Oliver Lohaj

Dr. Oliver Lohaj

orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
Oliver LohajOliver Lohaj
Oliver Lohaj
Ján ParaličJán Paralič
Ján Paralič
PBPeter Bednár
Peter Bednár
ZPZuzana Paraličová
Zuzana Paraličová
MHMatúš Huba
Matúš Huba
Machine learning (ML) has been used in different ways in the fight against COVID-19 disease. ML models have been developed, e.g., for diagnostic or prognostic purposes and using various modalities of data (e.g., textual, visual, or structured). Due to the many specific aspects of this disease and its evolution over time, there is still not enough understanding of all relevant factors influencing the course of COVID-19 in particular patients. In all aspects of our work, there was a strong involvement of a medical expert following the human-in-the-loop principle. This is a very important but usually neglected part of the ML and knowledge extraction (KE) process. Our research shows that explainable artificial intelligence (XAI) may significantly support this part of ML and KE. Our research focused on using ML for knowledge extraction in two specific scenarios. In the first scenario, we aimed to discover whether adding information about the predominant COVID-19 variant impacts the performance of the ML models. In the second scenario, we focused on prognostic classification models concerning the need for an intensive care unit for a given patient in connection with different explainability AI (XAI) methods. We have used nine ML algorithms, namely XGBoost, CatBoost, LightGBM, logistic regression, Naive Bayes, random forest, SGD, SVM-linear, and SVM-RBF. We measured the performance of the resulting models using precision, accuracy, and AUC metrics. Subsequently, we focused on knowledge extraction from the best-performing models using two different approaches as follows: (a) features extracted automatically by forward stepwise selection (FSS); (b) attributes and their interactions discovered by model explainability methods. Both were compared with the attributes selected by the medical experts in advance based on the domain expertise. Our experiments showed that adding information about the COVID-19 variant did not influence the performance of the resulting ML models. It also turned out that medical experts were much more precise in the identification of significant attributes than FSS. Explainability methods identified almost the same attributes as a medical expert and interesting interactions among them, which the expert discussed from a medical point of view. The results of our research and their consequences are discussed.

31 January 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
Ján ParaličJán Paralič
Ján Paralič
MKMichal Kolarik
Michal Kolarik
ZPZuzana Paraličová
Zuzana Paraličová
Oliver LohajOliver Lohaj
Oliver Lohaj
AJAdam Jozefík
Adam Jozefík
Deep neural network models have produced significant results in solving various challenging tasks, including medical diagnostics. To increase the credibility of these black-box models in the eyes of doctors, it is necessary to focus on their explainability. Several papers have been published combining deep learning methods with selected types of explainability methods, usually aimed at analyzing medical image data, including ECG images. The ECG is specific because its image representation is only a secondary visualization of stream data from sensors. However, explainability methods for stream data are rarely investigated. Therefore, in this article we focus on the explainability of black-box models for stream data from 12-lead ECG. We designed and implemented a perturbation explainability method and verified it in a user study on a group of medical students with experience in ECG tagging in their final years of study. The results demonstrate the suitability of the proposed method, as well as the importance of including multiple data sources in the diagnostic process.

28 April 2024 in Diagnostics
Verified
Journal Article
Open access
PDF file
Oliver LohajOliver Lohaj
Oliver Lohaj
Ján ParaličJán Paralič
Ján Paralič
Zuzana PellaZuzana Pella
Zuzana Pella
DPDominik Pella
Dominik Pella
APAdam Pavlíček
Adam Pavlíček
The focus of this study, and the subject of this article, resides in the conceptually funded usability evaluation of an application of descriptive models to a specific dataset obtained from the East Slovak Institute of Heart and Vascular Diseases targeting cardiovascular patients. Delving into the current state-of-the-art practices, we examine the extent of cardiovascular diseases, descriptive data analysis models, and their practical applications. Most importantly, our inquiry focuses on exploration of usability, encompassing its application and evaluation methodologies, including Van Welie’s layered model of usability and its inherent advantages and limitations. The primary objective of our research was to conceptualize, develop, and validate the usability of an application tailored to supporting cardiologists’ research through descriptive modeling. Using the R programming language, we engineered a Shiny dashboard application named DESSFOCA (Decision Support System For Cardiologists) that is structured around three core functionalities: discovering association rules, applying clustering methods, and identifying association rules within predefined clusters. To assess the usability of DESSFOCA, we employed the System Usability Scale (SUS) and conducted a comprehensive evaluation. Additionally, we proposed an extension to Van Welie’s layered model of usability, incorporating several crucial aspects deemed essential. Subsequently, we rigorously evaluated the proposed extension within the DESSFOCA application with respect to the extended usability model, drawing insightful conclusions from our findings.

26 June 2026 in Healthcare
Verified
Journal Article
Open access
PDF file
Oliver LohajOliver Lohaj
Oliver Lohaj
PKPavel Kočan
Pavel Kočan
ABAnna Biceková
Anna Biceková
DJDaniela Javorská
Daniela Javorská
Background/Objectives: Clostridioides difficile infection (CDI) represents an important healthcare-associated complication in hospitalized patients, particularly in those exposed to antibiotics, prolonged hospitalization, and intensive treatment during COVID-19. This study aimed to design, evaluate, and interpret machine learning models for predicting CDI occurrence in hospitalized COVID-19 patients across individual pandemic waves, with respect to administered treatment and clinical characteristics. Methods: Anonymized clinical data from 3848 COVID-19-positive patients treated at the University Hospital of L. Pasteur in Košice, Slovakia, were analyzed following the CRISP-DM methodology. Four classification models were compared: logistic regression, Random Forest, XGBoost, and a multilayer perceptron. Missing values were addressed using MICE and KNN imputation, and class imbalance was handled through oversampling techniques. Given the low CDI prevalence of 2.68%, model performance was primarily assessed using the precision–recall area under the curve (PR-AUC), with AUROC reported for comparability. Interpretability was supported using SHAP, LIME, and odds ratio analysis. Results: The best-performing models achieved PR-AUC values up to 0.160, representing more than a fivefold improvement over the random baseline of 0.027. XGBoost reached the highest AUROC of 0.823, followed by Random Forest with 0.798. Inflammatory markers were identified as important predictors of CDI risk. A Flask-based decision-support web application was developed to provide CDI risk estimation with patient-specific explanations. A preliminary pilot usability evaluation involving two physicians yielded a mean System Usability Scale score of 73.75; however, the very small evaluator sample limits the generalizability of this finding. Conclusions: Interpretable machine learning models can support clinically meaningful CDI risk stratification in highly imbalanced COVID-19 hospital datasets. The proposed decision-support tool shows potential for future integration into clinical workflows, although external and prospective validation is required.

2 March 2022
Verified
Conference Paper
Closed access
Oliver LohajOliver Lohaj
Oliver Lohaj
Zuzana PellaZuzana Pella
Zuzana Pella
Ján ParaličJán Paralič
Ján Paralič
The article deals with data analytics methods that analyze the influence of individual factors on the severity of cardiovascular disease. Cardiovascular diseases (CVD) are characterized by their severity and high impact also in terms of a financial burden on the human population. After analyzing the related work in this research field, we propose a methodology and perform experiments targeting on identification of the main factors of CVD. The experimental part of our research took place on a dataset from the East Slovak Institute of Cardiovascular Diseases in Kosice. Four classification models were applied to this dataset: C5.0, CART, Logistic Regression, and Random Forest. Our goal was to focus on the most commonly used attributes in the outputs of particular models, indicating the degree of influence of individual attributes on the severity of CVD. The CART and C5.0 models matched the highest number of attributes. The Forward Stepwise Selection selected three sets of attributes according to 3 different criteria that appear to be significant, which was confirmed by the percentage values of the resulting accuracy of the classification. All steps of the experimental part were continuously consulted with a domain expert working at the East Slovak Institute of Cardiovascular Diseases in Kosice.

1 December 2023 in Acta Electrotechnica et Informatica
Verified
Journal Article
Open access
PDF file
Oliver LohajOliver Lohaj
Oliver Lohaj
PFPeter Fedačko
Peter Fedačko
JPJán Paralič
Ján Paralič
This article deals with the issue of usability of an exploratory data analysis tool in the field of medicine. The text portion contains a description of the methods and the visualization procedure. It analyses the current state on usability, medical data visualization and presents the benefits of visualization tools. The goal of this research was to design and implement a suitable visualization tool for the provided dataset, which is the result of cooperation between Technical university of Košice and East Slovak Institute of Heart and Vascular Diseases (VÚSCH) and presents cardiovascular disease data. The resulting web application aimed to provide potential users with an overview of cardiovascular diseases and the influence of their lifestyle on the assumption of such a disease, and for experts to provide different forms of viewing the given data, as well as to facilitate their work. The application was then tested for usability in two iterations where corrections and improvements were made based on the given feedback. Thanks to this iterative procedure and careful reflection of the user feedback the application received endorsement as a valuable tool by the expert evaluator.