1 Department of Biochemistry and Molecular and Cellular Biology of Plants, Estación Experimental del Zaidín, Spanish National Research Council (CSIC), Profesor Albareda, 1, 18008, Granada, Spain.
2 Department of Microbiology, Science Faculty of Málaga University, Institute for Mediterranean and Subtropical Horticulture "La Mayora", Málaga University, Spanish National Research Council (IHSM-UMA-CSIC), Boulevard Louis Pasteur, 31, 29071, Málaga, Spain; and.
Plant Dis. 2019 Jun;103(6):1119-1125. doi: 10.1094/PDIS-10-18-1778-RE. Epub 2019 Apr 17.
White root rot, caused by the soilborne fungus , is an important constraint to production for a wide range of woody crop plants such as avocado trees. The current methods of detection of white root rot are based on microbial and molecular techniques, and their application at orchard scale is limited. In this study, physiological parameters provided by imaging techniques were analyzed by machine learning methods. Normalized difference vegetation index (NDVI) and normalized canopy temperature (canopy temperature - air temperature) were tested as predictors of disease by several algorithms. Among them, logistic regression analysis (LRA) trained on NDVI data showed the highest sensitivity and lowest rate of false negatives. This algorithm based on NDVI could be a quick and feasible method to detect trees potentially affected by white root rot in avocado orchards.
白根腐病是一种由土壤传播真菌引起的疾病，它严重限制了多种木本作物（如鳄梨树）的生产。目前，白根腐病的检测方法主要基于微生物和分子技术，其在果园规模上的应用受到限制。在这项研究中，我们通过机器学习方法分析了成像技术提供的生理参数。通过几种算法，测试了归一化差异植被指数（NDVI）和归一化冠层温度（冠层温度-空气温度）作为疾病预测因子的效果。其中，基于 NDVI 的逻辑回归分析（LRA）显示出最高的灵敏度和最低的假阴性率。这种基于 NDVI 的算法可以成为一种快速、可行的方法，用于检测鳄梨果园中可能受到白根腐病影响的树木。
