Fully supervised binary classification of skin lesions from dermatoscopic images using an ensemble of diverse CNN architectures (EfficientNet-B6, Inception-V3, SEResNeXt-101, SENet-154, DenseNet-16…

Weakly supervised 3D classification of multi-disease chest CT scans using multi-resolution deep segmentation features via dual-stage CNN architecture (DenseVNet, 3D Residual U-Net).

Fully supervised, healthy/malignant prostate detection in multi-parametric MRI (T2W, DWI, ADC), using a modified 2D RetinaNet model for medical object detection, built upon a shallow SEResNet backb…

Fully supervised, multi-class 3D brain segmentation in T1 MRI, using atlas-based segmentation algorithms (label propagation, tissue models, Expectation-Maximization algorithm).

Fully supervised, multi-class 3D brain segmentation in T1 MRI using an ensemble of diverse CNN architectures (3D FCN, 3D U-Net) with multi-scale input.

Unsupervised region proposal and supervised patch extraction algorithms for extracting candidate 2D ROIs to train SVM/CNN classifiers, for mass detection in mammograms.

Fully supervised binary classification of skin lesions from dermatoscopic images using multi-color space moments/texture features and Support Vector Machines/Random Forests.