8 papers
BreastMammo and DenseMammo: Benchmarks for Mammography Domain Generalization
Hongyi Pan, Gorkem Durak, Halil Ertugrul Aktas +18
Breast density classification is a critical component of breast cancer risk assessment, yet AI models often struggle to generalize across clinical sites due to vendor-specific acqu…
HamVision: Hamiltonian Dynamics as Inductive Bias for Medical Image Analysis
Mohamed A Mabrok
We present HamVision, a framework for medical image analysis that uses the damped harmonic oscillator, a fundamental building block of signal processing, as a structured inductive…
Latent Semantic Manifolds in Large Language Models
Mohamed A. Mabrok
Large Language Models (LLMs) perform internal computations in continuous vector spaces yet produce discrete tokens -- a fundamental mismatch whose geometric consequences remain poo…
Controllability Analysis of State Space-based Language Model
Mohamed Mabrok, Yalda Zafari
State-space models (SSMs), particularly Mamba, have become powerful architectures for sequence modeling, yet their internal dynamics remain poorly understood compared to attention-…
X-VMamba: Explainable Vision Mamba
Mohamed A. Mabrok, Yalda Zafari
State Space Models (SSMs), particularly the Mamba architecture, have recently emerged as powerful alternatives to Transformers for sequence modeling, offering linear computational…
MammoClean: Toward Reproducible and Bias-Aware AI in Mammography through Dataset Harmonization
Yalda Zafari, Hongyi Pan, Gorkem Durak +3
The development of clinically reliable artificial intelligence (AI) systems for mammography is hindered by profound heterogeneity in data quality, metadata standards, and populatio…