# Enhancing Tip Detection by Pre-Training with Synthetic Data for Ultrasound-Guided Intervention

## TL;DR

This paper introduces a new method for improving tip detection in ultrasound-guided procedures by using synthetic data generated with deep learning models.

## Contribution

The novel approach uses synthetic ultrasound puncture data generated with a diffusion model to pre-train tip detectors, enhancing generalization without expert annotations.

## Key findings

- The proposed method outperforms MSCOCO pre-training on a clinical puncture dataset with improvements of 1.27–7.19% in AP0.1:0.5.
- State-of-the-art detectors also show performance gains of 1.14–1.76% when using the proposed pre-training method.
- The method enhances generalization without relying on expert annotations or large real data sets.

## Abstract

Objectives: Automatic tip localization is critical in ultrasound (US)-guided interventions. Although deep learning (DL) has been widely used for precise tip detection, existing methods are limited by the availability of real puncture data and expert annotations. Methods: To address these challenges, we propose a novel method that uses synthetic US puncture data to pre-train DL-based tip detectors, improving their generalization. Synthetic data are generated by fusing clinical US images of healthy controls with tips created using generative DL models. To ensure clinical diversity, we constructed a dataset from scans of 20 volunteers, covering 20 organs or anatomical regions, obtained with six different US machines and performed by three physicians with varying expertise levels. Tip diversity is introduced by generating a wide range of synthetic tips using a denoising probabilistic diffusion model (DDPM). This method synthesizes a large volume of diverse US puncture data, which are used to pre-train tip detectors, followed by subsequently training with real puncture data. Results: Our method outperforms MSCOCO pre-training on a clinical puncture dataset, achieving a 1.27–7.19% improvement in AP0.1:0.5 with varying numbers of real samples. State-of-the-art detectors also show performance gains of 1.14–1.76% when applying the proposed method. Conclusions: The experimental results demonstrate that our method enhances the generalization of tip detectors without relying on expert annotations or large amounts of real data, offering significant potential for more accurate visual guidance during US-guided interventions and broader clinical applications.