An accurate and interpretable model for BCCT.core

Abstract

Abstract-Breast Cancer Conservative Treatment (BCCT) is considered nowadays to be the most widespread form of locorregional breast cancer treatment. However, aesthetic results are heterogeneous and difficult to evaluate in a standardized way. The limited reproducibility of subjective aesthetic evaluation in BCCT motivated the research towards objective methods. A recent computer system (BCCT.core) was developed to objectively and automatically evaluate the aesthetic result of BCCT. The system is centered on a support vector machine (SVM) classifier with a radial basis function (RBF) used to predict the overall cosmetic result from features computed on a digital photograph of the patient. However, this classifier is not ideal for the interpretation of the factors being used in the prediction. Therefore, an often suggested improvement is the interpretability of the model being used to assess the overall aesthetic result.In the current work we investigate the accuracy of different interpretable methods against the model currently deployed in the BCCT.core software. We compare the performance of decision trees and linear classifiers with the RBF SVM currently in BCCT.core. In the experimental study, these interpretable models shown a similar accuracy to the currently used RBF SVM, suggesting that the later can be replaced without sacrificing the performance of the BCCT.core.

Evaluation of Aesthetic and Quality-of-Life Results after Immediate Breast Reconstruction with Definitive Form-Stable Anatomical Implants

“…Results are classified automatically as excellent, good, fair, and poor. 11,13 Despite the fact that it was designed for breast-conserving treatment evaluation, one study found a statistically significant correlation between the BCCT. core and the Harris Scale for measuring aesthetic outcomes in unilateral breast reconstruction after mastectomy.…”

“…The final results were displayed by means of a four-point scale: excellent, good, fair, and poor. [11][12][13] To evaluate quality-of-life, all patients answered the BREAST-Q questionnaires, which were previously validated for application in Brazilian patients. 14 The BREAST-Q patient-reported outcome instrument breast reconstruction postoperative module comprises two domains: satisfaction domains (i.e., satisfaction with breasts, satisfaction with outcomes, and satisfaction with care) and quality-of-life domains (psychosocial well-being, physical well-being, and sexual well-being), which consist of nine scales.…”

Improving the BCCT.core model with lateral information

“…In a previous work [8], an accurate and interpretable model for the BeeT.core was found. The objective of that work was to compare between the previous model used on the evalu ation of the overall aesthetic result of the BeeT [6], with other methods in order to overcome the non-interpretability of the older one.…”

“…1), selecting the aspects with most impact to the overall cosmetic result: breast asymmetry, skin colour changes due to the radiotherapy treatment and surgical scar appearance. Posteriorly, a Support Vector Machine (SVM) classifier was trained to predict the overall cosmetic result from the recorded features [6], [8]. BeeT.core classifies the aesthetic outcome of BeeT into excellent, good, fair, and poor classes.…”

A 3D low-cost solution for the aesthetic evaluation of breast cancer conservative treatment

“…The selection of 2D features was based on previous studies Oliveira et al 2010), whereas the 3D features were those presented in this work. The model was designed using the selected features, considering all the possible subsets, using a leave-one-out scheme (Duda et al 2000).…”