Assessment of Upper Extremity Swelling Among Breast Cancer Survivors with a Commercial Infrared Sensor.

Abstract

Background: There are various methods of detecting and monitoring arm lymphedema after breast cancer (BC) treatment. Using volumetric measurements is the most common method performed clinically in the United States, and several techniques exist. Recently, the use of infrared cameras has been explored by several groups. This study shows the use of this technology and its ability to assess arm volume measurements comparable to the Perometer while being more inexpensive, having a smaller footprint in clinics, and being easy to use. Methods and Results: Seventy-three BC survivors were scanned with the Perometer and the Kinect IR system, and arm volumes of 146 arms were calculated using their respective algorithms. Good correlation in the arm volume (R-squared = 0.8799) and percent difference (R-squared range 0.6277-0.7098) in the ranges where lymphedema diagnosis is made, and good clinical agreement of lymphedema status based on commonly used thresholds were found between the two modalities. The Cohen's kappa coefficient indicates that based on a 10% arm volume difference, there was a fair agreement (κ = 0.2663), and based on a 200 mL difference, there was a moderate agreement (κ = 0.5475) between the clinical outcomes of these two modalities. The Bland-Altman analysis of volume differences indicated a bias of 6.016 mL; thus, only a minor correction factor is necessary between the two modalities. Conclusion: The Kinect IR system presented can be used clinically for arm volume measurements to detect and monitor patients at risk for or managing lymphedema.