Sebastian Niko Nagel1, Sebastian Wyschkon2, Stefan Schwartz3, Bernd Hamm4, Thomas Elgeti5. 1. Klinik und Hochschulambulanz für Radiologie, Charité - Universitätsmedizin Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, 12203 Berlin, Germany. Electronic address: sebastian.nagel@charite.de. 2. Klinik und Hochschulambulanz für Radiologie, Charité - Universitätsmedizin Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, 12203 Berlin, Germany. Electronic address: sebastian.wyschkon@charite.de. 3. Medizinische Klinik mit Schwerpunkt Hämatologie und Onkologie, Charité - Universitätsmedizin Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, 12203 Berlin, Germany. Electronic address: stefan.schwartz@charite.de. 4. Klinik und Hochschulambulanz für Radiologie, Charité - Universitätsmedizin Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, 12203 Berlin, Germany. Electronic address: bernd.hamm@charite.de. 5. Klinik und Hochschulambulanz für Radiologie, Charité - Universitätsmedizin Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, 12203 Berlin, Germany. Electronic address: thomas.elgeti@charite.de.
Abstract
OBJECTIVE: To prospectively evaluate a short MRI examination protocol for the detection of nodular pulmonary infiltrates in immunocompromised patients with hematologic diseases and suspected invasive fungal infections. METHODS: Patients with nodular infiltrates on CT scans were examined on a 3T MRI scanner. The standardized protocol included axial T2-weighted fast spin echo (FSE) sequences +/- fat saturation (FS), and axial T1-weighted gradient echo (GRE) sequences. Long and short axis diameters of nodular infiltrates and visibility were assessed on MR images at least six months after the CT scan, blinded to patient and examination data. Inter- and intra-reader reliability was assessed in two patients. Statistical testing included Wilcoxon-test, Cohen's kappa, and intra-class correlation coefficients. Bland-Altman plots were created to visualize differences in the measurements. RESULTS: In all 13 patients MRI examinations were completed successfully (average examination time 12 min and maximum breath-hold time of 8s). CT detected 409 nodules. Sensitivity of MRI was 93.2% when using all sequences in combination; considering nodules >5mm, sensitivity increased to 97.9%. Reliability analysis showed excellent correlations with an intra-class correlation coefficient of at least 0.89 for T2 FSE (95% CI 0.79-0.93, p<0.01) images for the intra-, and the lowest of 0.77 for T2 FSE (95% CI 0.55-0.89, p<0.01) images for the inter-reader comparison. Agreement on nodule visibility was at least kappa=0.95 (p<0.01) for the intra- and 0.72 (p<0.01) for the inter-reader analysis. CONCLUSION: With an average examination time of 12 min, pulmonary MRI at 3T is feasible in immunocompromised patients with hematologic diseases and suspected invasive fungal infections. MRI might serve as an alternative diagnostic tool during follow-up examinations.
OBJECTIVE: To prospectively evaluate a short MRI examination protocol for the detection of nodular pulmonary infiltrates in immunocompromised patients with hematologic diseases and suspected invasive fungal infections. METHODS:Patients with nodular infiltrates on CT scans were examined on a 3T MRI scanner. The standardized protocol included axial T2-weighted fast spin echo (FSE) sequences +/- fat saturation (FS), and axial T1-weighted gradient echo (GRE) sequences. Long and short axis diameters of nodular infiltrates and visibility were assessed on MR images at least six months after the CT scan, blinded to patient and examination data. Inter- and intra-reader reliability was assessed in two patients. Statistical testing included Wilcoxon-test, Cohen's kappa, and intra-class correlation coefficients. Bland-Altman plots were created to visualize differences in the measurements. RESULTS: In all 13 patients MRI examinations were completed successfully (average examination time 12 min and maximum breath-hold time of 8s). CT detected 409 nodules. Sensitivity of MRI was 93.2% when using all sequences in combination; considering nodules >5mm, sensitivity increased to 97.9%. Reliability analysis showed excellent correlations with an intra-class correlation coefficient of at least 0.89 for T2 FSE (95% CI 0.79-0.93, p<0.01) images for the intra-, and the lowest of 0.77 for T2 FSE (95% CI 0.55-0.89, p<0.01) images for the inter-reader comparison. Agreement on nodule visibility was at least kappa=0.95 (p<0.01) for the intra- and 0.72 (p<0.01) for the inter-reader analysis. CONCLUSION: With an average examination time of 12 min, pulmonary MRI at 3T is feasible in immunocompromised patients with hematologic diseases and suspected invasive fungal infections. MRI might serve as an alternative diagnostic tool during follow-up examinations.
Authors: Damon Kim; Thomas Elgeti; Tobias Penzkofer; Ingo G Steffen; Laura J Jensen; Stefan Schwartz; Bernd Hamm; Sebastian N Nagel Journal: Eur Radiol Date: 2020-08-21 Impact factor: 5.315