FungID: Innovative Fungi Identification Method with Chromogenic Profiling of Colony Color Patterns

Fungi play crucial roles in many ecosystems; however, traditional identification methods are often time- and labor-intensive. In this study, we introduce FungID, a pilot and novel deep learning algorithm, alongside its user-friendly software implementation, developed by analyzing various fungal species for identification based on chromogenic profiling of colony color patterns via a Convolutional Neural Network. Training and testing FungID upon a set of 269 images showed remarkable performance in terms of model robustness and classification efficacy. These findings demonstrate that FungID offers a potential method for rapid and reliable identification of fungal species through chromogenic profiling, providing additional tools to conventional techniques being employed in the fields of health, microbiology, biotechnology, and more. Our research underscores the promising role of deep learning algorithms in enhancing the understanding of the taxonomy and ecological functions of fungi that can be grown in pure cultures, while also emphasizing the importance of carefully assessing the scope and limitations of these methods.