Brain Tumor Detection Based on Features Extracted and Classified Using a Low-Complexity Neural Network

Papageorgiou, Vasileios E.

doi:10.18280/ts.380302

Cited by 20 publications

(6 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We note that the SEIRD scheme should be considered appropriate for the examination of mpox, since the publicly available datasets only include records for the number of infected and deceased cases, which is a frequently observed phenomenon is other cases of epidemics. Thus, the implementation of a more complex epidemiological structure, can lead to overfitting effects-that accompany both artificial intelligence and statistical models [25,51,52]-since any additional states added to the model, are not supported by real-time data.…”

Section: Discussionmentioning

confidence: 99%

A novel epidemiologically informed particle filter for assessing epidemic phenomena. Application to the monkeypox outbreak of 2022

Papageorgiou,

Kolias

2024

Inverse Problems

View full text Add to dashboard Cite

Contagious diseases are constantly affecting more and more people every day, resulting in widespread health crises especially in developing nations. Previous studies have developed deterministic and stochastic mathematical models to investigate the spread of epidemics. In the present study, a hybrid particle filtering epidemiological model is proposed, which combines the elements of a deterministic susceptible-exposed-infectious-recovered-deceased (SEIRD) model with the inclusion of stochastic and penalty factors, in order to efficiently evaluate the dynamics of the disease. The inclusion of penalty factors stands out as the main novelty of the proposed methodology, guaranteeing estimations that align with the unique aspects of the examined natural phenomenon. The model is applied to the monkeypox (mpox) data of the United States from June 25th to November 21th, 2022. Our approach is compared to four alternatives, corresponding to deterministic and stochastic approaches that are associated with either fixed or time-varying parameters. In all cases, the particle filtering models displayed better characteristics in terms of infectious cases and deaths compared to their deterministic counterpart. The final version of the proposed epidemiologically informed particle filtering (EI-PF) model exhibited significant potential and provided the best fitting/predictive performance compared to other examined methodologies. The predictive effectiveness of the proposed methodology has been thoroughly evaluated across various time intervals. Moreover, the inclusion of additional penalty factors in the weight computation procedure, assists in reducing fitting and prediction errors while simultaneously providing increased likelihood estimates. This modelling approach can be readily applied to other epidemics, both existing and emerging, where uncertainties in system dynamics and real-time observations hinder the accurate capture of the epidemic's progression. 

show abstract

Section: Discussionmentioning

confidence: 99%

A novel epidemiologically informed particle filter for assessing epidemic phenomena. Application to the monkeypox outbreak of 2022

Papageorgiou,

Kolias

2024

Inverse Problems

View full text Add to dashboard Cite

show abstract

“…In our CNN we used max pooling and the outputs of the pooling layers are produced according to

y_{i^{k}, j^{k}, d} = \max_{0 \leq i \leq m, 0 \leq j \leq n} x_{i^{k} \times m + i, j^{k} \times n + j, d}^{k},

where,

0 \leq i^{k} \leq M^{k}, 0 \leq j^{k} \leq N^{k}

and

0 \leq d \leq D^{k}

. Intuitively, pooling layers decrease the dimensions of the output tensors while maintain the most crucial detected patterns for the purposes of the classification 56 …”

Section: Methods and Mathematical Toolsmentioning

confidence: 99%

“…Intuitively, pooling layers decrease the dimensions of the output tensors while maintain the most crucial detected patterns for the purposes of the classification. 56…”

Section: Pooling Layermentioning

confidence: 99%

Analysis of digitalized ECG signals based on artificial intelligence and spectral analysis methods specialized in ARVC

Papageorgiou

Zegkos

Efthimiadis

et al. 2022

Numer Methods Biomed Eng

View full text Add to dashboard Cite

Arrhythmogenic right ventricular cardiomyopathy (ARVC) is an inherited heart muscle disease that appears between the second and forth decade of a patient's life, being responsible for 20% of sudden cardiac deaths before the age of 35. The effective and punctual diagnosis of this disease based on electrocardiograms (ECGs) could have a vital role in reducing premature cardiovascular mortality. In our analysis, we first outline the digitalization process of paperbased ECG signals enhanced by a spatial filter aiming to eliminate dark regions in the dataset's images that do not correspond to ECG waveform, producing undesirable noise. Next, we propose the utilization of a low-complexity convolutional neural network for the detection of an arrhythmogenic heart disease, that has not been studied through the usage of deep learning methodology to date, achieving high classification accuracy, namely 99.98% training and 98.6% testing accuracy, on a disease the major identification criterion of which are infinitesimal millivolt variations in the ECG's morphology, in contrast with other arrhythmogenic abnormalities. Finally, by performing spectral analysis we investigate significant differentiations in the field of frequencies between normal ECGs and ECGs corresponding to patients suffering from ARVC. In 16 out of the 18 frequencies where we encounter statistically significant differentiations, the normal ECGs are characterized by greater normalized amplitudes compared to the abnormal ones. The overall research carried out in this article highlights the importance of integrating mathematical methods into the examination and effective diagnosis of various diseases, aiming to a substantial contribution to their successful treatment.

show abstract

“…Moreover, most compartmental models do not consider the complete disease dynamics. Inevitably, this involves highly complex structures, resulting in severe overfitting, a phenomenon that can occur in any statistical or artificial intelligence model [20,21]. Therefore, a transition from a deterministic to a stochastic approach appears to be necessary.…”

Section: Introductionmentioning

confidence: 99%

A stochastic particle extended SEIRS model with repeated vaccination: Application to real data of COVID‐19 in Italy

Papageorgiou,

Tsaklidis

2024

Math Methods in App Sciences

View full text Add to dashboard Cite

The prediction of the evolution of epidemics plays an important role in limiting the transmissibility and the burdensome consequences of infectious diseases, which leads to the employment of mathematical modeling. In this paper, we propose a stochastic particle filtering extended SEIRS model with repeated vaccination and time‐dependent parameters, aiming to efficiently describe the demanding dynamics of time‐varying epidemics. The validity of our model is examined using daily records of COVID‐19 in Italy for a period of 525 days, revealing a notable capacity to uncover the hidden dynamics of the pandemic. The main findings include the estimation of asymptomatic cases, which is a well‐known feature of the current pandemic. Unlike other proposed models that employ extra compartments for asymptomatic cases, which force the estimation of this proportion and significantly increase the model's complexity, our approach leads to the evaluation of the hidden dynamics of COVID‐19 without additional computational burden. Other findings that confirm the model's appropriateness and robustness are its parameter evolution and the estimation of more ICU‐admitted cases compared to the official records during the most prevalent infection wave of January 2022, attributed to the intensified increase in admissions that may have led to full occupancy in ICUs. As the vast majority of datasets contain time series of total recovered and vaccinated cases, we propose a statistical algorithm to estimate the currently recovered and protected through vaccination cases. This necessity arises from the attenuation of antibodies after vaccination/infection and is necessary for long‐time interval predictions. Finally, we not only present a novel stochastic epidemiological model and test its efficiency but also investigate its mathematical properties, such as the existence and stability of epidemic equilibria, giving new insights to the literature. The latter provides additional details concerning the system's long‐term behavior, while the conclusions drawn from the index provide perspectives on the severity and future of the COVID‐19 pandemic.

show abstract

Brain Tumor Detection Based on Features Extracted and Classified Using a Low-Complexity Neural Network

Cited by 20 publications

References 30 publications

A novel epidemiologically informed particle filter for assessing epidemic phenomena. Application to the monkeypox outbreak of 2022

A novel epidemiologically informed particle filter for assessing epidemic phenomena. Application to the monkeypox outbreak of 2022

Analysis of digitalized ECG signals based on artificial intelligence and spectral analysis methods specialized in ARVC

A stochastic particle extended SEIRS model with repeated vaccination: Application to real data of COVID‐19 in Italy

Contact Info

Product

Resources

About