Deep Learning-based Classification of Viruses Using Transmission Electron Microscopy Images

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26220%2F22%3APU146546" target="_blank" >RIV/00216305:26220/22:PU146546 - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/abstract/document/9851305" target="_blank" >https://ieeexplore.ieee.org/abstract/document/9851305</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/TSP55681.2022.9851305" target="_blank" >10.1109/TSP55681.2022.9851305</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Deep Learning-based Classification of Viruses Using Transmission Electron Microscopy Images

Popis výsledku v původním jazyce

Humans have a strong urge to categorize natural organisms, and the categorization of viruses becomes more challenging. Viruses are not visible with the naked eyes, and their automatic classification based on images obtained with Transmission Electron Microscopy (TEM) can help a lot in the medical field. Their classification is more challenging due to their complicated intracellular structures and lighting conditions to capture the TEM images. The proposed architecture has been developed for the classification of the 14 different types of viruses. The dataset has been split into the training set, validation set and test set. The proposed model obtained better experimental results with 96.90% classification accuracy on the validation set and 96.10 % on the test set of unseen images. The performance of the proposed model has been compared with state-of-the-art pre-trained deep-learning models such that XceptionNet, MobileNet and DenseNet201. The model is accurate and computationally less complex, which supports faster processing suitable for microscopic cell image analysis for different medical applications.

Název v anglickém jazyce

Deep Learning-based Classification of Viruses Using Transmission Electron Microscopy Images

Popis výsledku anglicky

Humans have a strong urge to categorize natural organisms, and the categorization of viruses becomes more challenging. Viruses are not visible with the naked eyes, and their automatic classification based on images obtained with Transmission Electron Microscopy (TEM) can help a lot in the medical field. Their classification is more challenging due to their complicated intracellular structures and lighting conditions to capture the TEM images. The proposed architecture has been developed for the classification of the 14 different types of viruses. The dataset has been split into the training set, validation set and test set. The proposed model obtained better experimental results with 96.90% classification accuracy on the validation set and 96.10 % on the test set of unseen images. The performance of the proposed model has been compared with state-of-the-art pre-trained deep-learning models such that XceptionNet, MobileNet and DenseNet201. The model is accurate and computationally less complex, which supports faster processing suitable for microscopic cell image analysis for different medical applications.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

20203 - Telecommunications

Projekt

<a href="/cs/project/VI04000039" target="_blank" >VI04000039: Systém včasného záchytu infekce COVID-19 pro bezpečnost ohrožených skupin obyvatelstva s využitím umělé inteligence</a><br>

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

45th International Conference on Telecommunications and Signal Processing (TSP 2022). IEEE, 2022

ISBN

978-1-6654-6948-7

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

174-178

Název nakladatele

IEEE

Místo vydání

Prague, Czech Republic

Místo konání akce

Prague

Datum konání akce

13. 7. 2022

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—