Location: Subtropical Plant Pathology ResearchTitle: Visual rating and the use of image analysis for assessing different symptoms of citrus canker on grapefruit leaves) Author
|Cook, A. Z.|
Submitted to: Plant Disease
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/1/2008
Publication Date: 6/1/2008
Citation: Bock, C.H., Parker, P., Cook, A., Gottwald, T.R. 2008. Visual rating and the use of image analysis for assessing different symptoms of citrus canker on grapefruit leaves. Plant Disease. 92:530-541. Interpretive Summary:
Technical Abstract: Citrus canker is caused by the bacterial pathogen Xanthomonas axonopodis pv citri (Xac) and infects several citrus species in wet tropical and subtropical citrus growing regions. Accurate, precise and reproducible disease assessment is needed for monitoring epidemics and disease response in breeding material. The objective of this study was to assess reproducibility of image analysis (IA) for measuring severity of canker symptoms and to compare this to visual assessments made by three visual assessors (VA1-3) for various symptom types (lesion numbers, % area necrotic and % area necrotic+chlorotic), and to assess inter-, and intra-VA reproducibility. Digital images of 210 citrus leaves with a range of symptom severity were assessed on two separate occasions. IA was the most precise compared to VAs for all symptom types (inter-assessment correlation coefficients, r, for lesion counts by IA = 0.99, by VAs = 0.89-0.94; for %, r for % area necrotic+chloroitic for IA=0.97 and for VAs=0.86-0.89; and r for % area necrotic for IA=0.96 and for VAs=0.74-0.85). Accuracy based on Lin’s concordance coefficient also followed a similar pattern, with IA being most consistently accurate for all symptom types (bias correction factor, Cb = 0.99-1.00) compared to visual assessors (Cb = 0.85-1.00). Lesion number was the most reproducible symptom assessment (Lin’s concordance correlation coefficient, 'c, = 0.76-0.99), followed by % area necrotic+chlorotic ('c = 0.85-0.97), and finally % area necrotic ('c = 0.72-0.96). Based on the “true” value provided by IA, precision among VAs was reasonable for number of lesions per leaf (r = 0.88-0.94), slightly less precision for % area chlorotic+necrotic (r = 0.87-0.92), and poorest precision for % area necrotic (r = 0.77-0.83). Loss in accuracy was less, but showed a similar trend with counts of lesion numbers (Cb = 0.93-0.99) which was more consistently accurately reproduced by VAs than either % area necrotic (Cb = 0.85-0.99) or % area necrotic+chlorotic (Cb = 0.91-1.00). Thus visual assessors suffered both losses in precision and accuracy, with loss in precision estimating % area necrotic being the greatest. Indeed, only for % area necrotic was there a significant effect of assessor (a two-way random effects model ANOVA returned a P<0.001 and 0.016 for assessor in assessment 1 and 2, respectively). VAs showed a marked preference for clustering of % area severity estimates, especially at severity >20% area (eg 25, 30, 35, 40 etc), yet VAs were prepared to estimate disease of <1% area, and at 1% increments up to 20%. Their was a linear relationship between actual disease (IA assessments) and VAs. IA appears to provide a highly reproducible way to assess canker infected leaves for disease, but symptom characters (symptom heterogeneity, coalescence of lesions) could lead to discrepancies in results, and full automation of the system remains to be tested.