Table 30 GAN-augmented data analysis.

From: A federated learning-based privacy-preserving image processing framework for brain tumor detection from CT scans

Model variant

Accuracy

Precision

F1-score

Sensitivity

Specificity

MCC

SSIM

PSNR (dB)

Baseline (real data only)

0.9453

0.9385

0.9412

0.9361

0.9498

0.892

0.89

24.6

CycleGAN-Augmented

0.9611

0.9547

0.9582

0.9523

0.9684

0.921

0.93

26.9

CycleGAN (no SSIM filtering)

0.939

0.9302

0.9335

0.9281

0.9435

0.873

0.84

21.7

GAN-Augmented + Grad-CAM QA

0.9645

0.9589

0.9613

0.9572

0.9701

0.926

0.94

27.4

Back to article page