Related Experiment Video
Updated: Jan 12, 2026
Drug-Induced Sleep Endoscopy DISE with Target Controlled Infusion TCI and Bispectral Analysis in Obstructive Sleep Apnea
Published on: December 6, 2016
Opportunistic pharyngeal length measurement on routine brain MRI: A novel screening tool for severe obstructive sleep
Tho Dinh Phung1,2,3, Giap Van Vu4,5,6, Chung Thi Phan7
1Department of Cardiology, Hanoi Medical University, Hanoi, Vietnam.
Background:
Obstructive sleep apnea (OSA) affects two-thirds of acute ischemic stroke (AIS) patients and is linked to poor neurological recovery and recurrent stroke. Current OSA screening tools (e.g., STOP-BANG) are limited by subjectivity and impracticality in acute stroke settings. We evaluated the predictive value of upper airway dimensions measured on routine brain MRI (Magnetic Resonance Imaging) - a novel, objective strategy - for screening severe OSA in AIS patients.
Methods:
In this prospective single-center study (April-October 2024), 103 adults with AIS underwent routine brain MRI. Retropalatal and combined retropalatal-retroglossal pharyngeal length (CRR) were measured on sagittal T1-weighted sequences. OSA severity was confirmed via cardiorespiratory polygraphy. The diagnostic accuracy of these MRI-derived measurements for screening severe OSA was compared to the STOP-BANG questionnaire.
Results:
Participants (mean age 59.6 ± 11.2 years; 71.8% male) had mild-moderate stroke severity (median NIHSS 3 [IQR 2-4]) and mean BMI 24.1 ± 3.1 kg/m². CRR was significantly longer in severe OSA versus non-severe OSA (66.6 ± 8.9 mm vs. 62.5 ± 8.1 mm, p = 0.048). CRR demonstrated superior discriminative accuracy for severe OSA (AUC = 0.70, 95% CI: 0.61-0.78) compared to STOP-BANG ≥ 3 (AUC = 0.57, 95% CI: 0.48-0.65). Using cutoff ≥62 mm, CRR achieved 91% sensitivity and 49% specificity, outperforming STOP-BANG ≥ 3 (sensitivity = 90%, specificity = 24%).
Conclusions:
Opportunistic measurement of CRR on routine brain MRI is a rapid, non-invasive screening tool for severe OSA in AIS, leveraging existing neuroimaging data without added costs or patient burden. It's high sensitivity minimizes missed diagnoses, addressing a critical gap in acute stroke care.
