Deep Learning for MRI Slice Interpolation: The Critical Role of Problem Formulation
For medical imaging researchers and practitioners, this work highlights that problem formulation can have a far greater impact than architectural complexity, providing a simple yet effective guideline for improving through-plane resolution.
This paper investigates deep learning for MRI slice interpolation in prostate imaging, finding that reformulating the task to use adjacent slices (i-1, i+1) instead of distant slices (i-2, i+2) yields a 58% improvement in SSIM, with the U-Net achieving the best results (PSNR 30.08 dB, SSIM 0.898), outperforming linear interpolation by 10.1%.
Through-plane resolution in clinical MRI is typically much coarser than in-plane resolution, limiting diagnostic utility. This work investigates deep learning approaches to interpolate intermediate MRI slices in prostate imaging, effectively doubling through-plane resolution. I evaluated five architectures (CNN, U-Net, two GAN variants, and DDPM) and discovered that problem formulation has dramatically more impact than architectural complexity. By reformulating the interpolation task to use adjacent slices (i-1, i+1) rather than distant slices (i-2, i+2), I achieved a 58% improvement in SSIM performance across all deterministic architectures. The U-Net model achieved the best results with PSNR of 30.08 dB and SSIM of 0.898, representing a 10.1% improvement over linear interpolation baseline. A DDPM was also evaluated but showed poor reconstruction quality due to fundamental mismatch between stochastic generation and deterministic reconstruction requirements. These findings demonstrate that problem formulation can have 290x more impact than architectural sophistication in medical imaging tasks.