Premium
Brain Abscess Apparent Diffusion Coefficient is Associated With Microbial Culture Yields
Author(s) -
Toh Cheng Hong,
Siow Tiing Yee,
Wong Alex MunChing,
Castillo Mauricio
Publication year - 2021
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.27563
Subject(s) - medicine , abscess , effective diffusion coefficient , brain abscess , receiver operating characteristic , logistic regression , statistical significance , clinical significance , intraclass correlation , diffusion mri , nuclear medicine , radiology , surgery , magnetic resonance imaging , clinical psychology , psychometrics
Background Obtaining pus for microbial cultures is one of the surgical aims in patients with brain abscess. Predictors of microbial yields are necessary as they help in treatment planning. Purpose To investigate the relationship between microbial culture yields of brain abscesses and their apparent diffusion coefficient (ADC) values and clinical characteristics. Study Type Retrospective. Subjects Eighty ‐ four patients diagnosed with brain abscess by surgery and histopathology (59 with positive abscess cultures). Field Strength/Sequence Diffusion‐weighted, T2‐weigthed, and contrast‐enhanced T1‐weighted imaging at 1.5 T and 3 T. Assessment Contrast‐enhanced T1‐weighted images were co‐registered to ADC maps. Three neuroradiologists determined abscess imaging characteristics (distribution, location, and ventricular rupture), and two measured abscess volumes and ADC values. Clinical characteristics collected included sex, age, fever, underlying diseases, infection sources, white blood cell count, percentage of segmented neutrophils, C‐reactive protein level, regimen and duration of empirical antibiotics, and types of surgery. Statistical Tests Interobserver differences were assessed with Fleiss kappa and intraclass correlation coefficients. The differences in clinical and imaging factors between the positive and negative culture groups were compared with Chi‐square analysis or Student's t test. All factors were subjected to multivariable logistic regression analysis to assess their associations with microbial culture yields, and factors with statistical significance were evaluated with receiver operating characteristic curve analysis to assess their diagnostic performance in discriminating the two groups. Results Mean ADC (×10 −6 mm 2 /s) of culture‐negative abscesses (841 ± 173) was significantly higher ( P < 0.05) than that of culture‐positive abscesses (536 ± 90). On multivariable analysis, mean ADC was the only significant factor ( P < 0.05) related to culture yields. With 660 as the cutoff value, the sensitivity, specificity, and accuracy of ADC for discriminating culture yields were 93.2%, 88.0%, and 91.7%, respectively. Data Conclusion ADC could be used to discriminate between culture‐positive and culture‐negative abscesses. Evidence Level 4 Technical Efficacy Stage 2