Getting Started

AltaDB is a performant and flexible medical imaging cloud storage alternative to systems like AWS s3, Google Cloud Storage, and Orthanc. AltaDB provides a modern interface for storing DICOM data in the cloud, with medical imaging-specific features and organization.

This document walks you through getting started with AltaDB.

Signup and Organization Creation

1. Head to https://app.altadb.com/createaccount to sign up for a new account.

2. A confirmation code (sent to your e-mail) is required to complete account creation. If you don't receive a confirmation code, please check your spam.

3. After creating an account, create an Organization. Organizations are the parent entity of AltaDB, and all of your colleagues (Team Members) and assets (data) live inside of your Organization.

4. Invite collaborators from the Team tab.

1. AltaDB has a single permission level - Admin. Admins have access to all assets within an Organization.

2. API keys give developers Organization-wide access to all assets.

Dataset Creation and Importing Data

In a single Organization, you can organize your data into multiple Datasets. A single Dataset is a collection of several DICOM studies and / or individual series.

All DICOM data is indexed by series using the Series Instance UID DICOM headers and automatically organized into series within your dataset.

Once indexing is complete, you can view all DICOM headers and search/filter your dataset by Series Instance UID, Study UID, or Patient ID.

AltaDB v1.1.2 will not maintain any folder structure you upload. The folder structure will be traversed, and individual series will be extracted and indexed.

If AltaDB encounters an invalid DICOM file in a Series, the entire Series will be ignored.

Creating a Dataset

In the UI, you can easily create a Dataset by clicking on the + button next to the Datasets section in the lefthand toolbar or clicking on Create Dataset while on the Datasets Page.

Creating a Dataset in the AltaDB UI

In the CLI, after installing and authenticating AltaDB, you can create a Dataset with the following command:

altadb create [-h] dataset

# Create a dataset named "brats_batch_July_08_2024"
altadb create brats_batch_July_08_2024 

Importing Data

In the UI, you can take the following steps to upload your DICOM studies and series:

1. Navigate to a Dataset and open it;

2. Click on Import Data to open the data importer;

3. Either a) Drag & drop your DICOM files into the data importer or b) search for them on your device;

4. Once the upload is complete, AltaDB will index the DICOM headers and re-compress the DICOM pixel data with state-of-the-art compression. This process may take several minutes.

5. You can monitor the status of your imports inside the "Import" tab.

In the CLI, you can use the following command to designate a file path and destination Dataset for your upload (full docs here):

altadb upload [-h] [-n NAME] [-c CONCURRENCY] dataset path

# Upload all valid DICOM files in folder "patient_001" to the "brain_brats" Dataset 
altadb upload -n Aug_15_upload brain_brats ./patient_001

AltaDB SDK Mastery: you can indicate a top-level directory as your file path to recursively upload all valid DICOM files in all subdirectories.

AltaDB recognizes the following file extensions as valid DICOM files:

• *.dicom

• *.dcm

• *.ima

Viewing Data

Inside of a Dataset, you can preview your images by clicking on the thumbnail. The simple viewer offers a view of the imaging axis and supports a few key functions:

1. Change slice: Mouse scroll.

2. Windowing: Right-click+drag.

3. Zoom: Left-click+drag.

4. Pan: Middle mouse+drag.

Previewing images

Searching & Filtering

You can filter your dataset by the following parameters:

1. Series UID.

2. Study UID.

3. Import ID: each import session is assigned a unique import ID, which can be found on the Imports page.

4. Created by: The person who uploaded the series.

5. Imported at: The date when the upload took place.

Exporting JSON from AltaDB

Inside a Dataset, simply select the series that you'd like to export and click on Download RedBrick AI JSON in the snackbar at the bottom of the screen.

You will then be prompted to optionally sort your series entries by study.

If you group by study, the series you select will be grouped into multi-series Tasks for RedBrick AI where the Task name is the StudyInstanceUID.

If you do not group by study, the series you select will be exported as individual single-series Tasks where the Task name is the SeriesInstanceUID.