Mehrnaz Hosseininejad

• Evaluate generative MRI inpainting models from the BraTS challenge, generating inpainted volumes and testing them on external datasets. • Benchmark reference-based (pixel-wise, perceptual) against distribution-based image quality metrics, and analyze how well each correlates with downstream task performance. • Work with 3D brain MRI data in PyTorch on HPC/Kubernetes infrastructure.

Teaching Python and MATLAB programming, image processing, and machine learning

Focused on machine learning, neural networks, computer vision, database systems, stochastic processes, network science, and adversarial machine learning. Projects: • Designed and developed “DonorDash,” a food donation database system aimed at reducing food waste through coordination between restau

Focused on biomedical signal and image processing, biological modeling, pattern recognition, data mining, neural networks, and medical imaging. Thesis: Segmentation of CT images to improve contouring accuracy in pelvic radiotherapy treatment planning.