Article

Diabetic peripheral neuropathy (DPN) is a common complication leading to foot
ulceration and amputation. Several kinematic, kinetic and plantar pressure
measures have been proposed for DPN detection, however findings have been
inconsistent. In this work, we present new shape features that capture variations
in the plantar pressure using shape and entropy measures to the study of patients
with retinopathy, DPN and nephropathy, and a control diabetic group with no
complications. The change in the peak plantar pressure (PPP) position with each
step for both feet was represented as a convex polygon, asymmetry index, area of
the convex polygon, 2nd wavelet moment (WM2) and sample entropy (SamEn). WM2 and
the SamEn were more sensitive in capturing variations due to presence of
complications than the area and asymmetry measures. WM2 of the left heel (median:
1st IQ, 3rd IQ): 8.27 (4.6,14.8) and left forefoot: 9.2 (2.4,16) were
significantly lower for the DPN group compared to the control (CONT) group (heel
11.9 (5.0,16.4); forefoot: 10.3 (4.4,21.3), p < 0.05). SamEn for the DPN group
was significantly lower in the right foot compared to the left foot (1.3 (1.26,
1.37) and 1.33 (1.26,1.4), p < 0.01) compared to CONT (right foot: 1.37
(1.24,1.45) and left foot: 1.34 (1.25,1.42), P < 0.05).
These new shape and
regularity features have shown promising results in detecting diabetic peripheral
neuropathy and warrant further investigation.
CI - Copyright A(c) 2016 Elsevier B.V. All rights reserved.

Langue : ANGLAIS