Interactive Shape Sonification for Tumor Localization in Breast Cancer Surgery

TL;DR

Breast-conserving surgery suffers from substantial reoperation due to positive margins. The authors introduce shape sonification to encode both tumor margin and seed location, enabling simultaneous auditory guidance. Across three iterative studies with 12 volunteers and 4 breast surgeons (totaling 33 non-clinical participants and 4 clinicians), shape-based sonifications markedly improve localization accuracy (Dice coefficient rising from ~0.44 with Beep to ~0.74 with Sine in the surgeon study) and, in many cases, usability (SUS up to ~85 with Sine); some measures show no significant change due to added cognitive load and latency. The work demonstrates the practical potential of motion- and shape-encoded auditory displays to augment surgical precision and could inform next-generation multimodal guidance systems for tumor localization and margin assessment, with broader implications for sensory substitution and AR/MR interfaces. $p<0.001$ in several comparisons underlines the robustness of the reported improvements.

Abstract

About 20 percent of patients undergoing breast-conserving surgery require reoperation due to cancerous tissue remaining inside the breast. Breast cancer localization systems utilize auditory feedback to convey the distance between a localization probe and a small marker (seed) implanted into the breast tumor prior to surgery. However, no information on the location of the tumor margin is provided. To reduce the reoperation rate by improving the usability and accuracy of the surgical task, we developed an auditory display using shape sonification to assist with tumor margin localization. Accuracy and usability of the interactive shape sonification were determined on models of the female breast in three user studies with both breast surgeons and non-clinical participants. The comparative studies showed a significant increase in usability (p<0.05) and localization accuracy (p<0.001) of the shape sonification over the auditory feedback currently used in surgery.

Figures (8)

• Figure 1: Experiment setup and sonifications used in the three user studies. The setup included the transmitter of the trakSTAR™ electromagnetic (EM) tracking system, the Savi Scout® radar localization probe tracked using an EM sensor, a three-dimensional (3D) printed holder for the EM sensor, and a pen. Study 1 used a 3D printed board to hold the paper in place. Study 2 and 3 used a 3D printed board for holding the breast models in place.
• Figure 2: Sonification models in Study 1. Sounds indicating the seed location are colored in yellow, sounds indicating the tumor shape are colored in blue: (a) Beep (clinical status quo) and the newly proposed (b) Rhythm and (c) Synth shape sonifications are compared. (a) Beep uses a beeping sound (yellow) varying in frequency. (b) Rhythm uses one beating sound: the seed is represented by a xylophone (yellow), the tumor by a marimba (blue). The Rhythm sonification interpolates between the two instruments for the area between the tumor margin and the seed. (c) Synth uses two sounds to distinguish between margin and seed location: a continuous synthesizer sound (blue) represents the tumor margin, and a ticking sound (yellow) represents the seed. No interpolation is applied. The sounds are distinct.
• Figure 3: Box plots on localization accuracy in Study 1. *** = p < 0.001
• Figure 4: Display of the segmented ground truth, drawn shape, and their overlay from one of the user experiments guided by the Synth sonification model. In the overlay image, the ground truth appears in white and pink, while the drawn shape appears in white and green. For the above case, the Sørensen–Dice coefficient was 0.92.
• Figure 5: Sonification models in Study 2 and 3; sounds indicating the seed location are colored in yellow, the sound indicating the tumor shape is colored in blue: (a) Beep (clinical status quo) and (b) Sine (new shape sonification)