Bacterial colony counting with Convolutional Neural Networks in Digital Microbiology Imaging. (January 2017)

• Record Type: Journal Article
• Authors: Ferrari, Alessandro; Lombardi, Stefano; Signoroni, Alberto
• Is Part Of: Pattern recognition. Volume 61(2017:Jan.)
• Page Start: 629
• Page End: 640
• Publication Date: 2017-01
• Subjects: Convolutional Neural Networks -- Deep learning -- Image classification -- Handcrafted feature extraction -- Image analysis -- Bacterial colony counting -- Digital Microbiology Imaging -- Full Laboratory Automation; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2016.07.016
• Languages: English
• ISSNs: 0031-3203

Abstract

Abstract: Counting bacterial colonies on microbiological culture plates is a time-consuming, error-prone, nevertheless essential quantitative task in Clinical Microbiology Laboratories. With this work we explore the possibility to find effective solutions to the above issue by designing and testing two different machine learning approaches. The first one is based on the extraction of a complete set of handcrafted morphometric and radiometric features used within a Support Vector Machines solution. The second one is based on the design and configuration of a Convolutional Neural Networks deep learning architecture. To validate, in a real and challenging clinical scenario, the proposed bacterial load estimation techniques, we built and publicly released a fully labeled large and representative database of both single and aggregated bacterial colonies extracted from routine clinical laboratory culture plates. Dataset enhancement approaches have also been experimentally tested for performance optimization. The adopted deep learning approach outperformed the handcrafted feature based one, and also a conventional reference technique, by a large margin, becoming a preferable solution for the addressed Digital Microbiology Imaging quantification task, especially in the emerging context of Full Laboratory Automation systems. Abstract : Highlights: Bacterial colony counting on solid agar is an essential but challenging computer vision task in lab automation. Comparison between classification solutions for cardinality estimation of colony aggregates is proposed. Large and quality dataset (28.5k images) created and fully labeled for training and validation. Deep neural network compared to handcrafted feature approach and watershed count. Results are unique and relevant in the emerging field of Digital Microbiology Imaging.