Evaluating the accuracy of a smartphone-based artificial intelligence system, PlantVillage Nuru, in identifying the viral diseases of cassava

• Published in: bioRxiv
• Main Authors: Mrisho, Latifa M, Mbilinyi, Neema, Ndalahwa, Mathias, Ramcharan, Amanda M, Kehs, Annalyse, Mccloskey, Peter, Murithi, Harun, Hughes, David P, Legg, James P
• Format: Paper
• Language: English
• Published: Cold Spring Harbor Cold Spring Harbor Laboratory Press 30.01.2020
• Subjects: Artificial intelligence; Cassava; Diagnosis; Leaves; Pests; Plant diseases; Researchers
• DOI: 10.1101/2020.01.26.919449

Premise of the study: Nuru is an artificial intelligence system for diagnosis of plant diseases and pests developed as a public good by PlantVillage (Penn State University), FAO, IITA and CIMMYT. It provides a simple, inexpensive and robust means of conducting in-field diagnosis without requiring internet connection and provides real-time results and advice. The present work evaluates the effectiveness of Nuru as an in-field diagnostic tool by comparing the diagnosis capability of Nuru to that of cassava experts (researchers trained on cassava pests and diseases), agricultural extension agents and farmers. Methods: The diagnosis capability of Nuru and that of the assessed individuals was determined by inspecting cassava plants in-field and by using the cassava symptom recognition assessment tool (CaSRAT) to score images of cassava leaves. Results: Nuru's accuracy for symptom recognition when using six leaves (74 - 88%, depending on the condition) was similar to that of experts, 1.5-times higher than agricultural extension agents and two-times higher than farmers. Discussion: These findings suggests that Nuru can be an effective tool for in-field diagnosis of cassava diseases and has a potential of being a quick and cost-effective means of disseminating knowledge from researchers to agricultural extension agents and farmers.