CPAP Titration Using an Artificial Neural Network: A Randomized Controlled Study

Study was terminated due to lack of interest from subjects and no funding, only 1 subject signed consent but did not participate.

The purpose of the study is to determine the validity of the prediction model in reducing the rate of CPAP titration failure and in achieving a shorter time to optimal pressure

In order to derive the most effective pressure, CPAP titration is performed in the sleep laboratory during which the pressure is gradually increased until apneas and hypopneas are abolished in all sleep stages and in all body positions. The technique is however time consuming and labor intensive. Furthermore, the duration of the study may not be sufficient to attain this goal because of patient's poor ability to sleep in this environment or due to difficulty in attaining an appropriate pressure. A predictive algorithm based on demographic, anthropometric, and polysomnographic data was developed to facilitate the selection of a starting pressure during the overnight titration study. Yet, the performance of this model was inconsistent when validated by other centers. One of the potential reasons for the lack of reproducibility is the complex relation of behavioral processes with nonlinear attributes. In areas of complex interactions, the artificial neural network (ANN) has been found to be a more appropriate alternative to linear, parametric statistical tools due to its inherent property of seeking information embedded in relations among variables thought to be independent. Comparison: time to achieve optimal pressure in the conventional technique versus the intervention model

Inclusion Criteria: patients 18 years of age and older, documented OSA by sleep study defined as AHI > 5/hr Exclusion Criteria: previously treated OSA, unwilling to undergo a titration study, unable or unwilling to sign an informed consent.

El Solh AA, Aldik Z, Alnabhan M, Grant B. Predicting effective continuous positive airway pressure in sleep apnea using an artificial neural network. Sleep Med. 2007 Aug;8(5):471-7. doi: 10.1016/j.sleep.2006.09.005. Epub 2007 May 18.