Surface‐guided analysis of breast shape changes during postoperative radiotherapy by using a functional map framework

BACKGROUND: While positioning errors during radiotherapy are increasingly well-characterized and accounted for in the Planning Target Volume (PTV), little is known about the shape and volume changes of the breast in the course of radiotherapy. Yet they represent an unknown factor that may impact treatment effectiveness, highlighting the need to analyze inter-fractional breast deformations to enable promising approaches such as patient-specific adaptive irradiation planning. PURPOSE: In this study, we develop a geometric approach to analyze inter-fractional breast deformation throughout the radiotherapy treatment, based on a dataset of 3D surface scans of patients' frontal torso acquired during radiotherapy sessions. METHODS: We used a handheld 3D range scanner to collect surface scans of 22 patients prior to, during, and after radiotherapy sessions following breast-conserving surgery (lumpectomy). We begin by adapting functional map framework to compute inter- and intra-patient nonrigid correspondences. To study shape changes during radiotherapy, we then perform both intrinsic and extrinsic analyses of the collected breast surface shapes, by leveraging the intra- and inter-patient correspondences. These analysis provide complementary insights into how the breast changes over time: The intrinsic analysis evaluates deformation patterns directly on the shape manifolds, enabling the identification of regions exhibiting high conformal (bending) or area distortions (stretching). We also perform extrinsic analysis, where we align surface acquisitions of the treated breast with the CT-derived skin surface to assess pointwise surface displacements and volume changes in the treated area, thereby producing quantitative measures such as displacement amplitudes (in mm) and relative volume changes (in percentage). RESULTS: The qualitative shape collection analysis highlight deformations in the contra-lateral breast and armpit areas, along with positioning shifts on the head or abdominal regions. On average, displacements within the treated breast exhibit amplitudes of 1-2 mm across sessions, with higher values observed at the time of the 25th irradiation session. Volume changes, inferred from surface variations, reached up to 10% over the course of treatment, with the majority of values ranging between 2% and 5%. CONCLUSIONS: A -based geometric approach has been proposed for analyzing breast shape changes over the course of radiotherapy from surface data acquisitions. Based on a functional map framework, we develop a shape matching specialized for the breast data as well as a shape variability analysis within a unified space. Analysis across multiple patient datasets revealed a wide spectrum of breast changes despite a clinically acceptable quantitative average, demonstrating the power of our method to perform both quantitative and qualitative analysis of deformation patterns both on the individual and group level. TRIAL REGISTRATION: The clinical trial data used in this paper is registered under the ClinicalTrials.gov ID NCT03801850.