A clinical decision support system using multilayer perceptron neural network to assess well being in diabetes.

摘要: Background : Diabetes mellitus is an increasingly common life-style disorder whose management outcomes are measured in symptomatic, biochemical as well psychological areas. Well being outcome of treatment recognized a crucial component treatment. There little published literature on psychosocial and the factors influencing them. Therefore we have developed neural network system which trained to predict diabetes, using data generated real life. Material Methods We Multi Layer Perceptron Neural Network model, had been by back propagation algorithm. Data was used from cohort 241 individuals with diabetes. age, gender, weight, fasting plasma glucose set inputs predicted measures - (depression, anxiety, energy positive well-being). Results It observed that female patients report significantly higher levels depression than their male counter parts. Some slight high or no significant differences between males regard number persons whom they share anxieties fears regarding not much difference has both females. Also, Males pwb value when compared counterparts. this may be due women tend react more emotionally disease hence experience difficulty coping it. The present sample predominantly house wives worrying about health its problems. it that, gender total general well-being. With five (age, sex, glucose, bias), four outputs well-being) momentum rate 0.9, learning 0.7, 50. maximum individual error 0.001 iterations were 500, hidden layers 1 units layer 6, normalized 470.57. input samples 100, 150 200, keeping other variables constant, 419.61, 359.67 332.32 respectively. Similar values found for increased 7, 8 9 two layers, each containing 6,7,8,9,10,11 50,100,150, same found.. Women having weight 40kgs 85kgs men who 39kgs 102kgs. Conclusion prototype model well-being biological biographical given inputs. When greater fed system, can reduced. ©