SAMSAM: Segmentation for Analysis and Measurements in the Shoot Apical Meristem

• Grant number: 16/11853-2
• Support Opportunities: Scholarships abroad - Research Internship - Post-doctor
• Start date: October 15, 2016
• End date: April 14, 2017
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Alexandre Xavier Falcão
• Grantee: Thiago Vallin Spina
• Supervisor: Elliot Meyerowitz
• Host Institution: Instituto de Computação (IC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Institution abroad: California Institute of Technology, United States
• Associated to the scholarship: 15/09446-7 - Medical Image Segmentation: How to integrate object appearance/shape models and interactive correction with minimum user intervention?, BP.PD

Abstract

Problem: The shoot apical meristem is a network of cells located at the top of plants responsiblefor all above ground plant growth. Since most of what humans eat, such as grains and fruits, is derived from the shoot, biologists around the world investigate the regulatory mechanisms that drive a healthy growth of the meristem and the crops that depend on it. Computer simulations done on phantom models with idealized cell shapes, sizes, and connectivity have recently shown how transport of proteins and hormones responsible for growth in the shoot can be modeled to generate patterns observed in the wet lab. It is desirable, however, to move to more realistic models derived directly from real data. This type of modeling requires the segmentation of hundreds of cells in 3D confocal microscopy images, for posterior cell lineage tracking using time-lapse imaging. Fully automatic segmentation leads to errors that may be propagated to cell lineage tracking if not fixed. Hence, user-assisted object segmentation must be performed to ensure high accuracy, which calls for robust methods that minimize the required amount of user intervention to prevent mistakes caused by weariness and carelessness Proposal: This postdoctoral research project aims to investigate solutions that integrate in an effective and efficient way automatic image segmentation and interactive correction. In this context, we argue that the appearance of the cells in the images calls for methods that attempt to segment them automatically, providing at the same time a set of suitable seed voxels that can be interactively edited for greater effectiveness. We thus propose to develop interactive and automatic techniques derived from the Image Foresting Transform (IFT) framework to this end, a tool for the design of image processing operators based on optimum connectivity. To further reduce user effort and increase robustness of the result, we shall implement our techniques in an on-line collaborative segmentation platform developed at the California Institute of Technology (Caltech). The users will segment smaller portions of the image, which will then be merged through consensus analysis to obtain the final result. Justification and expected results: The project will be developed at Caltech, under the supervision of Prof. Dr. Elliot Meyerowitz and Dr. Alexandre Cunha, thereby increasing the candidate's international experience and our group's international cooperation network. They are interested in long-term collaboration and have great expertise on the subject, while possessing hundreds of images that can be mined for producing the aforementioned models. We expect that the on-line application will be useful for plant biologists worldwide, and that it may be extended to domains such as medical image segmentation, following the principle of integration between automatic and interactive techniques.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

MARTINS, SAMUEL BOTTER; SPINA, THIAGO VALLIN; YASUDA, CLARISSA; FALCAO, ALEXANDRE X.; STYNER, MA; ANGELINI, ED. A Multi-Object Statistical Atlas Adaptive for Deformable Registration Errors in Anomalous Medical Image Segmentation. MEDICAL IMAGING 2017: IMAGE PROCESSING, v. 10133, p. 8-pg., 2017-01-01. (15/09446-7, 14/12236-1, 13/07559-3, 16/11853-2)

STEGMAIER, JOHANNES; SPINA, THIAGO V.; FALCAO, ALEXANDRE X.; BARTSCHAT, ANDREAS; MIKUT, RALF; MEYEROWITZ, ELLIOT; CUNHA, ALEXANDRE; IEEE. CELL SEGMENTATION IN 3D CONFOCAL IMAGES USING SUPERVOXEL MERGE-FORESTS WITH CNN-BASED HYPOTHESIS SELECTION. 2018 IEEE 15TH INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (ISBI 2018), v. N/A, p. 5-pg., 2018-01-01. (15/09446-7, 16/11853-2, 14/12236-1)

MOREIRA TAVARES, ANDERSON CARLOS; VECHIATTO MIRANDA, PAULO ANDRE; SPINA, THIAGO VALLIN; FALCAO, ALEXANDRE XAVIER; ANGULO, J; VELASCOFORERO, S; MEYER, F. A Supervoxel-Based Solution to Resume Segmentation for Interactive Correction by Differential Image-Foresting Transforms. MATHEMATICAL MORPHOLOGY AND ITS APPLICATIONS TO SIGNAL AND IMAGE PROCESSING (ISMM 2017), v. 10225, p. 12-pg., 2017-01-01. (16/11853-2, 15/09446-7, 11/50761-2, 14/12236-1)