Güliz Acker1, Nicolas Schlinkmann2, Sophie Käthe Piper3, Julia Onken2, Peter Vajkoczy2, Thomas Picht4. 1. Department of Neurosurgery and Center for Stroke Research Berlin, Charité-Universitätsmedizin Berlin, Berlin, Germany; Berlin Institute of Health, Berlin, Germany. 2. Department of Neurosurgery and Center for Stroke Research Berlin, Charité-Universitätsmedizin Berlin, Berlin, Germany. 3. Institute of Biometry and Clinical Epidemiology, Charité-Universitätsmedizin Berlin, Berlin, Germany; Berlin Institute of Health, Berlin, Germany. 4. Department of Neurosurgery and Center for Stroke Research Berlin, Charité-Universitätsmedizin Berlin, Berlin, Germany; Exzellenzcluster Bild Wissen Gestaltung, Humboldt Universität zu Berlin, Berlin, Germany. Electronic address: thomas.picht@charite.de.
Abstract
OBJECTIVE: Stereoscopic viewing of computed tomographic angiography (CT-A) or magnetic resonance angiograms might increase the diagnostic potential of these imaging techniques. Our aim was to evaluate the benefits of a novel stereoscopic viewing system for aneurysm detection compared with standard monoscopic viewing. METHODS: Retrospective patient data were used for 2 different evaluations. First, monoscopic and stereoscopic CT-A viewing was compared by 14 clinicians in 10 patients with challenging (i.e., small and initially CT-A negative) aneurysms. Second, stereoscopic CT-As and the reference standard, digital subtraction angiography (DSA), were compared in 15 patients with randomly selected aneurysms by 12 clinicians. The study participants rated the presence and location of any aneurysm and its morphological characteristics. The detection rates and interrater reliability were calculated. RESULTS: The first evaluation showed superior aneurysm detection in challenging cases using stereoscopic versus monoscopic CT-A viewing (median: monoscopic, 20%; interquartile range [IQR], 10%-32.5%; stereoscopic, 40%; IQR, 27.5%-42.5%). The interrater reliability analysis revealed good to excellent agreement among raters for aneurysm detection in both viewing modalities (monoscopic, intraclass correlation coefficient [ICC(2,1)], 0.798; 95% confidence interval [CI], 0.549-0.941; stereoscopic viewing, ICC(2,1), 0.895; 95% CI, 0.770-0.968). The second part demonstrated that stereoscopic CT-A viewing is comparable to DSA viewing for aneurysm detection (median: DSA, 80%; IQR, 73%-100%; stereoscopic CT-A, 87%; IQR, 87%-93%). The interrater reliability analysis revealed excellent absolute agreement in aneurysm detection between DSA and stereoscopic CT-A viewing (DSA: ICC(2,1), 0.971; 95% CI, 0.944-0.989; stereoscopic CT-A: ICC(2,1), 0.972; 95% CI, 0.945-0.989). The aneurysm detection rates correlated significantly with the participants' years of experience. CONCLUSIONS: Stereoscopic viewing of CT-As increases the diagnostic accuracy and represents a promising technique to reduce the need for invasive DSA.
OBJECTIVE: Stereoscopic viewing of computed tomographic angiography (CT-A) or magnetic resonance angiograms might increase the diagnostic potential of these imaging techniques. Our aim was to evaluate the benefits of a novel stereoscopic viewing system for aneurysm detection compared with standard monoscopic viewing. METHODS: Retrospective patient data were used for 2 different evaluations. First, monoscopic and stereoscopic CT-A viewing was compared by 14 clinicians in 10 patients with challenging (i.e., small and initially CT-A negative) aneurysms. Second, stereoscopic CT-As and the reference standard, digital subtraction angiography (DSA), were compared in 15 patients with randomly selected aneurysms by 12 clinicians. The study participants rated the presence and location of any aneurysm and its morphological characteristics. The detection rates and interrater reliability were calculated. RESULTS: The first evaluation showed superior aneurysm detection in challenging cases using stereoscopic versus monoscopic CT-A viewing (median: monoscopic, 20%; interquartile range [IQR], 10%-32.5%; stereoscopic, 40%; IQR, 27.5%-42.5%). The interrater reliability analysis revealed good to excellent agreement among raters for aneurysm detection in both viewing modalities (monoscopic, intraclass correlation coefficient [ICC(2,1)], 0.798; 95% confidence interval [CI], 0.549-0.941; stereoscopic viewing, ICC(2,1), 0.895; 95% CI, 0.770-0.968). The second part demonstrated that stereoscopic CT-A viewing is comparable to DSA viewing for aneurysm detection (median: DSA, 80%; IQR, 73%-100%; stereoscopic CT-A, 87%; IQR, 87%-93%). The interrater reliability analysis revealed excellent absolute agreement in aneurysm detection between DSA and stereoscopic CT-A viewing (DSA: ICC(2,1), 0.971; 95% CI, 0.944-0.989; stereoscopic CT-A: ICC(2,1), 0.972; 95% CI, 0.945-0.989). The aneurysm detection rates correlated significantly with the participants' years of experience. CONCLUSIONS: Stereoscopic viewing of CT-As increases the diagnostic accuracy and represents a promising technique to reduce the need for invasive DSA.
Authors: Nicolas Schlinkmann; Rutvik Khakhar; Thomas Picht; Sophie K Piper; Lucius S Fekonja; Peter Vajkoczy; Gueliz Acker Journal: Neurosurg Rev Date: 2021-09-22 Impact factor: 3.042