Ph.D. Candidate · NCCA, Bournemouth University · MSCA Fellow

Armin

Sheibanifard

AI researcher specialising in implicit neural representations, transformer-based architectures, and representation learning — applied to 3D medical imaging for compression, reconstruction, segmentation, and super-resolution.

About

I am a Marie Skłodowska-Curie Actions fellow pursuing a fully funded Ph.D. at the National Centre for Computer Animation (NCCA), Bournemouth University. My research designs end-to-end machine learning pipelines for 3D medical image compression and reconstruction — integrating implicit neural representations (INRs), 3D Gaussian Splatting, and vision transformers. I am interested in robust, generalisable AI systems, online adaptation, and efficient model design for real-world clinical deployment.

Publications

An End-to-End Implicit Neural Representation Architecture for Medical Volume Data

Sheibanifard, A., Yu, H., Ruan, Z., Zhang, J.J.

PLOS ONE, 20(1): e0314944 · 2025

Integrated pipeline combining downsampling, INR, and super-resolution for multi-parametric MRI. Achieved 97.5% compression ratio with 40.05 dB PSNR and 0.96 SSIM.

INR Compression MRI DOI ↗

A Novel Implicit Neural Representation for Volume Data

Sheibanifard, A., Yu, H.

Applied Sciences, 13(3242) · 2023

INR framework integrating Lanczos downsampling, SIREN-based networks, and SRDenseNet — reducing training time and GPU memory while improving compression and reconstruction quality.

SIREN Super-Resolution Volumetric DOI ↗

Novel XAI Method & Explainable AI for Implicit Neural Representations

Sheibanifard, A. et al.

Under Submission · 2025

Two papers in preparation: a new interpretability approach for modern ML models, and explainability techniques applied to INR-based models to support transparency of learned representations.

XAI Interpretability In Preparation

Skills

Research Expertise

Implicit Neural Representations 3D Gaussian Splatting Vision Transformers (ViT, Swin) Medical Image Compression Reconstruction & Segmentation X-ray Enhancement Multi-Modal Medical Imaging

ML Competencies

Self-Supervised Learning Contrastive Learning Continual & Online Learning Few-Shot Learning Semi-Supervised Learning Deep Representation Learning

Technical Stack

Python PyTorch TensorFlow MONAI OpenCV SimpleITK NiBabel Docker CUDA Git Weights & Biases scikit-learn

Honours & Awards

Marie Skłodowska-Curie Fellowship (MSCA)

2025

Competitive EU funding for the final year of PhD research in implicit neural representations and 3D medical imaging.

Doctoral College Outstanding Contribution Award

2024

Recognised by Bournemouth University for service and community leadership within the Doctoral College.

Doctoral College Seminar Grant — Write Wise & AI in Research

2024

Organised two seminar series on academic writing, effective media communication, domain-tailored AI, and future AI trends.

Fully Funded Ph.D. Studentship — NCCA, Bournemouth University

2022–Now

Three-year full scholarship for doctoral studies in AI and 3D medical imaging.