Posture and Limb Detection for Pressure Ulcer Prevention R. Yousefi, S. Ostadabbas, M. Faezipour, M. Nourani, L. Tamil Quality of Life Technology Laboratory The University of Texas at Dallas, Richardson, TX 75083 {rxy091020, sarahostad, miad.fz, nourani, tamil}@utdallas.edu M. Pompeo, M.D. Presbyterian Wound Care Clinic Dallas, TX, 75231 healerone@aol.com ABSTRACT Pressure ulcers have affected humans for ages and address- ing the prevention of pressure ulcers is a prominent issue in our healthcare system. Once developed, the treatment is costly and it increases length of hospital stay. This is particularly true for patients with impaired sensation including diabetics, advanced age or prolonged immobility. In this work, we have developed a processing platform that unobtrusively records patient’s bed posture and tracks different limbs along with associated statistical pressure data. The proposed algorithm has a training and test step. K Nearest Neighbor (kNN) algorithm is used to classify different postures. A cardboard body model is assigned to training samples for body limb detection. Keywords: Pressure Ulcer, Posture Classification, Limb De- tection, Pressure Mapping System. I. I NTRODUCTION Pressure ulcer (PU) is a skin breakdown which develops over a bony prominence as a result of pressure as main factors and other factors like shear stress and friction [1]. In hospitalized patients, the prevalence ranges from about 3% to 11% (approximately 1.5-3.0 million patients in the United States). Groups known to have a high risk of developing pressure ulcer include bedridden patients , wheelchair-bound individuals, frail elderly [2] with no or limited mobility, as well as individuals with diabetes, poor nutrition, and chronic blood-flow diseases [3]. Pressure ulcers imposes an enormous burden on our health care system [4]. Once developed, PUs represent an acute health condition that results in increased costs and suffering over many months and even years. Pressure ulcers result in both an increased length of hospital stay and increased hospital costs [5]. The current cost to our health care system resulting from PUs is more than $1.2 billion annually [6]. Effective ulcer prevention and early detection will greatly reduce patient suffering/discomfort. Strong motivation for this work comes from the high cost of PU treatment and the growing shortage of trained health care providers. In 2000, the shortage of nurses was estimated at 6%. This shortage is expected to grow to 20% by 2015 and, if not addressed, to 29% by 2020. We have developed a monitoring platform using commercial pressure mapping system that records patient’s bed posture and tracks different limbs along with associated statistical pressure data. Turning the patient every two hours, as traditionally advised to hospital staff, is neither efficient nor practical. Our methodology allows care givers to schedule and reposition patient more effectively. It also allows continuous risk assess- ment and provides related information for a managed healing process. The remaining of this paper is as follows: A review of previous work is done in Section II. Our data collection platform is described in Section III. The proposed algorithm and associated experimental results for posture classification and body limb detection are discussed in Section Section IV. Section V contains concluding remarks. II. PREVIOUS WORK To deal with pressure as the main contributing factor in PUs, different support surfaces have been introduced to distribute load over the contacted areas of the human body. Features added to support surface to do pressure redistribution includes air fluidized, alternating pressure, low air loss and multi- zoned surfaces. There has been always a need to evaluate the effectiveness and optimize usage of support surface technology and other resources to prevent PUs. Pressure sensors are good means to monitor pressure on different areas of the body on support surfaces. Adding processing capabilities to pressure sensors could help us extract more useful information for pressure ulcer management including patient’s posture on the bed. Pressure sensors have been used for bed posture detec- tion in many research works. A low-cost sensor mat with minimum number of sensors with a feed-forward neural net was presented in [7] for detection and distinction of body position. Harada et al. have proposed a template-based human posture detection [8] and a pressure image-based human motion tracking system [9]. In [10], a bed robotic system was proposed for elderly and the disabled, and pressure sensors embedded in the mattress are used to estimate the pose of the patient. A bed posture detection method is developed in [11] using Bayesian classification for the elderly where statistical kurtosis and skewness measures are estimated as feature vector to represent the shape of pressure contour using the pressure values received from sensors. To achieve a better performance, a multimodal approach to human sleeping posture classification, using pressure sensor array and video camera as complementary modalities was proposed in [12]. A comparison of different sleeping posture classification using cost-effective pressure sensitive mattress was studied in [13]. III. DATA COLLECTION PLATFORM Force Sensing Array (FSA) [14] is used to collect pressure data on the bed. The FSA system is a flexible mat that contains 2048 (32×64) uniformly distributed sensors which cover the 1