GALILEO EDUCATION

User Guide

  • Device Overview

    This user manual describes GBrain MRI functionalities and operation.
    CAUTION: Federal law restricts the medical device to sale by or on the order of a qualified clinician.

    Intended Use and Indications for Use

    GBrain MRI is a fully automated post processing medical device software intended for analyzing known signal hyperintensities in brain MR images. It is not intended for detection of any disease or signal.

    GBrain MRI is intended to aid radiologists in quantitatively reporting the extent of FLAIR hyperintensities and or contrast enhancement hyperintensities in the corresponding brain MRI series. GBrain MRI is intended to provide automatic segmentation, quantification, and reporting of derived image metrics.

    GBrain MRI should not be used in-lieu of a full evaluation of the patient’s MRI scans. The physician retains the ultimate responsibility for making the final diagnosis and treatment decision.

    CAUTION:

    • Users should only upload T2w FLAIR series when the user has confirmed the presence of hyperintensities and wishes to measure their volume. GBrain MRI is not intended to be used to detect the presence of hyperintensities.
    • Users should only upload T1 pre and post contrast series when the user has confirmed the presence of contrast enhancement and wishes to measure its volume. GBrain MRI is not intended to be used to detect the presence of contrast enhancement.

    Intended User

    • To ensure safe use of GBrain MRI, the device is intended for use by a clinician, physician, or radiologist who works with medical imaging.

    Intended Patient Population

    • The GBrain MRI is intended to be used for adult patients over 18 years of age undergoing standard clinical brain MRI scans. GBrain MRI is not intended for use in pediatric populations, nor in patients who have undergone brain tumor or other resection.
    • GBrain provided segmentations are not intended for direct image guided treatments, i.e. radiotherapy or surgical tumor resection.

    Use Environment

    • GBrain MRI is intended to be used in the inpatient or outpatient clinical setting in conjunction with ACR certified MRI scanners under a board certified radiologist’s supervision.

    User Training

    • Required user training involves reading the entirety of this User Manual and watching the GBrain MRI Getting Started video prior to using the device.

    Contraindications

    • GBrain MRI is contraindicated for use in pediatric patient populations, and in patients who have undergone brain tumor or other resection surgery.

    Prerequisites

    To operate GBrain MRI safely and according to its intended use, the following prerequisites must be met:

    • The user has read and understood the Indications for Use, Contraindications, Warnings, and Operating Instructions included in this user manual.

    Warnings

    • Do not use GBrain MRI without review of this User Manual first.
    • GBrain MRI provided segmentations are not intended for direct image guided treatments, i.e. radiotherapy or surgical tumor resection. They are not intended as sole criteria for any treatment.
    • GBrain MRI is not currently designed to be used in patients who have had tumor or other resection surgery.
    • GBrain MRI must not be used for the detection of T2w FLAIR hyperintensities nor T1 Contrast Enhancement. Radiologists should confirm the presence of T2w FLAIR hyperintensities, or T1 Contrast Enhancement before using GBrain MRI.
    • GBrain MRI is not designed to measure the volume of hyperintensities outside the brain, for example in the sinuses or the skull.
    • Radiologists should confirm the appropriateness of the hyperintensities used in the volume calculations by looking at the segmentation overlays provided by GBrain MRI, before using the volume and maximum diameter.
    • Radiologists should review the GBrain MRI annotated images concurrently with original images before using the calculated volumes provided by GBrain MRI.
    • Radiologists should review original images for all suspected pathologies. GBrain MRI is an adjunct tool and does not replace the role of the radiologist. Radiologists must not use the GBrain MRI generated output as the primary interpretation.
    • All images must have proper DICOM tags. GBrain MRI performance may be reduced on images that do not have correct DICOM tags.
  • Device Description

    Summary

    • GBrain MRI is designed to aid the radiologist in measuring and reporting abnormally increased (hyperintense) T2w FLAIR and contrast enhancement, by providing automatic quantitative size data, specifically maximum diameters and volumes. Automated measurements generally minimize subjectivity, decrease both intra as well as inter-rater variability, reduce costs, and save radiologist time.
    • While not a specific disease diagnosis, these measurements may subsequently be used for clinical management at the radiologist's discretion. Measurements are often used in evaluating disease extent, disease progression, and response to treatments. Due to human nature, technical and time constraints, radiologists tend to report size qualitatively (e.g. small, medium, large). However, it is generally agreed that quantitative reporting of size is more desirable.

    Workflow

    • NOTE: While GBrain MRI is designed to measure the volumes of hyperintense T2w FLAIR and abnormal contrast enhancement, it should not be used as a detection tool for these image features. GBrain MRI should only be used when the radiologist or clinician has confirmed the presence of these image features, and wishes to measure them.
    • From a workflow perspective, GBrain MRI is packaged as a computing appliance that is capable of supporting DICOM file transfer for input and output of results. The software is designed without the need for a user interface after installation. Any processing errors are reported either in the output series report, or in the system log files.
    • The proposed Brain MRI workflow is shown in the figure above. A radiologist reading the patient’s study notices the presence of FLAIR hyperintensities, or Contrast Enhancing tissue, and wishes to measure the volume. The radiologist sends the T2w FLAIR, T1 pre and post contrast series to GBrain MRI device from the PACs. GBrain MRI sends back a report containing the total volume of hyperintensities, and two new series, “GBrain FLAIR” & “GBrain Contrast”, both with segmentation overlays, to show the radiologist which hyperintensities were used to calculate the total volumes in each of those series.
    • The radiologist reviews the segmentation overlays, along with the original MR images to assess whether they agree with the hyperintensities used in the volume calculations. If they agree with the choice of hyperintensity regions used, they can include the volume information in their report or decision making process. However, if they do not agree with the regions of hyperintensities used by GBrain MRI, they must make measurements manually, using methods they would have used in the absence of GBrain MRI.

    Principles of Operation

    • GBrain MRI primarily relies on trained neural networks to segment hyperintensities in the T2-weighted Fluid-attenuated inversion recovery (FLAIR) and post contrast T1 sequences. GBrain relies on separate neural networks for the FLAIR and for the Contrast Enhancement measurements. The masks generated by these neural networks are then used to calculate the total volumes of these hyperintensities. GBrain MRI uses specially trained Convolutional Neural Networks using the U-Net architecture.
    • Next, for the identified hyperintensities, the maximum diameter measurement in the axial plane is presented to the user. This maximum diameter measurement is brute force calculated by calculating the distance between every two points on the boundary of the largest contiguous segmented region, on each axial slice of that region. Then the largest of all the maximum diameters measured across all slices is presented to the user.

    Device Inputs

    For the measurement of hyperintensities in T2w FLAIR, GBrain MRI accepts the following MRI series as inputs.

    • T2-weighted Fluid-attenuated inversion recovery (FLAIR)

    For the measurement of contrast enhancement, GBrain MRI accepts the following MRI series as inputs. Both the pre and post contrast series are necessary for the measurement of contrast enhancement.

    • T1-weighted pre-contrast
    • T1-weighted post-contrast
    Scan Parameters

    The input series should conform to the following parameters for best results.

    Deviation from the recommended protocols can result in unexpected GBrain MRI outputs, and those protocols should be validated prior to clinical use.

    FLAIR and Post Contrast images must be axial for the device to work correctly.

    Parameter T2-weighted FLAIR T1-weighted T1-weighted with contrast
    Manufacturer Siemens, Philips, GE, Canon Siemens, Philips, GE, Canon Siemens, Philips, GE, Canon
    Field Strength 1.5T or 3T 1.5T or 3T 1.5T or 3T
    Scan Mode 2D, 3D 2D, 3D 2D, 3D
    Plane Axial Axial or Sagittal Axial
    Tr (ms) 5000 – 13,930 ms 250 – 2212 ms 250 – 2212 ms
    Te (ms) 85 – 394 ms 2.87 – 20 ms 3.11 – 20 ms
    Ti (ms) 1800 -2800 ms 0 – 950 ms 0 – 950 ms
    Flip Angle Variable 90 90
    Fat Suppression Yes or no Yes or no Yes or no
    NEX/NSA >= 1 >= 1 >= 1
    Slice Thickness 1 to 5 mm in the axial plane 1 to 5 mm in the axial plane 1 to 5 mm in the axial plane
    Spacing between Slices 1 – 5 mm 1 – 5 mm 1 – 5 mm
    Number of Slices 16 – 240 20 – 384 16 – 192
    FOV 180-240 mm 180-240 mm 180-240 mm
    Voxel Spacing 0.41 – 1.015 0.429 – 1.145 0.39 – 1.145
    Needed Metadata >Patient ID, Patient Age/DOB, and Scan Date (unless this is handled by the AI Marketplace to confirm age appropriateness of patient)
    Exclusions No patients with shunts or other artifact causing items.
    No patients with resections, or any intracranial surgery
    No pediatric patients.

    Device Outputs

    Five outputs will be created by the GBrain MRI device, when processing has been completed. These outputs are sent back to the PACs.

    First two segmentation overlays will be created in separate DICOM series.

    • GBrain FLAIR: This series shows hyperintensities in FLAIR with a semi-transparent red overlay. Additionally, the volume of the hyperintensities and the maximum diameter are annotated on the series.
    • GBrain Contrast: This series shows contrast enhancement in the T1w post-contrast image with a semi-transparent red overlay. Additionally, the volume of the contrast enhancement and the maximum diameter are annotated on the series.

    Additionally reports are created in two formats.

    • A DICOM encapsulated pdf report: This report is provided so that a user can share information outside the normal PACS environment if required.
    • A DICOM Structured Report: This report is provided so that a user can review the text summary in the normal PACS environment.

    Lastly, since some PACS are unable to open pdf reports, a DICOM secondary capture image is created from the pdf report.

    • DICOM secondary capture image of the pdf report: This allows a user to view the pdf report on the PACS. Each page of the pdf report is a “slice” in the secondary capture.

    💡 It is imperative for users to review the GBrain MRI annotated DICOM series alongside the patient’s original MRI series. The utilization of volume measurement data should be contingent upon the user’s satisfaction with the accuracy of the segmentation, as indicated by the red overlay produced by the device.

    Example Device Outputs

    Example GBrain FLAIR

    The image below shows a slice in the GBrain FLAIR series, which is a composite image of the patient’s current original scan, the same original image with an annotation for the maximum diameter that was calculated from the segmentations, and the same original scan with all the areas of segmented hyperintensities shown with a semi-transparent red color overlay. The user must agree with the segmented regions in order to use the volumetric and maximum diameter measurement.

    Understanding the GBrain FLAIR series

    The above figure shows the data presented in the GBrain FLAIR series. The total volume is the volume of all FLAIR hyperintensities that the GBrain device found, and were annotated in red for review by the user. The max diameter is the maximum diameter of the largest (in terms of volume) contiguous red region amongst all the regions that were annotated in red.

    Example GBrain Contrast

    The image below shows a slice in the GBrain Contrast series, which is a composite image of the patient’s original post contrast scan, the same original scan with an annotation for the maximum diameter that was calculated from the segmentations, and the same original scan with all the areas of contrast enhancement having a semi-transparent red color overlay. The user must agree with the segmented regions in order to use the volumetric and maximum diameter measurement.

    Understanding the GBrain Contrast series

    The above figure shows the data presented in the GBrain Contrast series. The total volume is the volume of all contrast enhancing regions that the GBrain device found, and were annotated in red for review by the user. The max diameter is the maximum diameter among all the regions in all the slice that were annotated in red.

    Example PDF Report

    An example of the pdf report is attached in Appendix A of this document.

  • Device Operating Instructions

    Login

    • When deployed on-prem, there is no login information associated with the GBrain device at the individual user level.

    Image Upload

    • When deployed on-prem, the GBrain MRI device is typically shown as a separate DICOM server or destination on the PACS.
    • Users using the PACS can send a specific series belonging to a patient study to the GBrain MRI device. Users must send only the T2-w FLAIR, pre-contrast and post-contrast T1-w series to the GBrain MRI device.
    • When deployed on an AI Marketplace, the user must use the AI Marketplace vendor’s instructions to send specific MR series from a patient study to the GBrain MRI device.

    CAUTION:

    • Users should only upload T2w FLAIR series when the user has confirmed the presence of hyperintensities and wishes to measure them. GBrain MRI is not intended to be used to detect the presence of hyperintensities.
    • Users should only upload T1 pre and post contrast series when the user has confirmed the presence of contrast enhancement and wishes to measure its volume. GBrain MRI is not intended to be used to detect the presence of contrast enhancement.
  • System Requirements

    As a software application, the software device does not operate on its own and can only function in conjunction with computer hardware and software. The use of the application is restricted to the specific computer hardware and software outlined in the user documentation. The manufacturer provides the software device solely for its designated medical purpose, and the software itself is considered a medical device. Any other necessary components, hardware, or software required to ensure the workstation’s functionality (e.g., operating systems, network structures, and displays) are not included. This also encompasses potentially required runtime dependencies, even if provided as part of the installation package for user convenience. However, the effects of such components on the medical device are considered and managed.

    The capabilities and performance of the software device may be affected by limitations and failures of the hardware, network, or other installed software on the system. Therefore, proper operation and maintenance of the hardware and network are critical. It is necessary to adhere to the manufacturer’s instructions concerning hardware and software specifications, as well as software maintenance.

    Hardware Requirements

    • Standard off-the-shelf commercially available computer hardware that has
      • 1. At least 2 CPUs
      • 2. Greater than 8 GB of RAM
      • 3. A GPU with at least 8 GB RAM
      • 4. Storage requirements are variable depending on the size of the obtained images
  • AI Marketplace Setup

    Installation and setup instructions are specific to each AI Marketplace, since every marketplace has different input and output protocols for studies as well as reports.

    Contact Galileo for assistance in installing GBrain MRI for your specific AI Marketplace vendor.

    As Galileo GBrain MRI is available on different marketplaces, the list will be updated here

    AI Marketplace Docker Command

      • The AI Marketplace will invoke the following Docker run command:
      •  docker run --gpus all -v tmp/input:/workspace/input -v tmp/output/:/workspace/output --cap-add SYS_ADMIN --device /dev/fuse --security-opt apparmor:unconfined gbrain /bin/bash run.sh
              
      • Note: to specify a specific GPU, replace the --gpus all option above with --gpus '"device=3"'
  • On-Prem Setup

    A trained IT staff on the customer’s site must follow the following steps to deploy GBrain MRI on their VM. The VM will be provided by the customer. GBrain MRI will only provide the software.

    Deployment

    To deploy GBrain MRI with it’s default settings to a VM, do the following:

    • Galileo will send customer the setup bash script onprem.sh
    • The customer must ensure the following are installed on the VM
      • Docker
      • Docker Nvidia drivers
    • Ensure that the following ports are opened on the VM
      • Port: 4242 (the DICOM port)
      • Port: 8042 (the port for accessing the Orthanc web portal)
    • Run the setup bash script with the following parameters, where the PACS is the PACS used by the radiologists to view their images. This is where the GBrain device will send its output DICOM files.
      • Usage: onprem.sh <PACS_AET> <PACS_IP> <PACS_PORT>
      • PACS_AET: the AET of the customer’s PACS
      • PACS_IP: the IP address of the customer’s PACS
      • PACS_PORT: the port on the PACS that will be used for receiving DICOMs from GBrain
    • Configure the GBrain AET on your PACS with the following parameters
      • AET: GBRAIN
      • IP: <the internal IP address of the VM>
      • Port: 4242
    • All logs will be routed to stdout and stderr. The customer can pipe them to local storage somewhere.

    Enable SSL (HTTPS) for DICOM

    To enable SSL for DICOM queries within the customer’s intranet, do the following:

    1. 1. Obtain a X.509 certificate in the PEM format.
    2. 2. Prepend this certificate with the content of your private key.
    3. 3. Open up the onprem.sh file sent to you by GBrain MRI and Update the following fields in the Orthanc DICOM server configuration:
      1. SslEnabled: Set this to true
      2. SslCertificate: Set this to the certificate that you obtained in step 1.

    Encryption at Rest

    GBrain MRI is designed to run as a Docker container on a Linux based operating system. We recommend that trained IT staff at the customer’s site enable encryption at rest on the host operating system and the storage volume that is mounted to the Docker container.

    Here are some encryption systems we recommend. The exact system used should be determined by the IT staff based on best practices as the customer’s site and the specific Linux distribution deployed on the VM.

    Security Scanning

    GBrain MRI is designed to run as a Docker container on a Linux based operating system. We recommend that trained IT staff at the customer’s site enable security scanning tools on the host operating system.

    Here are some security scanning systems we recommend. The exact system used should be determined by the IT staff based on best practices as the customer’s site and the specific Linux distribution deployed on the VM.

    Maintenance

    GBrain MRI recommends that trained IT staff at the customer’s site

    • Review the device logs on a monthly basis
    • Ensure that the operating system packages on the provisioned VM are kept up to date.
    • Monitor the space usage of the VM and periodically clear the temporary cache to avoid device downtime.
    • Update the PACS configuration details such as AET, IP address and port if settings change other parts of the customer’s IT infrastructure.
  • Cybersecurity

    Responsibilities

    Customers must ensure the system is used responsibly by trained facility staff. This includes but may not be limited to:

    • Monitoring proper use of the SaMD when deployed on-prem.
    • Ensuring effective administration of access to the SaMD.
    • Providing effective staff training.
    • Customers should maintain proper updates for all devices related to this product.

    Configuration

    When deployed to an AI Marketplace, cybersecurity configuration will be handled by the marketplace itself.

    When deployed on-prem, customers must provide hardware for the GBrain MRI to run on. The provisioning, cybersecurity, and configuration of this hardware is the responsibility of the customer. The provisioned hardware must meet the following requirements:

    • The hardware requirements specified above
    • An up to date Linux based operating system that supports Docker
    • An operating system that supports Nvidia GPUs and the relevant drivers

    SBOM

    A software bill of materials (SBOM) will be provided upon request. Please email info@galileocds.com. Please provide the GBrain MRI version number with your request.

    The SBOM is updated automatically for each released version of GBrain MRI and it uses the SPDX 2.2 file format.

    Ports and Networking

    When deployed via an AI Marketplace, all the relevant networking and port changes will have already been made by the marketplace when it integrates with the hospital’s IT ecosystem. No additional networking and port changes outside of that will be required for GBrain MRI.

    When deployed on-prem, IT staff will need to configure the GBrain MRI DICOM SCP with their PACS with the AE Title “GBRAIN”. The IP address and port for the DICOM SCP will be determined during device onboarding.

    Backup and Restore

    GBrain MRI does not store any medical images long term. Those data will only be stored in the customer’s PACS system.GBrain MRI will only temporarily access those data for analysis and report generation purposes. Data, i.e. input MR images, are only stored as long as they are being processed. Once processing is complete the docker container is stopped, and the container along with all data stored within it is deleted.

    The customer is strictly responsible for all backups of their medical images on their own PACS.

    Customers are also responsible for backups of GBrain MRI logs. When deployed to an AI Marketplace, these backups will be managed by the marketplace itself. When deployed on-prem, we recommend that IT staff set up an recurring background job to do a file system backup of the VM that the device is running on.

    GBrain MRI does not have a database and does not rely on persistent data whether deployed to an AI Marketplace or on-prem. The device is stateless. No data outside of the software is required to restore the device back into proper functioning.

    Responding to Cybersecurity Incidents

    When deployed to an AI Marketplace, the marketplace will manage cybersecurity incident response.

    When deployed on-prem,GBrain MRI will run on hardware that the customer provides. The security of this device is the responsibility of the customer. The only data source GBrain MRI communicates with is the customer’s PACS. There are no other external connections.

    If a cybersecurity incident is detected, we recommend customers monitor the activity on their PACS for abnormalities.

    Logs

    GBrain MRI will log the following events:

    • Timestamps for when new data is sent to the device for processing and timestamps for when the processing is complete.
    • Core device algorithm troubleshooting information such as warnings and info messages from various stages of the algorithm.
    • Troubleshooting details related to the AI model that is running on the device.
    • Startup and shutdown of the device.
    • Other device warnings and errors

    End of Life

    To stop using the GBrain MRI:

    • Uninstall the software
      • On-prem: terminate and wipe the VM that the device is running on.
      • AI Marketplace: contact the marketplace and ask GBrain MRI to be removed.
    • Remove any DICOM SCP configurations on the PACS.
    • Email info@galileocds.com and notify the support team that you plan to no longer use GBrain MRI.
  • Product Information

    Account Registration

    On Premise Installation

    There are no users accounts for an on-prem installation. The installation should be done by trained IT administrators and PACS administrators from the customer site, with assistance from the Galileo team.

    AI Marketplace Installation

    There are no users accounts for an on-prem installation. The installation should be done by trained IT administrators and PACS administrators from the customer site, with assistance from the Galileo team.

    Software Installation

    Clone the private repository via Git using the following command –

    git clone https://github.com/galileo-cds/GBrain

    Transfer the MODEL zip files provided by Galileo to the GBrain Repository.

    mv GBRAIN_MODELS/* GBrain

    Add DICOM Series names based on the hospital location for FLAIR, T1 and T1CE in the following json file –

    GBrain/seq-id/hospital_heuristics/general.json

    following is an example of how you can add a new series name –

    { "FLAIR": [ "3D CUBE T2 FLAIR AX REF", "3D T2_FLAIR_SAG_ISO FS POST", "NEW FLAIR NAME" ## new ], "T1": [ "AX 3D BRAVO", "AX 3D T1 BRAVO", "NEW T1 NAME" ## new ], "CT1": [ "AX 3D T1 MPRAGE ISO POST", "AX 3D T1 MPRAGE POST", "NEW CT1 NAME" ## new ] }

    Note that the values in this json file must match exactly the value that appears in the DICOM SerriesDescription (0008, 103E) header.
    Once this is completed, run the following command to complete the installation:

    sudo docker build . –tag gbrain

    User Manual Access and Product Label

    The latest version of the GBrain MRI user manual can be accessed at the following link https://galileocds.com/index.php/GBrainManual/.

    Additionally, links to the GBrain MRI user manual are also added to the pdf reports generated by the GBrain device.

    The GBrain MRI product label can be accessed…

    Account Deletion

    There are no user accounts associated with an on-premise deployment of GBrain MRI. To delete the GBrain MRI device itself, terminate and wipe the VM that the GBrain MRI device is running on.

    User accounts associated with an AI Marketplace deployment are handled by the vendor of the AI Marketplace.

  • Customer Support

    If you have any questions, problems, or comments regarding GBrain MRI functionalities or this user manual, please contact the Customer support team:

    Email: customer.support@galileocds.com

    Troubleshooting

    This section contains guidelines for users to identify and resolve basic issues that may arise while using GBrain MRI. For additional customer support, please contact customer.support@galileocds.com.

    Problem Cause Solution
    Docker run command returns Nvidia driver related error. Nvidia Docker Driver not installed. Make sure to install Nvidia Docker Driver.
    No DICOMs found in output folder. Docker container not mounted to the correct volumes. Double check that volumes that are mounted onto the Docker container.

    Unknown Anomalies

    TODO: need to fill out

  • Glossary

    Regulatory Symbols
    Symbol Symbol Title / Reference Number Symbol Definition Standard Title
    Manufacturer
    (5.1.1)
    Indicates the medical device manufacturer ISO 15223-1: 2021-07 Medical Devices – Symbols to be Used with information to be supplied by the manufacturer – Part 1: General requirements
    Serial Number
    (5.1.7)
    Indicates the manufacturer’s serial number so that a specific medical device can be identified ISO 15223-1: 2021-07 Medical Devices – Symbols to be Used with information to be supplied by the manufacturer – Part 1: General requirements
    Consult instructions for use or consult electronic instructions for use
    (5.4.3)
    Indicates the need for the user to consult the instructions for use ISO 15223-1: 2021-07 Medical Devices – Symbols to be Used with information to be supplied by the manufacturer – Part 1: General requirements
    Prescription use only Indicates that Federal law restricts the medical device to sale by or on the order of a qualified clinician. 21 CFR 801.109 Prescription devices
    Terms and Definitions
    Term Definition
    Study An imaging study is a radiologic exam/test/scan uploaded and submitted to the platform for reading, diagnosis, and interpretation by a well-trained and qualified radiologist.
    Modalities Imaging modalities refer to the different techniques used to create medical images, such as X-ray, CT, and MRI. Each modality uses a different method to produce images and is best suited for certain types of diagnostic exams.
    Medical professionals Medical personnel and medical professionals are the radiologists, physicians, and clinicians, who receive, ask for, or give feedback about radiologic reports and medical images, in hospitals, and clinics.
    AI Marketplace An AI Marketplace refers to a marketplace where medical device algorithms from multiple vendors are available for healthcare organizations to try, evaluate, and install for their use. The most common AI Marketplaces are Nuance’s Precision Imaging Network, Sectra’s Amplifier Marketplace, Bayer Calantic Digital Solutions, Siemens Healthineers Digital Marketplace, and the GE Healthcare marketplace.

    Galileo
    7000 N Mopac Expy, Suite 200, Austin, TX 78731

Interested? Contact Us at 

Email: education@galileocds.com

Url: https://galileocdslearning.com/