About Me
I am Shayan Kebriti, a Computer Engineering undergraduate at Shahid Beheshti University in Tehran, Iran.
My research interests include Machine Learning, Computer Vision, Medical Image Computing, and Computational Modeling.
I am open to working on projects in these areas and I would love to contribute. I am also seeking Master’s or PhD opportunities.
Email: shayankebriti@gmail.com
Publications
FractMorph: A Fractional Fourier-Based Multi-Domain Transformer for Deformable Image Registration
FractMorph is a novel 3D dual-parallel transformer for deformable image registration. Its Fractional Cross-Attention blends multi-scale spectral–spatial features via fractional Fourier transforms, achieving state-of-the-art results on intra-patient cardiac and atlas-to-patient cerebral MRI benchmarks.
1 Faculty of Computer Science and Engineering, Shahid Beheshti University, Tehran, Iran
2 School of Computing Science, University of Glasgow, Glasgow, UK
3 Alan Turing Institute, London, UK
Design and Implementation of a Software for Measurement of Brain Perfusion Using an Optimized Model
This work is carried out under the supervision of Dr. Oghabian1,2 🎓 🔗, in collaboration with Dr. Mohammadi1,2 🎓 🔗, Dr. Irandoost1,2 🔗 and Dr. Zare1,2 🔗
Project Description
The NIAG perfusion software is developed for the quantitative analysis of brain perfusion MRI. It supports standard medical image formats, provides preprocessing tools such as motion correction, denoising, segmentation, and slice-time correction, and computes key DSC perfusion parameters (CBF, CBV, MTT, AUC). It also offers advanced visualization features like mirror-mode maps and customizable ROI definition for detailed regional assessment of cerebral hemodynamics. The software is intended for real clinical use in the most reputable hospitals in Iran, under the supervision of Tehran University of Medical Sciences.
My Contributions
My current contributions have focused on enhancing the software’s analytical pipeline, including the refinement of brain segmentation methods for complex tumor cases, improvements in preprocessing routines, and extension of manual ROI handling and visualization modules. I have also increased the quality and resolution of perfusion MRI images while ensuring precise metric calculations, and carried out the implementation and debugging of perfusion parameter calculations.
1 Department of Medical Physics and Biomedical Engineering, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran
2 Neuroimaging and Analysis Group, Research Center for Molecular and Cellular Imaging, Advanced Medical Technologies and Equipment Institute, Tehran University of Medical Sciences, Tehran, Iran
Bachelor Thesis
A Framework for Anatomical Mesh Reconstruction From Medical Images for In-Silico Simulations
This ongoing work aims to develop a novel deep learning framework for anatomical mesh generation from medical images without requiring manual annotations. The framework consists of an anatomical structure segmentation network, GNN-based mesh control-handles displacement prediction, and a final deformation stage. The generated meshes are designed for use in in-silico simulations for multiple modalities such as CT, MR, and US scans, enabling virtual experiments and computational modeling of anatomy for surgical planning, medical device testing, and further researches. More updates coming soon!
Projects
You can browse a categorized list of my open-source projects on my Github profile. Here is a selection of them:
MedMesh: An easy to use, lightweight tool for converting medical images into 3D meshes
GitHub
Multi-Asset Trading with TD3 RL Agent and Wavelet-Coherence Graph Neural Networks
GitHub
Conditional Diffusion with Simple and Attention-Based U-Nets for CIFAR-10 Image Generation
GitHub
HIV Inhibitors Classification with Multiple GNN Message Passing Methods
GitHub
Rock–Paper–Scissors Automation with YOLOv11 and Cheat/Win Overlays
GitHub
Variational Autoencoder (VAE) Analysis & Denoising GAN
GitHub
Fashion Product Multilabel Classification & Image-to-Text Title Generation Using RNNs
GitHub
Siamese Face Recognition Network Featuring A Graphical Web Interface
GitHub
A puzzle-solving game implementing DFS, BFS, and A* search agents (designed as a TA assignment)
GitHub
Image Compressing & Watermarking Using Spatial and Spectral (DCT, DFT and DWT) Methods
GitHub
Classical Image Processing & Non Local Means (NLM) Filter CPU and GPU Implementations
GitHub
My Deep Reinforcement Learning Assignments
GitHub
My Machine Learning Assignments
GitHubEducation
Bachelor, Computer Engineering, Shahid Beheshti Univeristy, Tehran, Iran
Diploma, Mathematics & Physics, National Organization for Development of Exceptional Talents (Sampad), Iran
Experience
For additional details about each experience, please see my CV or contact me.
Endorsements
Here is what colleagues and supervisors have said about working with me:
“It is my pleasure to strongly recommend Mr. Shayan Kebriti, one of my undergraduate students, for his exceptional academic performance, research contributions, and outstanding personal qualities. I had the privilege of teaching Shayan in two courses: Computer Vision with Deep Learning, in which he achieved a perfect score of 20 out of 20, and Operating Systems Lab, where he ranked in the top 5% of the class. His performance in both courses demonstrated not only his deep understanding of the material but also his remarkable ability to apply theoretical concepts to practical challenges. Beyond coursework, Shayan worked under my supervision on a research project in deformable image registration, which is currently under review in a reputable journal. His contributions to this work were significant, showing creativity, strong problem-solving skills, and a clear passion for advancing the field of artificial intelligence and medical image computing. Shayan is a delighted, talented, and hardworking individual who consistently surprises me with his intellectual curiosity and eagerness to learn new concepts. He is particularly enthusiastic about exploring new developments in AI and medical imaging, a quality that makes him an ideal candidate for advanced research and professional growth. On a personal level, Shayan is polite, understanding, and highly reliable. His positive attitude, professionalism, and strong interpersonal skills make him a pleasure to work with, both in academic and research settings.
In summary, I have no doubt that Shayan will excel in any academic or professional environment he chooses to pursue. I wholeheartedly recommend him and am confident that he will make valuable contributions wherever he goes.”
“I have known Shayan since 2024 and have worked with him on NIAG’s perfusion MRI toolkit. As a computer engineering student, he showed exceptional eagerness to master medical imaging with a strong focus on perfusion MRI and became productive very quickly. He is organized and meticulous, with sharp attention to detail. His reports are clear, thorough, and actionable. Shayan communicates effectively with clinicians, actively seeks feedback, and turns it into practical improvements. He brings strong image processing expertise, programming skills, and the ability to research imaging methods and adapt them to clinical applications. He applies this knowledge creatively to solve difficult cases while maintaining scientific rigor. He is always on time, consistent in his execution, and shows strong responsibility for his work. I recommend Shayan for roles in medical image computing and I am confident he will elevate any project or team he joins.”
“Working with Shayan from 2022-2024 on our personalized guide creation app, I witnessed a software engineer who excelled at turning complex challenges into elegant solutions through creative problem-solving and systematic organization. His pre-AI era code showcased genuine innovation and craftsmanship, while his passionate advocacy for agile practices and thorough code reviews elevated our entire team's performance. Shayan's rare combination of technical creativity, structured execution, and unwavering reliability made him the engineer we trusted with our most critical features—any team would be fortunate to have him.”
Courses
Hobbies
The motto I live by is “Carpe Diem”, which means “Seize the Day”.
I believe the most valuable asset we have is the present moment, which is why I consider myself adventurous in all aspects of life, from work to personal experiences. I almost never say no to new opportunities, experiences, or challenges.
In my free time, I love doing sports, going trekking in nature, and capturing photographs.
Here are some of my works:
In case you’re curious, you can explore more of my adventures and photography on Unsplash.