Public Repository

Last pushed: a year ago
Short Description
This image contains a MongoDB based database implementation to manage nuclear segmentation results.
Full Description

This image contains a MongoDB based database implementation, called FeatureDB, to store and manage imaging features computed from segmentation of nuclei in whole slide tissue images.

You can pull the image as follows:

docker pull sbubmi/pathomics_featuredb:1.0

Please clone the following GitHub repository for scripts to interact with the Docker image:

git clone https://github.com/SBU-BMI/pathomics_featuredb

You can interact with the Docker image (start a container, load metadata and data, and kill and remove the container) using the run_docker_featuredb.sh shell script (located in the "script" folder).

Usage: run_docker_featuredb.sh -h|<command> [options]

Commands:

start - start Docker instance
arguments: <docker name> --dbpath <host database folder> --qryport <host port for query server> --dbport <host port for mongodb> [--image <docker image> --root <yes>]
--root <yes> : to run as root in the Docker container. The default is to run as the local user who started the Docker container.

remove - kill and remove Docker instance
arguments: <docker name>

create - create featuredb
arguments: <docker name> <database name>

imgmeta - load image metadata
arguments: <docker name> <database name> <file> <--image <cancer_type> <subject id> <case id> | --metadata>
metadata - file is a QUIP image metadata CSV file containing metadata about multiple images.
image - file is a whole tissue image. Need to provide cancer type, subject id, and case id.

loadfile - load a single file containing analysis results
arguments: <docker name> <database name> <input file> <csv|mask> <additional arguments>
required additional arguments:
--cid <case_id> Case ID (Image ID)
--eid <exec_id> Analysis execution id.
--sid <subject_id> Subject ID
optional additional arguments:
--etitle <title> Analysis title for FeatureDB storage and visualization.
--etype <type> Analysis type: human|computer.
--fromdb Get image metadata from FeatureDB and normalize coordinates.
--namespace <namespace> Namespace for feature attribute names.
--norm <w,h> Normalize polygon coordinates in mask image using
using width (w) and height (h).
--shift <x,y> Shift in X and Y dimensions for a mask image.
--simplify Simplify polygons in CSV files using the JTS library.
--sizefilter <minSize,maxSize> Filter polygons in CSV files based on polygon area.

loadzip - load multiple files stored in a zip file
arguments: <docker name> <database name> <input zip file> <csv|mask> <additional arguments>
required additional arguments:
--eid <exec_id> Analysis execution id.
optional additional arguments:
--etitle <title> Analysis title for FeatureDB storage and visualization.
--etype <type> Analysis type: human|computer.
--fromdb Get image metadata from FeatureDB and normalize coordinates.
--namespace <namespace> Namespace for feature attribute names.
--simplify Simplify polygons in CSV files using the JTS library.
--sizefilter <minSize,maxSize> Filter polygons in CSV files based on polygon area.

loadquip - load quip analysis file collection generated from sbubmi/pathomics_nucleus:1.0 Docker container in a zip file
arguments: <docker name> <database name> <input zip file> <csv|mask> <additional arguments>
optional additional arguments:
--namespace <namespace> Namespace for feature attribute names.
--simplify Simplify polygons in CSV files using the JTS library.
--sizefilter <minSize,maxSize> Filter polygons in CSV files based on polygon area.

Example run:

$ run_docker_featuredb.sh start mydb1 --dbpath /home/<userfolder>/test/db (the folder /home/<username>/test/db must exist).

Here mydb1 is the container name. If mydb1 already exists, you will get an error message.

$ run_docker_featuredb.sh create mydb1 u24_test

Here mydb1 is the Docker container name and u24_test is the name of the database where image metadata and analysis results will be stored.

$ run_docker_featuredb.sh imgmeta mydb1 u24_test TCGA-06-0148-01Z-00-DX1.3b19c82d-c52d-4514-8bf6-5b0f629c18de.svs --image gbm TCGA-06-0148 TCGA-06-0148-01Z-00-DX1

This will load metadata (image width, image height, etc) for image TCGA-06-0148-01Z-00-DX1.3b19c82d-c52d-4514-8bf6-5b0f629c18de.svs to the database.

$ run_docker_featuredb.sh loadquip mydb1 u24_test results.zip csv

This will load results stored in results.zip (which was generated from a segmentation run in the sbubmi/pathomics_nucleus:1.0 Docker container) to the database.

Docker Pull Command
Owner
sbubmi

Comments (0)