A fully automatic framework for evaluating cosmetic results of breast conserving therapy.

The breast cosmetic outcome after breast conserving therapy is essential for evaluating breast treatment and determining patient's remedy selection. This prompts the need of objective and efficient methods for breast cosmesis evaluations. However, current evaluation methods rely on ratings from a small group of physicians or semi-automated pipelines, making the processes time-consuming and their results inconsistent. To solve the problem, in this study, we proposed: 1. a fully-automatic Machine Learning Breast Cosmetic evaluation algorithm leveraging the state-of-the-art Deep Learning algorithms for breast detection and contour annotation, 2. a novel set of Breast Cosmesis features, 3. a new Breast Cosmetic dataset consisting 3k+ images from three clinical trials with human annotations on both breast components and their cosmesis scores. We show our fully-automatic framework can achieve comparable performance to state-of-the-art without the need of human inputs, leading to a more objective, low-cost and scalable solution for breast cosmetic evaluation in breast cancer treatment.

Adding Knowledge to Unsupervised Algorithms for the Recognition of Intent.

Computer vision algorithms performance are near or superior to humans in the visual problems including object recognition (especially those of fine-grained categories), segmentation, and 3D object reconstruction from 2D views. Humans are, however, capable of higher-level image analyses. A clear example, involving theory of mind, is our ability to determine whether a perceived behavior or action was performed intentionally or not. In this paper, we derive an algorithm that can infer whether the behavior of an agent in a scene is intentional or unintentional based on its 3D kinematics, using the knowledge of self-propelled motion, Newtonian motion and their relationship. We show how the addition of this basic knowledge leads to a simple, unsupervised algorithm. To test the derived algorithm, we constructed three dedicated datasets from abstract geometric animation to realistic videos of agents performing intentional and non-intentional actions. Experiments on these datasets show that our algorithm can recognize whether an action is intentional or not, even without training data. The performance is comparable to various supervised baselines quantitatively, with sensible intentionality segmentation qualitatively.