SurgicaL-CD: Generating Surgical Images via Unpaired Image Translation with Latent Consistency Diffusion Models

TL;DR

SurgicaL-CD introduces a novel diffusion-based method for generating high-quality, diverse surgical images from unpaired data, aiding training of computer-assisted surgery systems without extensive labeled datasets.

Contribution

The paper presents a new consistency-distilled diffusion model that produces realistic surgical images efficiently from unpaired data, surpassing previous GAN and diffusion methods.

Findings

Outperforms GANs and diffusion approaches in image quality.

Generates diverse surgical images with minimal sampling steps.

Enhances training datasets for surgical AI applications.

Abstract

Computer-assisted surgery (CAS) systems are designed to assist surgeons during procedures, thereby reducing complications and enhancing patient care. Training machine learning models for these systems requires a large corpus of annotated datasets, which is challenging to obtain in the surgical domain due to patient privacy concerns and the significant labeling effort required from doctors. Previous methods have explored unpaired image translation using generative models to create realistic surgical images from simulations. However, these approaches have struggled to produce high-quality, diverse surgical images. In this work, we introduce \emph{SurgicaL-CD}, a consistency-distilled diffusion method to generate realistic surgical images with only a few sampling steps without paired data. We evaluate our approach on three datasets, assessing the generated images in terms of quality and utility as downstream training datasets. Our results demonstrate that our method outperforms GANs and diffusion-based approaches. Our code is available at https://gitlab.com/nct_tso_public/gan2diffusion.