Bed posture classification for pressure ulcer prevention

Abstract

Pressure ulcer is an age-old problem imposing a huge cost to our health care system. Detecting and keeping record of the patient's posture on bed, help care givers reposition patient more efficiently and reduce the risk of developing pressure ulcer. In this paper, a commercial pressure mapping system is used to create a time-stamped, whole-body pressure map of the patient. An image-based processing algorithm is developed to keep an unobtrusive and informative record of patient's bed posture over time. The experimental results show that proposed algorithm can predict patient's bed posture with up to 97.7% average accuracy. This algorithm could ultimately be used with current support surface technologies to reduce the risk of ulcer development.

Automatic Sleep System Recommendation by Multi-modal RBG-Depth-Pressure Anthropometric Analysis

“…As with 3D camera devices, digital human models have also been applied to pressure maps (Harada et al 1999(Harada et al , 2001, which are able to estimate the position of the different body parts. Yousefi et al (2011) used a 2D parametric model to detect areas of the body which are at risk of developing a pressure ulcer. However, most of the works rely on pressure signal features combined with classifiers such as SVM or k-Nearest Neighbors (kNN) for posture recognition (Foubert 2010), or spatial and geometrical features based on a grid division of the pressure map combined with Hidden Markov Models (HMM) for continuous posture evaluation (Liu et al 2013(Liu et al , 2014.…”

Wearable accelerometer based extended sleep position recognition

“…For this approach different methodologies are used including: pressure amplitude value change embedded in a logistic regression model [3], pressure image analysis [12], [17], Bayesian classifier [10], and an algorithm combining normalization, Eigenspace projection, and a kNN classifier [24]. The approaches try to distinguish up to 13 different positions [16] showing difficulties with distinguishing the prone from the supine position.…”

“…In the studies, data from a different number of subjects were collected. In most cases, data was collected in a simulated setting, from two [10], three [3], six [24], nine [16] and 20 [17] participants. Liu et al [12] collected data from 14 subjects in a simulated, and three in a real world situation.…”

“…Postures with a higher granularity need to be considered, as leg positioning could lead to hip rotation resulting in back pain [8]. In general, motivations for sleep position tracking are often found in the prevention of pressure ulcers [17], [24] or sleep apnea [5]. Research has been done to monitor the four basic sleep postures taking into account simulated and real world data using different sleep position trackers.…”

“…Those trackers can be grouped into three categories. The first category uses sensors installed in the bed [3], [10], [16], [17], [24]. The issue with this method is, that it cannot distinguish between the supine and prone positions.…”