A two-stage approach to positioning and identification of tracheal intubation using LED-based lightwand and acoustic models.

This paper proposes a two-stage approach to positioning and identifying tracheal intubations by computer-assisted diagnosis system. First, an innovated LED-based lightwand is developed for positioning the opening of the trachea and inserting the endotracheal tube rapidly. The proposed LED-based lightwand instrument significantly reduces the effects of temperature reaction without changing transillumination through the tissue. After intubation, breath sound analysis instrumentation is developed and assists for identifying the accurate tube position. To overcome the fast and overstepped variations of amplitude, a high sensitivity omni dimensional microphone and an automatic gain control device with linear phase property are calibrated and conducted into experiments of operating intubations. User-friendly interface software is also developed to analyze and transcribe the physiological characteristics of the collected breath sound corpus. Several subjective and objective experiments are performed to examine the practicability of the proposed approach and systems. The preliminary results show that the proposed LED-based lightwand instrument outperformed in the tip temperature without the influence on transillumination. The proposed parameters of average vibration parameter variance (AVPV) and frequency-domain energy variance (FEV) revealed the potential for distinguishing between the tracheal and esophageal intubations.