Effect of Amplitude Criteria on Operating Characteristic of Detection for OSAH Events with Oxygen Saturation

Abstract

Effect of amplitude criteria on the operating characteristics of algorithms for detecting OSAH events based on the analysis of oxygen saturation alone is investigated. The objective is to establish that there exists an oxygen desaturation level that leverages these algorithms to be more sensitive or more specific, irrespective of the differences in detection mechanism and database, a first ever attempt. Linear classification of algorithms from previous studies discovered that a drop in oxygen saturation of 3% or less makes the detection algorithms more sensitive while a drop of 4% or more makes it more specific. Results from two algorithms developed here also supported this. This finding explains the contradiction cited in the performance of algorithms from the different authors, which casts doubts on their detection ability. It could lead to the establishment of standard oxygen desaturation levels for screening and diagnosis of moderate/severe OSA, thus providing a more credible comparison basis for automated detection algorithms or even clinical tests.

Assessment of computerized analysis algorithms for one dimension physiologic data with multiple databases

“…Several issues contributing to this are suggested and publications pertaining to this are invited. It is a surprise that our research into the computerized analysis of oxygen saturation to detect for desaturation events [2][3] reveals supportive evidence to issues raised in [1].…”

“…Secondly, annotators tend to underscore, dictated by their general practice, so as not to over-prescribe medication while the automated algorithms are designed to over-detect SSW to increase Se. Hence, discrepancy exists in the perception of 'TP' by clinician and engineers, which explains why computerassisted methods always lead to low sensitivity [3]. Moreover, the intra-scorer and inter-scorer discrepancy in golden standard are statistically proven [18].…”

“…1 denotes the start and end of a SSW event? This explains why most authors derive their own thresholds empirically from their own datasets (see Table 1, [3]). Objective determination of threshold by maximizing sum of sensitivity (Se) and specificity (Sp) [13] and by varying the number of annotation segments for highest Se [18] are observed.…”

“…These methods devise ways to replace the qualitative expressions with mathematic equations. Analogous to detection for desaturation events [2][3], it is found meaningless to ascertain the relative merits of these studies due to discrepancies in database, features description and criteria, performance evaluation.…”

“…Review of detection algorithms here is focused on spike events only since an extensive account of this for desaturation events is presented in [2][3].…”