Surface guided analysis of breast changes during post-operative radiotherapy by using a functional map framework

TL;DR

This work addresses non-rigid breast deformations during post-operative radiotherapy by bringing a surface-based, functional-map framework to bear on 3D scans collected throughout treatment. It introduces a complete workflow that constructs intra- and inter-patient correspondences, aligns surface data to planning CT, and analyzes shape changes with intrinsic (area/conformal) and extrinsic (displacement/volume) metrics. A key contribution is the Cross-Collection Functional Map Network and Global Latent Bases, enabling robust, scalable comparisons across many patients and sessions within a shared spectral domain. The approach is validated on a clinical dataset of hundreds of torso surface shapes, revealing non-negligible breast shape and volume changes that could inform personalized radiotherapy planning and future SGRT-driven dose adaptation.

Abstract

The treatment of breast cancer using radiotherapy involves uncertainties regarding breast positioning. As the studies progress, more is known about the expected breast positioning errors, which are taken into account in the Planning Target Volume (PTV) in the form of the margin around the clinical target volume. However, little is known about the non-rigid deformations of the breast in the course of radiotherapy, which is a non-negligible factor to the treatment. Purpose: Taking into account such inter-fractional breast deformations would help develop a promising future direction, such as patient-specific adjustable irradiation plannings. Methods: In this study, we develop a geometric approach to analyze inter-fractional breast deformation throughout the radiotherapy treatment. Our data consists of 3D surface scans of patients acquired during radiotherapy sessions using a handheld scanner. We adapt functional map framework to compute inter-and intra-patient non-rigid correspondences, which are then used to analyze intra-patient changes and inter-patient variability. Results: The qualitative shape collection analysis highlight deformations in the contralateral breast and armpit areas, along with positioning shifts on the head or abdominal regions. We also perform extrinsic analysis, where we align surface acquisitions of the treated breast with the CT-derived skin surface to assess displacements and volume changes in the treated area. On average, displacements within the treated breast exhibit amplitudes of 1-2 mm across sessions, with higher values observed at the time of the 25 th irradiation session. Volume changes, inferred from surface variations, reached up to 10%, with values ranging between 2% and 5% over the course of treatment. Conclusions: We propose a comprehensive workflow for analyzing and modeling breast deformations during radiotherapy using surface acquisitions, incorporating a novel inter-collection shape matching approach to model shape variability within a i shared space across multiple patient shape collections. We validate our method using 3D surface data acquired from patients during External Beam Radiotherapy (EBRT) sessions, demonstrating its effectiveness. The clinical trial data used in this paper is registered under the ClinicalTrials.gov ID NCT03801850.

Figures (16)

• Figure 1: Illustration of our clinical trial data acquired for each patient. Optical scans were captured as textured 3D meshes during the trial (shown below). The surfaces $A$ and $B$ have been acquired on the same day as the planning $CT$ scan, while $S_X$ has been acquired during the $X^{th}$ irradiation session. Finally $M_1$ and $M_3$ have been acquired one month and three months after the last irradiation. In addition to the clinical trial surfaces, we also consider the point clouds representing the patient skin (blue) and treated breast (pink), annotated from the CT scan (shown at the top).
• Figure 2: Schematic illustration of our approach and its components. From the clinical dataset, we first compute correspondences between the scans of the same patient focusing on the ROI (1). From these intra-patient correspondences, and derive efficiently inter-patient correspondences (2) that are used for the breast shape evolution throughout radiotherapy (3).
• Figure 3: An example of the scan-to-CT initial correspondences. Note that all scan meshes cover only the front part of the torso, except for the CT skin mesh which also includes the backside.
• Figure 4: ROI projected from the CT skin to the other acquisitions on four different patients undergoing right (two first rows) or left (two last rows) breast radiotherapy.
• Figure 5: Function Map Network (FMN) graph of patient $P^{id}$, before (left) and after (right) cycle-consistency optimization. The output after the functional maps consistency optimization is a latent space representing the functional domain of an average shape $L_{id}$ and its relations $Y_n$ to the other shapes of the collection.