Dataset Information

Shape-to-graph mapping method for efficient characterization and classification of complex geometries in biological images.

ABSTRACT: With the ever-increasing quality and quantity of imaging data in biomedical research comes the demand for computational methodologies that enable efficient and reliable automated extraction of the quantitative information contained within these images. One of the challenges in providing such methodology is the need for tailoring algorithms to the specifics of the data, limiting their areas of application. Here we present a broadly applicable approach to quantification and classification of complex shapes and patterns in biological or other multi-component formations. This approach integrates the mapping of all shape boundaries within an image onto a global information-rich graph and machine learning on the multidimensional measures of the graph. We demonstrated the power of this method by (1) extracting subtle structural differences from visually indistinguishable images in our phenotype rescue experiments using the endothelial tube formations assay, (2) training the algorithm to identify biophysical parameters underlying the formation of different multicellular networks in our simulation model of collective cell behavior, and (3) analyzing the response of U2OS cell cultures to a broad array of small molecule perturbations.

SUBMITTER: Pilcher W

PROVIDER: S-EPMC7494120 | biostudies-literature | 2020 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Shape-to-graph mapping method for efficient characterization and classification of complex geometries in biological images.

Pilcher William W Yang Xingyu X Zhurikhina Anastasia A Chernaya Olga O Xu Yinghan Y Qiu Peng P Tsygankov Denis D

PLoS computational biology 20200903 9

With the ever-increasing quality and quantity of imaging data in biomedical research comes the demand for computational methodologies that enable efficient and reliable automated extraction of the quantitative information contained within these images. One of the challenges in providing such methodology is the need for tailoring algorithms to the specifics of the data, limiting their areas of application. Here we present a broadly applicable approach to quantification and classification of compl ...[more]

PMID: 32881897

Similar Datasets

Project description:PurposePeople with drug-refractory epilepsy are potential candidates for surgery. In many cases, epileptogenic zone localization requires intracranial investigations, e.g., via ElectroCorticoGraphy (ECoG), which uses subdural electrodes to map eloquent areas of large cortical regions. Precise electrodes localization on cortical surface is mandatory to delineate the seizure onset zone. Simple thresholding operations performed on patients' computed tomography (CT) volumes recognize electrodes but also other metal objects (e.g., wires, stitches), which need to be manually removed. A new automated method based on shape analysis is proposed, which provides substantially improved performances in ECoG electrodes recognition.MethodsThe proposed method was retrospectively tested on 24 CT volumes of subjects with drug-refractory focal epilepsy, presenting a large number (> 1700) of round platinum electrodes. After CT volume thresholding, six geometric features of voxel clusters (volume, symmetry axes lengths, circularity and cylinder similarity) were used to recognize the actual electrodes among all metal objects via a Gaussian support vector machine (G-SVM). The proposed method was further tested on seven CT volumes from a public repository. Simultaneous recognition of depth and ECoG electrodes was also investigated on three additional CT volumes, containing penetrating depth electrodes.ResultsThe G-SVM provided a 99.74% mean classification accuracy across all 24 single-patient datasets, as well as on the combined dataset. High accuracies were obtained also on the CT volumes from public repository (98.27% across all patients, 99.68% on combined dataset). An overall accuracy of 99.34% was achieved for the recognition of depth and ECoG electrodes.ConclusionsThe proposed method accomplishes automated ECoG electrodes localization with unprecedented accuracy and can be easily implemented into existing software for preoperative analysis process. The preliminary yet surprisingly good results achieved for the simultaneous depth and ECoG electrodes recognition are encouraging. Ethical approval n°NCT04479410 by "IRCCS Neuromed" (Pozzilli, Italy), 30th July 2020.

Dataset Information

Shape-to-graph mapping method for efficient characterization and classification of complex geometries in biological images.

Publications

Shape-to-graph mapping method for efficient characterization and classification of complex geometries in biological images.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets