Lehel Dénes-Fazakas

Dr. Lehel Dénes-Fazakas

Physiological Controls Research Center, Óbuda University, Budapest, Hungary
+3
orcid

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
21 September 2024 in Biomedicines
Verified
Journal Article
Open access
PDF file
Lehel Dénes-FazakasLehel Dénes-Fazakas
Lehel Dénes-Fazakas
László SzilágyiLászló Szilágyi
László Szilágyi
Levente KovácsLevente Kovács
Levente Kovács
Andrea De GaetanoAndrea De Gaetano
Andrea De Gaetano
György EignerGyörgy Eigner
György Eigner
Background/Objectives: Managing blood glucose levels effectively remains a significant challenge for individuals with diabetes. Traditional methods often lack the flexibility needed for personalized care. This study explores the potential of reinforcement learning-based approaches, which mimic human learning and adapt strategies through ongoing interactions, in creating dynamic and personalized blood glucose management plans. Methods: We developed a mathematical model specifically for patients with type IVP diabetes, validated with data from 10 patients and 17 key parameters. The model includes continuous glucose monitoring (CGM) noise and random carbohydrate intake to simulate real-life conditions. A closed-loop system was designed to enable the application of reinforcement learning algorithms. Results: By implementing a Policy Optimization (PPO) branch, we achieved an average Time in Range (TIR) metric of 73%, indicating improved blood glucose control. Conclusions: This study presents a personalized insulin therapy solution using reinforcement learning. Our closed-loop model offers a promising approach for improving blood glucose regulation, with potential applications in personalized diabetes management.

1 May 2019
Verified
Conference Paper
Closed access
MAMargit Antal
Margit Antal
Lehel Dénes-FazakasLehel Dénes-Fazakas
Lehel Dénes-Fazakas
In this paper, we compare the performance of user verification systems based on three publicly available mouse dynamics data sets. One of these datasets is our new DFL data set which contains the mouse dynamics of 21 different users. Two aspects of mouse dynamics based user verification systems are investigated: the effect of the quantity of training data on the performance, and the effect on the performance of the number of consecutive mouse actions used for user identity predictions. Measurements show that it is advisable to use approximately 1000 mouse actions for training (60 minutes of active mouse movements on average), and at least 10 mouse actions for user identity prediction.

7 November 2022 in Sensors
Verified
Journal Article
Open access
PDF file
Lehel Dénes-FazakasLehel Dénes-Fazakas
Lehel Dénes-Fazakas
MSMáté Siket
Máté Siket
László SzilágyiLászló Szilágyi
László Szilágyi
Levente KovácsLevente Kovács
Levente Kovács
György EignerGyörgy Eigner
György Eigner
Non-coordinated physical activity may lead to hypoglycemia, which is a dangerous condition for diabetic people. Decision support systems related to type 1 diabetes mellitus (T1DM) still lack the capability of automated therapy modification by recognizing and categorizing the physical activity. Further, this desired adaptive therapy should be achieved without increasing the administrative load, which is already high for the diabetic community. These requirements can be satisfied by using artificial intelligence-based solutions, signals collected by wearable devices, and relying on the already available data sources, such as continuous glucose monitoring systems. In this work, we focus on the detection of physical activity by using a continuous glucose monitoring system and a wearable sensor providing the heart rate—the latter is accessible even in the cheapest wearables. Our results show that the detection of physical activity is possible based on these data sources, even if only low-complexity artificial intelligence models are deployed. In general, our models achieved approximately 90% accuracy in the detection of physical activity.

18 October 2024 in Biomedicines
Verified
Journal Article
Open access
PDF file
Lehel Dénes-FazakasLehel Dénes-Fazakas
Lehel Dénes-Fazakas
Levente KovácsLevente Kovács
Levente Kovács
György EignerGyörgy Eigner
György Eigner
László SzilágyiLászló Szilágyi
László Szilágyi
Background: Brain tumors are highly complex, making their detection and classification a significant challenge in modern medical diagnostics. The accurate segmentation and classification of brain tumors from MRI images are crucial for effective treatment planning. This study aims to develop an advanced neural network architecture that addresses these challenges. Methods: We propose L-net, a novel architecture combining U-net for tumor boundary segmentation and a convolutional neural network (CNN) for tumor classification. These two units are coupled such a way that the CNN classifies the MRI images based on the features extracted by the U-net while segmenting the tumor, instead of relying on the original input images. The model is trained on a dataset of 3064 high-resolution MRI images, encompassing gliomas, meningiomas, and pituitary tumors, ensuring robust performance across different tumor types. Results: L-net achieved a classification accuracy of up to 99.6%, surpassing existing models in both segmentation and classification tasks. The model demonstrated effectiveness even with lower image resolutions, making it suitable for diverse clinical settings. Conclusions: The proposed L-net model provides an accurate and unified approach to brain tumor segmentation and classification. Its enhanced performance contributes to more reliable and precise diagnosis, supporting early detection and treatment in clinical applications.

25 September 2024 in Computers in Biology and Medicine
Verified
Journal Article
Open access
Lehel Dénes-FazakasLehel Dénes-Fazakas
Lehel Dénes-Fazakas
BSBarbara Simon
Barbara Simon
AHAdam Hartveg
Adam Hartveg
László SzilágyiLászló Szilágyi
László Szilágyi
Levente KovácsLevente Kovács
Levente Kovács
György EignerGyörgy Eigner
György Eigner
For individuals diagnosed with diabetes mellitus, it is crucial to keep a record of the carbohydrates consumed during meals, as this should be done at least three times daily, amounting to an average of six meals. Unfortunately, many individuals tend to overlook this essential task. For those who use an artificial pancreas, carbohydrate intake proves to be a critical factor, as it can activate the insulin pump in the artificial pancreas to deliver insulin to the body. To address this need, we have developed personalized deep learning model that can accurately detect carbohydrate intake with a high degree of accuracy. Our study employed a publicly available dataset gathered by an Inertial Measurement Unit (IMU), which included accelerometer and gyroscope data. The data was sampled at a rate of 15 Hz, necessitating preprocessing. For our tailored to the patient model, we utilized a recurrent network comprising Long short-term memory (LSTM) layers. Our findings revealed a median F1 score of 0.99, indicating a high level of accuracy. Additionally, the confusion matrix displayed a difference of only 6 s, further validating the model's accuracy. Therefore, we can confidently assert that our model architecture exhibits a high degree of accuracy. Our model performed well above 90% on the dataset, with most results between 98%-99%. The recurrent networks improved the problem-solving capabilities significantly, though some outliers remained. The model's average prediction latency was 5.5 s, suggesting that later meal predictions result in extended meal progress predictions. The dataset's limitation of mostly single-day data points raises questions about multi-day performance, which could be explored by collecting multi-day data, including night periods. Future enhancements might involve transformer networks and shorter time windows to improve model responsiveness and accuracy. Therefore, we can confidently assert that our model exhibits a high degree of accuracy.

5 November 2020
Verified
Conference Paper
Closed access
Lehel Dénes-FazakasLehel Dénes-Fazakas
Lehel Dénes-Fazakas
LSLaszla Szilagyi
Laszla Szilagyi
JTJelena Tasic
Jelena Tasic
LKLevente Kovacs
Levente Kovacs
GEGyorgy Eigner
Gyorgy Eigner
In the case of diabetes mellitus physical activity does have a high effect on the glycemic state of the patients. This is especially regarding the patients with Type 1 diabetes mellitus, who need external insulin administration in their daily life. Nevertheless, physical activity - as one source of stress - is underrepresented in the decisions of patients and medical staff and in the decisions of the available automated glucose regulatory devices. The goal of the study was to build up a simulation framework for data generation and to assess which machine learning solution can be the most accurate in the identification of physical activity.