A new algorithm for 3D registration and its application in self-monitoring and early detection of lymphedema

Early detection and management of lymphedema (LE) can significantly reduce the potential of new symptoms and complications. In fact, effective diagnostic of LE can potentially affect the lives in the U.S. of nearly 500 000 current LE patients and over 2.4 million breast cancer survivors who are at-risk for developing this disease at some point in their lives. However, many cancer patients fail to seek medical assistance at the first sign of the disease in part due to the additional burden that constant monitoring imposes on patients. In that sense, a self-monitoring system could represent a major improvement in health-care management and delivery.The main challenges in self monitoring limb volume are in the design of a system that is at the same time inexpensive, easy to use, and accurate, despite being operated by a person without any training. In this paper, we present such a system. The proposed framework relies on off-the-shelf video gaming devices – the Kinect infrared device with a set of inertial sensors. In order to achieve high accuracy despite the typical low-texture and smoothness of the human skin, a new algorithm for 3D registration of clouds of points – Iterative Clustered Closest Points (ICCP) – is also proposed. The final result is a device that can be operated in the comfort of the patient's homes – i.e. it can be operated by professionals as well as non-professionals.In order to validate the system, we first tested its individual parts, more specifically the proposed ICCP algorithm for 3D registration and reconstruction of challenging objects. Then, we tested the complete system for the target application of limb-volume measurement by comparing our system against the gold standards: water displacement and perometer.