Estimation of Lung Properties Using ANN-Based Inverse Modeling of Spirometric Data

Spirometry is the most commonly used test of lung function because the forced expiratory flow-volume (FV) curve is effort-independent and simultaneously sensitive to pathological processes in the lungs. Despite this, a method for the estimation of respiratory system parameters, based on this association, has not been yet proposed. The aim of this work was to explore a feedforward neural network (FFNN) approximating the inverse mapping between the FV curve and respiratory parameters. To this end, the sensitivity analysis of the reduced model for forced expiration has been carried out, showing its local identifiability and the importance of particular parameters. This forward model was then applied to simulate spirometric data (8000 elements), used for training, validating, optimizing and testing the FFNN. The suboptimal FFNN structure had 52 input neurons (for spirometric data), two hidden nonlinear layers with 30 and 20 neurons respectively, and 10 output neurons (for parameter estimates). The total relative error of estimation of individual parameters was between 11 and 28%. Parameter estimates yielded by this inverse FFNN will be used as starting points for a more precise local estimation algorithm.

[1]  A. G. Polak,et al.  A forward model for maximum expiration , 1998, Comput. Biol. Medicine.

[2]  J E Cotes,et al.  Lung volumes and forced ventilatory flows , 1993, European Respiratory Journal.

[3]  Kerry L Hedges,et al.  Simulation of Forced Expiration in a Biophysical Model, With Homogeneous and Clustered Bronchoconstriction. , 2016, Journal of biomechanical engineering.

[4]  L Battolla,et al.  Relationship between extent of pulmonary emphysema by high-resolution computed tomography and lung elastic recoil in patients with chronic obstructive pulmonary disease. , 2001, American journal of respiratory and critical care medicine.

[5]  Kurt Hornik,et al.  Approximation capabilities of multilayer feedforward networks , 1991, Neural Networks.

[6]  R. Castile,et al.  Model of forced expiratory flows and airway geometry in infants. , 2004, Journal of applied physiology.

[7]  A. Tikhonov,et al.  Numerical Methods for the Solution of Ill-Posed Problems , 1995 .

[9]  Claudio Cobelli,et al.  Generalized Sensitivity Functions in Physiological System Identification , 1999, Annals of Biomedical Engineering.

[11]  Lalita Udpa,et al.  Neural network-based inversion algorithms in magnetic flux leakage nondestructive evaluation , 2003 .

[12]  R. Lambert Sensitivity and specificity of the computational model for maximal expiratory flow. , 1984, Journal of applied physiology: respiratory, environmental and exercise physiology.

[13]  Qi-Jun Zhang,et al.  Neural Network Inverse Modeling and Applications to Microwave Filter Design , 2008, IEEE Transactions on Microwave Theory and Techniques.

[14]  A. G. Polak,et al.  Computational Model for Forced Expiration from Asymmetric Normal Lungs , 2003, Annals of Biomedical Engineering.

[15]  J. Hankinson,et al.  Interpretative strategies for lung function tests , 2005, European Respiratory Journal.

[16]  E. Weibel Morphometry of the Human Lung , 1965, Springer Berlin Heidelberg.

[17]  T A Wilson,et al.  A computational model for expiratory flow. , 1982, Journal of applied physiology: respiratory, environmental and exercise physiology.

[18]  U. Ruttimann,et al.  The relationship between left mainstem bronchial diameter and patient size. , 1995, Journal of cardiothoracic and vascular anesthesia.

[19]  R. Brown,et al.  Effect of lung inflation and airway muscle tone on airway diameter in vivo. , 1996, Journal of applied physiology.

[20]  K. Beck,et al.  Airway area distribution from the forced expiration maneuver. , 2004, Journal of applied physiology.

[21]  M. Canals,et al.  A Simple Geometrical Pattern for the Branching Distribution of the Bronchial Tree, Useful to Estimate Optimality Departures , 2004, Acta biotheoretica.

[22]  Hyungsun Lim,et al.  Measurements of the length and diameter of main bronchi on three-dimensional images in Asian adult patients in comparison with the height of patients. , 2014, Journal of cardiothoracic and vascular anesthesia.

[23]  Adam G. Polak,et al.  Indirect measurements: combining parameter selection with ridge regression , 2001 .

[24]  Paul M. J. Van den Hof,et al.  Test for local structural identifiability of high-order non-linearly parametrized state space models , 1996, Autom..

[25]  Vera Kurková,et al.  Kolmogorov's theorem and multilayer neural networks , 1992, Neural Networks.

[26]  Adam G. Polak,et al.  Simulation of lung function evolution after heart-lung transplantation using a numerical model , 2005, IEEE Transactions on Biomedical Engineering.

[27]  Arnab Majumdar,et al.  Relating airway diameter distributions to regular branching asymmetry in the lung. , 2005, Physical review letters.

[28]  P. Sterk,et al.  Pressure-volume analysis of the lung with an exponential and linear-exponential model in asthma and COPD. Dutch CNSLD Study Group. , 1995, The European respiratory journal.