Region Specific Weight Measuring System for Bedridden patients

Abstract

Weight measurement is one important aspect, especially for monitoring long term bedridden patients or for post-operative subjects. Presently, though overall weight is measured regularly, periodicity is in the order days owing to practical difficulties like portability of equipment and shifting of patients. Even in this case, region-wise weight measurement is not done. This problem is addressed in the proposed method, where localization of region that contributes to weight change can be identified along with amount of weight change. Since no movement of subject is required, observation interval can be reduced to hours from days. This is achieved by dividing the mild steel sheet holding the subject and placing load sensors beneath the sheets. This is done after understanding and critically reviewing load and height distribution in human body. Load sensors, along with artificial neural network is able to sense the change up to 1 gram with information about location. Entire experiment was done using human phantom model. Result of the proposed method exhibits accuracy of 96% and above with linearity in measurement. Our present efforts are towards making this prototype into a real time weight measuring system.