Background: Despite the increased use of mechanical thrombectomy (MT) in recent years, there remains a lack of research on in-hospital mortality rates following the procedure, the primary factors influencing these rates, and the potential for predicting them. This study aimed to utilize interpretable machine learning (ML) to help clarify these uncertainties. Methods: This retrospective study involved patients with anterior circulation large vessel occlusion (LVO)-related ischemic stroke who underwent MT. The patient division was made into two groups: (I) the in-hospital death group, referred to as miserable outcome, and (II) the in-hospital survival group, or favorable outcome. Python 3.10.9 was utilized to develop the machine learning models, which consisted of two types based on input features: (I) the Pre-MT model, incorporating baseline features, and (II) the Post-MT model, which included both baseline and MT-related features. After a feature selection process, the models were trained, internally evaluated, and tested, after which interpretation frameworks were employed to clarify the decision-making processes. Results: This study included 602 patients with a median age of 76 years (interquartile range (IQR) 65–83), out of which 54% (n = 328) were female, and 22% (n = 133) had miserable outcomes. Selected baseline features were age, baseline National Institutes of Health Stroke Scale (NIHSS) value, neutrophil-to-lymphocyte ratio (NLR), international normalized ratio (INR), the type of the affected vessel (‘Vessel type’), peripheral arterial disease (PAD), baseline glycemia, and premorbid modified Rankin scale (pre-mRS). The highest odds ratio of 4.504 was observed with the presence of peripheral arterial disease (95% confidence interval (CI), 2.120–9.569). The Pre-MT model achieved an area under the curve (AUC) value of around 79% utilizing these features, and the interpretable framework discovered the baseline NIHSS value as the most influential factor. In the second data set, selected features were the same, excluding pre-mRS and including puncture-to-procedure-end time (PET) and onset-to-puncture time (OPT). The AUC value of the Post-MT model was around 84% with age being the highest-ranked feature. Conclusions: This study demonstrates the moderate to strong effectiveness of interpretable machine learning models in predicting in-hospital mortality following mechanical thrombectomy for ischemic stroke, with AUCs of 0.792 for the Pre-MT model and 0.837 for the Post-MT model. Key predictors included patient age, baseline NIHSS, NLR, INR, occluded vessel type, PAD, baseline glycemia, pre-mRS, PET, and OPT. These findings provide valuable insights into risk factors and could improve post-procedural patient management.

Introduction: While the machine learning (ML) model’s black‐box nature presents a significant barrier to effective clinical application, the dynamic nature of stroke patients’ recovery further undermines the reliability of established predictive scores and models, making them less suitable for accurate prediction and appropriate patient care. This research is aimed at building and evaluating an interpretable ML‐based model, which would perform outcome prediction at different time points of patients’ recovery, giving more secure and understandable output through interpretable packages. Materials and Methods: A retrospective analysis was conducted on acute ischemic stroke (AIS) patients treated with alteplase at the Neurology Clinic of the University Clinical Center of Vojvodina (Novi Sad, Serbia), for 14 years. Clinical data were grouped into four categories based on collection time—baseline, 2‐h, 24‐h, and discharge features—serving as inputs for three different classifiers—support vector machine (SVM), logistic regression (LR), and random forest (RF). The 90‐day modified Rankin scale (mRS) was used as the outcome measure, distinguishing between favorable (mRS ≤ 2) and unfavorable outcomes (mRS ≥ 3). Results: The sample was described with 49 features and included 355 patients, with a median age of 67 years (interquartile range (IQR) 60–74 years), 66% being male. The models achieved strong discrimination in the testing set, with area under the curve (AUC) values ranging from 0.80 to 0.96. Additionally, they were compared with a model based on the DRAGON score, which showed an AUC of 0.760 (95% confidence interval (CI), 0.640–0.862). The decision‐making process was more thoroughly understood using interpretable packages: Shapley additive explanation (SHAP) and local interpretable model–agnostic explanation (LIME). They revealed the most significant features at both the group and individual patient levels. Conclusions and Clinical Implications: This study demonstrated the moderate to strong efficacy of interpretable ML‐based models in predicting the functional outcomes of alteplase‐treated AIS patients. In all constructed models, age, onset‐to‐treatment time, and platelet count were recognized as the important predictors, followed by clinical parameters measured at different time points, such as the National Institutes of Health Stroke Scale (NIHSS) and systolic and diastolic blood pressure values. The dynamic approach, coupled with interpretable models, can aid in providing insights into the potential factors that could be modified and thus contribute to a better outcome.

Background and aims: Knowing that heart failure (HF) with preserved ejection fraction (HFpEF) can be difficult to diagnose, a clinical decision support system (CDSS), with functional, morphological and biomarker domains, was built. Therefore, this research aimed to analyse the possibility of implementing CDSS, compliance of the physicians with the CDSS recommendations, as well as identify obstacles and facilitators to its usage. Methods: This mixed-methods study was conducted over a 1 year period at the Institute of Cardiovascular Diseases of Vojvodina (Sremska Kamenica, Serbia). The retrospective component included an analysis of patients who had symptoms and/or signs of HF, with an ejection fraction (EF) of ≥50%. The Heart Failure Association pre-test assessment, echocardiography and natriuretic peptide, functional testing, final aetiology (HFA-PEFF) score was used to build a CDSS and to calculate the probability scores for these patients. The prospective component involved an electronic survey that was used to gain feedback from physicians about the CDSS. Results: The sample included 225 patients, out of which 194 (86.2%) had the HFA-PEFF score of 5–6, suggesting probable HFpEF. This syndrome was clinically documented in the discharge report in only 32 (16.5%) of these patients. The response rate of the survey was 83.3%, including physicians at different levels of education. Barriers to CDSS usage included a lack of echocardiographic skills and knowledge about phenotype profiling in HFpEF patients. An increase in explainability and transparency, as well as staff education, were mentioned as the main facilitators. Conclusions: The CDSS demonstrated reasonable effectiveness in diagnosing and classifying patients with symptoms of HF if the measured EF was ≥50%. The conducted survey highlighted that physician education, in both CDSS- and HFpEF-related domains, is a key factor in the battle against HFpEF.

Background: as a relatively high number of ST-segment elevation myocardial infarction (STEMI) patients develop major adverse cardiovascular events (MACE) following percutaneous coronary intervention (PCI), our aim was to determine the significance, and possible predictive value of QRS complex width and ST-segment elevation. Methods: our patient sample included 200 PCI-treated STEMI patients, which were divided into two groups based on the following duration of symptoms: (I) less than 6 h, and (II) 6 to 12 h. For every patient, an ECG was performed at six different time points, patients were followed for up to six years for the occurrence of MACE. Results: the mean age was 60.6 ± 11.39 years, and 142 (71%) were male. The 6–12 h group had significantly wider QRS complex, higher ST-segment elevation, lower prevalence of ST-segment resolution as well as MACE prevalence (p < 0.05). ECG parameters, QRS width, and magnitude of ST-segment elevation were proved to be independent significant predictors of MACE in all measured time points (p < 0.05). Even after controlling for biomarkers of myocardial injury, these ECG parameters remained statistically significant predictors of MACE (p < 0.05). Conclusion: our study highlights that wider QRS complex and a more pronounced ST-segment elevation are associated with longer total ischemic time and higher risk of long-term MACE.

Background/Aim. To date, most of the scientific attention has been aimed at the morphometric analysis of the nephrogenic zone (NZ) of the fetal renal cortex, while the quantification and analysis of the maturation zone (MZ) and other indicators of renal maturity were missing. The aim of the study was to examine the characteristics of fetal kidney cortex maturation, as well as to propose the development of a new cortical maturity index (CMI). Methods. The study included 42 paraffin molds of the fetal kidney, divided into three groups according to gestational age (GA). After hematoxylin and eosin staining, tissue sections were analyzed through the following parameters: the thickness of the NZ and MZ, the renal corpuscles area (RCa) and the glomerular capillary tuft area (GCTa), and the maturation stages of the glomeruli. In addition, a new parameter, CMI, was formed as a ratio of NZ and MZ thickness. The collected data were statistically processed. Results. Changes in NZ and MZ thickness were statistically significant, and they correlated with GA. A value of CMI higher than 0.2 was recorded in the kidney samples of fetuses younger than the 20th gestational week (GW), while a value lower than 0.1 was recorded in the samples older than the 30th GW. With an increase in GA in all zones of the renal cortex, RCa and GCTa decreased. A statistically significant reduction of GCTa was observed in the oldest group in the juxtamedullary and intermediate zones of the cortex (p < 0.01). Glomeruli located in the deeper parts of the cortex were more mature than the superficial ones. Conclusion. The measured parameters can serve as a starting point for future studies that would analyze the histomorphological characteristics of the fetal kidney cortex. In the absence of clinical data, a newly formed parameter CMI can represent assistance with the determination of GA, as it significantly correlates with GA (p < 0.01).

Background and Objectives: Long-term real-world data regarding smoking cessation after myocardial infarction (MI) remain scarce in Southeastern Europe. Our objective was to estimate the prevalence of smoking status categories and to assess their association with long-term all-cause mortality in percutaneous coronary intervention (PCI) treated MI patients. Materials and Methods: We retrospectively collected data from electronic medical records of hospitalized MI patients who underwent PCI from 1 January 2016 to 1 January 2019. Smoking status and primary outcome (all-cause mortality) were assessed at baseline, during follow-up visits, and during teleconsultation until January 2024, and patients were categorized accordingly. Results: Never-smokers were older than persistent smokers (67.6 ± 11.4 vs. 57.3 ± 9.6 years; p < 0.01) and had a higher burden of hypertension and diabetes (both p < 0.01). At baseline, 47.9% were active smokers, 32.3% of them quit after an MI (with 77.7% being men). Over a median follow-up of 6.7 years, 220 deaths (17.3%) occurred. In a Cox regression model adjusted for age, persistent post-MI smoking significantly increased mortality risk (aHR 2.18; 95% CI 1.52–3.12; p < 0.001), former smokers had an aHR was 1.07 (p = 0.694), while quitters had 30% reduction in death risk (aHR 0.69; p = 0.217). Additionally, in the fully adjusted model, besides persistent smoking (aHR 2.18; p = 0.001), diabetes (aHR 1.69; p = 0.001) and age (aHR 1.09; p < 0.001) were independent predictors of all-cause mortality. Conclusions: Post-MI smoking more than doubles mortality risk, while sustained cessation reduces risk to near never-smoker levels, underscoring the need for aggressive, tailored cessation strategies in low-quit-rate regions such as Southeastern Europe.