Search

• An alternative approach using the artificial neural network (ANN) in better identifying the insulinaemic pharmacokinetic parameters by mitigating the outlying data that could affect the overall outcome of the research (in terms of insulin sensitivity value and endogenous insulin secretion profile). • The low value of residual error plots provides additional support for this study's finding. Thus, it shows a considerably smaller, yet consistent trend regarding the measured data as a result. This change implies the adapted approach captures previously un-modelled effects in the interstitial insulin kinetic model. • The adapted approach yields no wrong conclusions when applied to typical data, making it more suitable for modelling tasks including outlier data or unmodeled effects. • This approach is likely to be useful for analysis of the pathogenesis of T2D as it captures the physiological determinants of participant-specific U N profiles. The identification of insulinaemic pharmacokinetic parameters using the least-squares criterion approach is easily influenced by outlying data due to its sensitivity. Furthermore, the least-squares criterion has a tendency to overfit and produce incorrect results. Hence, this research proposes an alternative approach using the artificial neural network (ANN) with two hidden layers to optimize the identifying of insulinaemic pharmacokinetic parameters. The ANN is selected for its ability to avoid overfitting parameters and its faster speed in processing data. 18 voluntarily participants were recruited from the Canterbury and Otago region of New Zealand to take part in a Dynamic Insulin Sensitivity and Secretion Test (DISST) clinical trial. A total of 46 DISST data were collected. However, due to ambiguous and inconsistency, 4 data had to be removed. Analysis was done using MATLAB 2020a. Results show that, with 42 gathered dataset, the ANN generates higher gains, ∅ P = 20.73 [12.21, 28.57] mU·L·mmol−1·min−1 and ∅ D = 60.42 [26.85, 131.38] mU·L·mmol−1 as compared to the linear least square method, ∅ P = 19.67 [11.81, 28.02] mU·L·mmol−1 ·min−1 and ∅ D = 46.21 [7.25, 116.71] mU·L·mmol−1. The average value of the insulin sensitivity (SI) of ANN is lower with, SI = 16 × 10−4 L·mU−1 ·min−1 than the linear least square, SI = 17 × 10−4 L·mU−1 ·min−1. Although the ANN analysis provided a lower SI value, the results were more dependable than the linear least square model because the ANN approach yielded a better model fitting accuracy than the linear least square method with a lower residual error of less than 5%. With the implementation of this ANN architecture, it shows that ANN able to produce minimal error during optimization process particularly when dealing with outlying data. The findings may provide extra information to clinicians, allowing them to gain a better knowledge of the heterogenous aetiology of diabetes and therapeutic intervention options. [ABSTRACT FROM AUTHOR]