A p = 0.008; ES p = 0.01), overload (A p = 0.01; ES p = 0.001), job security (A p < 0.0001; ES p = 0.002), job control (A p = 0.001; ES p = 0.009), resources and communication (A p = 0.0002; ES p < 0.0001), and job conditions (A p = 0.005; ES p = 0.03). The conscientiousness dimension was statistically associated with job control (p = 0.02). Doctors from different specialties experienced stress differently, with psychiatrists and doctors from preclinical specialties reporting the lowest levels of stress. Internists and surgeons reported higher levels of stress. This study showed that the dimensions of agreeableness and emotional stability were both associated with variables indicative of the level of stress felt at work.
Machine learning (ML) models have been proposed as a prognostic clinical tool and superiority over clinical risk scores is yet to be established. Our aim was to analyse the performance of predicting ...3-year all-cause- and cardiovascular cause mortality using ML techniques and compare it with clinical scores in a percutaneous coronary intervention (PCI) population.
An all-comers patient population treated by PCI in a tertiary cardiovascular centre that have been included prospectively in the local registry between January 2016–December 2017 was analysed. The ML model was trained to predict 3-year mortality and prediction performance was compared with that of GRACE, ACEF, SYNTAX II 2020 and TIMI scores.
A total number of 2242 patients were included with 12.1% and 14.9% 3-year cardiovascular and -all-cause mortality, respectively. The area under receiver operator characteristic curve for the ML model was higher than that of GRACE, ACEF, SYNTAX II and TIMI scores: 0.886 vs. 0.797, 0.792, 0.757 and 0.696 for 3-year cardiovascular- and 0.854 vs. 0.762, 0.764, 0.730 and 0.691 for 3-year all-cause mortality prediction, respectively (all p ≤ 0.001). Similarly, the area under precision-recall curve for the ML model was higher than that of GRACE, ACEF, SYNTAX II and TIMI scores: 0.729 vs. 0.474, 0.469, 0.365 and 0.389 for 3-year cardiovascular- and 0.718 vs. 0.483, 0.466, 0.388 and 0.395 for 3-year all-cause mortality prediction, respectively (all p ≤ 0.001).
The ML model was superior in predicting 3-year cardiovascular- and all-cause mortality when compared to clinical scores in a prospective PCI registry.
Display omitted
•Machine learning (ML) outperformed risk scores such as GRACE, ACEF, SYNTAX II, and TIMI scores in a prospective percutaneous coronary intervention (PCI) registry.•ML model was superior in predicting in-hospital, out-of-hospital, and 3-year all-cause and cardiovascular cause mortality.•ML model used LVEF, age, creatinine, neutrophil/lymphocyte ratio and minimum lymphocyte count as most important variables.
Abstract Introduction Healthcare-associated infections have a significant impact on public health, and many patients and their next-of-kin are seeking information on the internet. The study aimed to ...assess the quality of online written content about healthcare-associated infections available in English, Romanian, and Hungarian languages. Materials and methods The study sample included 75 websites, 25 for each language subgroup. The assessment involved examining the general characteristics, adherence to established credibility criteria, and the completeness and accuracy of informational content. The evaluation was conducted using a topic-specific, evidence-based benchmark. Two evaluators independently graded completeness and accuracy; scores were recorded on a scale from 0 to 10. A comparative analysis of websites was performed, considering pertinent characteristics, and potential factors influencing information quality were subjected to testing. The statistical significance was set at 0.05. Results For the overall study sample, the average credibility, completeness, and accuracy scores were 5.1 (SD 1.7), 2.4 (SD 1.5), and 5.9 (SD 1.0), respectively. Pairwise comparison tests revealed that English websites rated significantly higher than Romanian and Hungarian websites on all three quality measures (P<0.05). Website specialization, ownership, and main goal were not associated with credibility or content ratings. However, conventional medicine websites consistently scored higher than alternative medicine and other websites across all three information quality measures (P<0.05). Credibility scores were positively but weakly correlated with completeness (rho=0.273; P=0.0176) and accuracy scores (rho=0.365; P=0.0016). Conclusions The overall quality ratings of information about healthcare-associated infections on English, Romanian, and Hungarian websites ranged from intermediate to low. The description of information regarding the symptoms and prevention of healthcare-associated infections was notably unsatisfactory. The study identified website characteristics possibly associated with higher-quality online sources about healthcare-associated infections, but additional research is needed to establish robust evidence.
Background: A transparent machine learning model enables the end-user to investigate the decision process from (clinical) input data to the final prediction of the model. Modern (Deep Learning) ...machine learning models are highly complex and opaque - lacking the transparency required by both the clinician and the patient. Not knowing why a model predicted a particular outcome in a clinical context would likely hinder the clinician's trust in the model and ultimately impact the patient's health outcome. Aim. To integrate the model training into a unique automated pipeline that will output the prediction along with the relevant explanation. Materials and Methods:. The workflow was build using the Julia programming language and employing a wide range of machine learning packages from the MLJ universe. The interoperability with the explainer packages from the Python ecosystem was enabled using the PyCall.jl package. Results. An automated machine learning pipeline that a) exposes the (clinical) decision process inside various blackbox models (convolutional neural networks, random forest, boosted models), b) enables the training of explainable models without compromising the performance metrics, c) validates the resulted method against the medical domain-knowledge. Our pipeline combines the known non-healthcare-specific explainers - LIME, SHAP, ShapML, and InterpretML into a global explainer tailored to healthcarespecific data. Conclusions. The performance metrics of machine learning models trained on healthcare data are not necessarily sacrificed on the altar of transparency and interpretability. Using model-agnostic and model-specific explainers, we can satisfy the clinician's need for a transparent decision and diagnosis process and the superior performance metrics that ensure the predictive power of the model.
Background: Not all datasets are created equal. There are some happy scenarios when the researcher has the luxury of curating the dataset and ensuring all the desired fields are filled. However, ...especially when retrieving data from large EMR databases out in the wild, missing values are the norm (e.g., only some patients would have blood sugar readings, only a portion of the patients had transaminases determined, etc.). But the vast majority of machine learning models are not supporting missing values - hence, traditionally, we call the imputation methods to the rescue. But when a significant portion of values is missing, the imputation methods might insert incorrect data. Also, dropping the cases with missing values is not feasible when working with EMR data (too many instances are removed). Aim. To implement a missing-values-proof ensemble modeling method without sacrificing the predictive power. Materials and Methods:. Using a realistic synthetic patient generator (Synthea), we generated large FHIR datasets under various settings. We implemented a cartesian genetic programming model to develop an automated acceptance test that, in turn, extracted missing-values-free subsets from the training partition of the original dataset. One of the core functions of the acceptance test is to ensure the potential missing values patterns are not going to insert bias in the model training step. We trained one or more models on each missing-values-free training subset. The resulted models are integrated into a global ensemble model. Results. Our approach resulted in superior model performance metrics without the need for missing values imputation methods. Conclusions. Machine learning on datasets containing missing values is feasible when employing specialized training dataset generation pipelines. Removing the missing values imputation methods from the workflow eliminates potentially incorrect data insertion, resulting in more robust models.
(1) Background: Although transcatheter aortic valve replacement (TAVR) significantly improves long-term outcomes of symptomatic severe aortic stenosis (AS) patients, long-term mortality rates are ...still high. The aim of our study was to identify potential inflammatory biomarkers with predictive capacity for post-TAVR adverse events from a wide panel of routine biomarkers by employing ML techniques. (2) Methods: All patients diagnosed with symptomatic severe AS and treated by TAVR since January 2016 in a tertiary center were included in the present study. Three separate analyses were performed: (a) using only inflammatory biomarkers, (b) using inflammatory biomarkers, age, creatinine, and left ventricular ejection fraction (LVEF), and (c) using all collected parameters. (3) Results: A total of 338 patients were included in the study, of which 56 (16.5%) patients died during follow-up. Inflammatory biomarkers assessed using ML techniques have predictive value for adverse events post-TAVR with an AUC-ROC of 0.743 and an AUC-PR of 0.329; most important variables were CRP, WBC count and Neu/Lym ratio. When adding age, creatinine and LVEF to inflammatory panel, the ML performance increased to an AUC-ROC of 0.860 and an AUC-PR of 0.574; even though LVEF was the most important predictor, inflammatory parameters retained their value. When using the entire dataset (inflammatory parameters and complete patient characteristics), the ML performance was the highest with an AUC-ROC of 0.916 and an AUC-PR of 0.676; in this setting, the CRP and Neu/Lym ratio were also among the most important predictors of events. (4) Conclusions: ML models identified the CRP, Neu/Lym ratio, WBC count and fibrinogen as important variables for adverse events post-TAVR.
Background: Health on the Net (HON) promotes access to reliable medical information. The study aimed to investigate the prevalence of HON certification and whether it is associated with the quality ...of informational content. Methods: The study included the results of 21 researches conducted between 2018-2021 about the quality of online medical information (1040 Romanian and English language webpages). HON certification was assessed using the HONcode toolbar application. Completeness and accuracy scores were extracted from previous research. The frequency of certification and compliance with individual HON principles was calculated. Associations between HON certification and information quality scores (completeness and accuracy) were tested using the student t-test. The threshold for statistical significance was set at 0.05. Results. Health on the Net certification was present on 1.1% and 21.6% of the Romanian and English websites, respectively. Compliance with the HON principles varied between 9.7% and 82.7% of Romanian-language websites and between 35.6% and 88.5% of English-language ones. The principle with the highest compliance was transparency regarding the texts' authors in both languages. The minimum compliance was with the principle of displaying the date of the last update in the case of Romanian-language websites and of attributing sources in the case of English-language websites. Association tests between HON certification and information quality showed the following results: HON certification vs. completeness p=0.201 for the Romanian subsample and p<0.0001 for the English-language subsample; HON certification vs. accuracy p=0.018 for the Romanian subsample and p=0.886 for the English-language subsample. Conclusions. Health on the Net certification of Romanian-language sites was almost absent, while English-language websites had relatively higher, but still low certification rates. HON certification was associated with content accuracy for Romanian-language websites and with the completeness of information for English-language sites. If subsequent research confirms the results of this meta-analysis, users should reconsider their reliance on HON certification.