LighTDiff: Surgical Endoscopic Image Low-Light Enhancement with T-Diffusion

TL;DR

LighTDiff is a lightweight diffusion-based model designed for low-light endoscopic image enhancement, combining a T-shape architecture, a Temporal Light Unit, and a Chroma Balancer to improve performance and efficiency in medical imaging.

Contribution

The paper introduces LighTDiff, a novel lightweight DDPM tailored for endoscopic images, incorporating a T-shape architecture, TLU, and CB to enhance stability, efficiency, and spectral fidelity.

Findings

LighTDiff outperforms existing LLIE methods in quality and speed.

The TLU improves training stability and denoising performance.

The Chroma Balancer reduces spectral shifts during image recovery.

Abstract

Advances in endoscopy use in surgeries face challenges like inadequate lighting. Deep learning, notably the Denoising Diffusion Probabilistic Model (DDPM), holds promise for low-light image enhancement in the medical field. However, DDPMs are computationally demanding and slow, limiting their practical medical applications. To bridge this gap, we propose a lightweight DDPM, dubbed LighTDiff. It adopts a T-shape model architecture to capture global structural information using low-resolution images and gradually recover the details in subsequent denoising steps. We further prone the model to significantly reduce the model size while retaining performance. While discarding certain downsampling operations to save parameters leads to instability and low efficiency in convergence during the training, we introduce a Temporal Light Unit (TLU), a plug-and-play module, for more stable training and better performance. TLU associates time steps with denoised image features, establishing temporal dependencies of the denoising steps and improving denoising outcomes. Moreover, while recovering images using the diffusion model, potential spectral shifts were noted. We further introduce a Chroma Balancer (CB) to mitigate this issue. Our LighTDiff outperforms many competitive LLIE methods with exceptional computational efficiency.