Return to search

Comparing automatically and manually scored apnea hypopnea index and investigating if differences are affected by central apneas and home sleep apnea test signal quality

Introduction: Sleep apnea is a pathological health condition with repeatedly paused breathing during sleep. The condition can cause serious health problems and decrease quality of life. Offering a fast diagnosis and treatment could prevent further progress of the condition. The severity of sleep apnea is indicated by an apnea hypopnea index (AHI), which is scored based on a home sleep apnea test (HSAT). The purpose: This study compared the differences between manually and automatically scored AHI, to examine if the automatic scoring is an acceptable singular method for sleep apnea diagnostics. This study also examined if AHI differences could be predicted by HSAT airflow signal quality and the degree of central or mixed apneas. Methods: Sleep apnea patients were instructed by the author how to use the HSAT equipment, data of 182 one-night HSAT recordings were then collected. Each recording was analyzed automatically and manually by a sleep specialist, using the software Noxturnal 6.3. Results: There was a great correlation between the two AHI scoring methods (Spearman’s r 0,97), but a statistically significant difference was found. The positive predictive value (PPV) and negative predictive value (NPV) of the automatic method were 96% and 97%, respectively, sensitivity was 99% and specificity 84%. A moderate, negative correlation between signal quality and AHI differences (Pearson’s r -0,31) was found, but none with central apneas. Conclusion: The results were contradictory, but considering a low Cohen’s d, this study still concludes that clinical use of automatic AHI scoring should be sufficient if AHI > 15.

Identiferoai:union.ndltd.org:UPSALLA1/oai:DiVA.org:uu-530793
Date January 2024
CreatorsStrandberg, Johanna
PublisherUppsala universitet, Institutionen för medicinsk cellbiologi
Source SetsDiVA Archive at Upsalla University
LanguageEnglish
Detected LanguageEnglish
TypeStudent thesis, info:eu-repo/semantics/bachelorThesis, text
Formatapplication/pdf
Rightsinfo:eu-repo/semantics/openAccess

Page generated in 0.0016 seconds