EUVIMEDEUVIMED entdecken
Crossref

ENHANCING BRAIN TUMOR CLASSIFICATION FROM MRI USING GENERATIVE DATA AUGMENTATION

Kolli.Bala Bhaskara Rao, R.VIJAYA KUMAR REDDY

International Journal of Drug Delivery Technology · 04.04.2026 · 16 (88s)

DOI: 10.25258/ijddt.16.89s.2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 04.04.2026. Verfasst von Kolli.Bala Bhaskara Rao und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In medical diagnostics, DL based models often fail to perform well because of the high level of class imbalance in the training set especially if one is dealing with a rare pathology. Such inner data bias can result in biased classifiers and sub-optimal generalisation. In this article, we propose a data-generative approach of dataaugmentation to address this constraint in the case of brain-tumor detection utilising MRI. A conditional GAN was built and trained to produce high fidelity T1-weighted images with contrast agent. Using the 3,064 photos in three categories of the Figshare Brain Tumour Dataset, the network was able to learn the morphological traits unique to different types of tumours. The trained GAN was used to create new plausible samples, notably of under-represented tumour types, leading to a more balanced training corpus. A second independent deep-learning classification model, trained with the synthetically-augmented data, achieved 96% accuracy in diagnosing a different test set of original, unseen, photos. The results support the pragmatic effectiveness of GAN-based data synthesis as a formidable weapon of battling the illogical outputs of data-paucity and for greatly boosting the trustworthiness and predictability of the automated-systems of creating medical-imaging data. “Keywords: Brain Tumor Classification, Data Augmentation, Generative Adversarial Networks (GANs), Medical Imaging, Deep Learning, Class Imbalance, MRI."

Zitieren

APA

Rao, K. B., & REDDY, R. K. (2026). ENHANCING BRAIN TUMOR CLASSIFICATION FROM MRI USING GENERATIVE DATA AUGMENTATION. International Journal of Drug Delivery Technology, 16(88s). https://doi.org/10.25258/ijddt.16.89s.2

Vancouver

Rao KB, REDDY RK. ENHANCING BRAIN TUMOR CLASSIFICATION FROM MRI USING GENERATIVE DATA AUGMENTATION. International Journal of Drug Delivery Technology. 2026;16(88s). doi:10.25258/ijddt.16.89s.2

BibTeX
@article{Rao2026,
  author = {Rao, Kolli.Bala Bhaskara and REDDY, R.VIJAYA KUMAR},
  title = {ENHANCING BRAIN TUMOR CLASSIFICATION FROM MRI USING GENERATIVE DATA AUGMENTATION},
  journal = {International Journal of Drug Delivery Technology},
  year = {2026},
  volume = {16},
  number = {88s},
  doi = {10.25258/ijddt.16.89s.2}
}
Kostenlos starten →