Comparative Evaluation of Custom Convolutional Neural Networks and EfficientNet-B3 for Malaria Cell Image Classification: Impact of Targeted Data Augmentation on Model Performance

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11110%2F25%3A10506820" target="_blank" >RIV/00216208:11110/25:10506820 - isvavai.cz</a>

Výsledek na webu

<a href="https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=TQi6mnB7yC" target="_blank" >https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=TQi6mnB7yC</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.61882/ijbc.17.3.62" target="_blank" >10.61882/ijbc.17.3.62</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Comparative Evaluation of Custom Convolutional Neural Networks and EfficientNet-B3 for Malaria Cell Image Classification: Impact of Targeted Data Augmentation on Model Performance

Popis výsledku v původním jazyce

Background: Malaria diagnosis with thin blood smears remains labor-intensive and relies on the operator. Deep learning could enable accurate automation. Objective: Compare four convolutional approaches for classifying parasitized versus uninfected erythrocytes and to evaluate whether targeted image-quality augmentations enhance performance. Materials and Methods: We used the balanced NIH/Kaggle dataset, which included 13,780 parasitized and 13,780 uninfected samples. Data were split stratified into training, validation, and test sets (70/15/15). Images were resized to 256x256 and normalized. Four experiments were conducted: (1) a custom CNN; (2) the same CNN with targeted augmentation applied to 20% of training samples per class-using Contrast Limited Adaptive Histogram Equalization [CLAHE] and controlled brightness adjustment-and augmented images were added back to the training set (totaling 30,864 images); (3) a soft-attention parallel CNN (SPCNN); and (4) transfer learning with EfficientNet-B3 on 300x300 inputs with full fine-tuning. Evaluation metrics included accuracy, precision, recall, F1 score, and AUC-ROC. Results: EfficientNet-B3 achieved the highest performance with a validation accuracy of 0.9741, 98% precision, 96% recall, an F1 score of 0.97, and an AUC-ROC of 0.9964. SPCNN was competitive but slightly lower, with a validation accuracy of 0.9652, 98% precision, 95% recall, an F1 score of 0.96, and an AUC-ROC of 0.9909. The baseline CNN had a validation accuracy of 0.9649, 97% precision, 94% recall, an F1 score of 0.96, and an AUC-ROC of 0.9910. Targeted augmentation resulted in negligible change compared to the baseline CNN, with a validation accuracy of 0.9647, an F1 score of 0.96, and an AUC-ROC of 0.9908, indicating limited added discriminative value for this dataset. Conclusion: EfficientNet-B3 outperformed SPCNN and custom CNNs. The CLAHE/brightness strategy applied to 20% of training images and added back to the dataset did not significantly improve generalization. External validation and prospective field testing are necessary before clinical deployment.

Název v anglickém jazyce

Comparative Evaluation of Custom Convolutional Neural Networks and EfficientNet-B3 for Malaria Cell Image Classification: Impact of Targeted Data Augmentation on Model Performance

Popis výsledku anglicky

Background: Malaria diagnosis with thin blood smears remains labor-intensive and relies on the operator. Deep learning could enable accurate automation. Objective: Compare four convolutional approaches for classifying parasitized versus uninfected erythrocytes and to evaluate whether targeted image-quality augmentations enhance performance. Materials and Methods: We used the balanced NIH/Kaggle dataset, which included 13,780 parasitized and 13,780 uninfected samples. Data were split stratified into training, validation, and test sets (70/15/15). Images were resized to 256x256 and normalized. Four experiments were conducted: (1) a custom CNN; (2) the same CNN with targeted augmentation applied to 20% of training samples per class-using Contrast Limited Adaptive Histogram Equalization [CLAHE] and controlled brightness adjustment-and augmented images were added back to the training set (totaling 30,864 images); (3) a soft-attention parallel CNN (SPCNN); and (4) transfer learning with EfficientNet-B3 on 300x300 inputs with full fine-tuning. Evaluation metrics included accuracy, precision, recall, F1 score, and AUC-ROC. Results: EfficientNet-B3 achieved the highest performance with a validation accuracy of 0.9741, 98% precision, 96% recall, an F1 score of 0.97, and an AUC-ROC of 0.9964. SPCNN was competitive but slightly lower, with a validation accuracy of 0.9652, 98% precision, 95% recall, an F1 score of 0.96, and an AUC-ROC of 0.9909. The baseline CNN had a validation accuracy of 0.9649, 97% precision, 94% recall, an F1 score of 0.96, and an AUC-ROC of 0.9910. Targeted augmentation resulted in negligible change compared to the baseline CNN, with a validation accuracy of 0.9647, an F1 score of 0.96, and an AUC-ROC of 0.9908, indicating limited added discriminative value for this dataset. Conclusion: EfficientNet-B3 outperformed SPCNN and custom CNNs. The CLAHE/brightness strategy applied to 20% of training images and added back to the dataset did not significantly improve generalization. External validation and prospective field testing are necessary before clinical deployment.

Druh

J_SC - Článek v periodiku v databázi SCOPUS

CEP obor

—

OECD FORD obor

30502 - Other medical science

Projekt

—

Návaznosti

V - Vyzkumna aktivita podporovana z jinych verejnych zdroju

Rok uplatnění

2025

Kód důvěrnosti údajů

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

Název periodika

Iranian Journal of Blood and Cancer

ISSN

2008-4595

e-ISSN

2008-4609

Svazek periodika

17

Číslo periodika v rámci svazku

3

Stát vydavatele periodika

IR - Íránská islámská republika

Počet stran výsledku

11

Strana od-do

62-72

Kód UT WoS článku

—

EID výsledku v databázi Scopus

2-s2.0-105021799286