Background

Quantification of lesion area is widely used in plant pathology to assess disease severity, compare pathogen virulence, and evaluate host resistance in plant–pathogen interaction studies [1]. Necrotic lesions on leaves are common readouts in experimental systems involving viral, bacterial, and fungal pathogens, making accurate and reproducible lesion measurement important for comparative analyses across genotypes, treatments, and time points [2].

Several approaches have been used to quantify lesion areas, including visual scoring, manual tracing, and, more recently, digital image analysis, which has become the conventional method [3]. Among digital tools, ImageJ has become widely adopted because it is easy to use, freely available, and frequently used for color-based segmentation approaches such as global thresholding [4]. However, color-based threshold methods can perform poorly when lesions are heterogeneous in color, have diffuse or irregular borders, or are captured under non-uniform lighting conditions. Under such conditions, lesions may be incompletely detected, adjacent lesions may be merged, or healthy tissues with similar color may be misclassified as diseased areas. While advanced machine learning and automated segmentation tools offer powerful alternatives to thresholding [5], they frequently require specialized computational expertise, extensive training datasets, or high-end hardware. Furthermore, studies reporting ImageJ-based lesion quantification often describe the software without specifying workflows or parameters, hampering reproduction by readers.

The protocol described here addresses these gaps by providing a highly controlled and standardized manual ImageJ workflow for quantifying irregular necrotic lesions on plant leaves. It requires no specialized equipment beyond standard image acquisition and freely available software and is well-suited for experiments involving individual lesions or a moderate number of lesions per sample. Compared with automated color thresholding–based methods, this approach offers controlled precision in lesion boundary selection, particularly when lesion shape, color, or contrast complicates segmentation. Although the workflow becomes labor-intensive when lesions are extremely numerous, it remains well-suited to datasets with several tens of lesions per leaf. This manual ImageJ-based approach can be adapted to quantify other visible, localized plant symptoms on the leaves of different plant species. For example, the workflow can measure symptoms caused by diverse pathogens that induce localized tissue damage, necrosis, or pathogen-associated mycelial growth. It is also applicable to distinct discolorations on leaves of varying colors (such as chlorotic spots), as well as physical deformations like hypertrophic regions or insect herbivory (chewing damage). However, the suitability of this approach depends on how clearly symptom boundaries can be defined.