Multimodal Cross-Device and Marker-Free Co-Registration of Preclinical Imaging Modalities
October 27th, 2023
The combination of multiple imaging modalities is often necessary to gain a comprehensive understanding of pathophysiology. This approach utilizes phantoms to generate a differential transformation between the coordinate systems of two modalities, which is then applied for co-registration. This method eliminates the need for fiducials in production scans.
We developed software tools for preclinical imaging, including industry construction, fusion, segmentation, and quantification. Our software is meant to be user friendly, efficient, and compatible with any small animal imaging device to enable reproducible quantitative analysis of the acquired multimodal image data. Automation is an important topic in our field because manual analysis is often too time consuming, error prone, and hardly reproducible.
Software-based automation can make the analysis efficient and reproducible, but the challenge is in making the automation robust enough. Small animal imaging devices provide volumetric image data of anesthetized laboratory mice and rats, and allow non-invasive assessment of pathological and physiological changes. Integrated multimodal devices combine the strengths of multiple modalities, such as micro CT and PET.
It is also possible to combine two separate devices. We developed software for the automated fusion of special modalities, such as micro CT and FMT. We also provided efficient software for interactive image fusion and defined a new file format to store the image data considering aspects from data curation.
Last year we published an approach for retrospective respiratory gating and the high throughput settings. Our protocol results in a transformation matrix, which is generated using a calibration step that involves visible structures such as fiduciary markers. Then the image fusion can be automated using this transformation matrix, which eliminates the need for markers.
Begin by loading the images of both modalities. To do so, open the analysis software. To load the CT Image A as underlay, go to File, followed by Underlay, and click Load Underlay.
In the dialogue that appears, choose and open the image file. Similarly, to load the PET Image A as an overlay click File, Overlay, Load Overlay, and choose the image file to open it. For marker based fusion, hide the underlay by navigating to Menu, View, and Layer Visibility.
Then uncheck overlay and keep underlay. Switch the selected layer to the underlay if required. To show multiple slice views, press alt plus A for the axial, alt plus S for the sagittal, and alt plus C for the coronal views.
If necessary, press CTRL plus W and adjust the left and right vertical bars in the windowing dialogue to better see the fiducials. Click on Okay to close the dialogue. Click the marker symbol on the vertical toolbar on the left side to activate the mouse action mode, Create Marker.
Navigate to one fiducial of the phantom. Place the mouse pointer in its center and left click. Select Okay To save the marker, place the mouse pointer over the view of a plane, and use alt plus mouse wheel to slice through the planes.
Adjust the viewing settings by navigating to Menu, View, Layer Visibility, and check Overlay to show the overlay. Switch the selected layer to overlay. Adjust the windowing as demonstrated if necessary.
Navigate to each fiducial of the phantom. Place the mouse pointer in the center of the fiducial and left click. Click on Okay to save the marker.
Place the mouse pointer over the view of a plane and use alt plus mouse wheel to slice through the planes. Align the markers of underlay and overlay by navigating to Fusion, followed by Register Overlay to Underlay, and selecting Compute Rotation and Translation. Note this measurement and click Okay.
Check the result of the alignment. As demonstrated previously, activate the views of underlay and overlay layers. Click on the symbol on the left side vertical toolbar to activate the mouse mode Interactive Image Fusion.
Check overlay, uncheck underlay and segmentation. Perform translation by placing the mouse pointer near the center of a view. Hold the left mouse button and move the mouse to move the overlay.
Perform rotation by placing the pointer near the edge of a view, axial, coronal, or sagittal. Hold the left mouse button and move the mouse to rotate the overlay. Keep performing rotations and translations on the different views until the underlay and overlay are aligned as best as possible.
Create and save the differential transformation by navigating Fusion, Overlay Transformation, and select Create and Save Differential Transformation. In the first dialogue, select the original overlay file and click on Open. In the second dialogue, enter a file name for the differential transformation and press Save.
To begin, launch the analysis software. Import the CT file as an underlay by clicking on File, Underlay, and Load Underlay. In the ensuing dialogue, select the CT image file to open it.
Load the PET file as an overlay by sequentially selecting File, Overlay, and Load Overlay. In the dialogue that appears, choose the PET image file. To enable the display of both layers, go to Menu, View, followed by Layer Visibility, and check both Underlay and Overlay.
To load and apply the previously saved differential transformation matrix, click on Fusion, Overlay Transformation, and then Load and Apply Transform. Then select the file containing the matrix from the calibration process to open it.
Overview
This study presents software tools designed for preclinical imaging, focusing on automation for multimodal image analysis. The tools enhance efficiency and reproducibility in analyzing volumetric image data from small animal imaging devices.
Key Study Components
Area of Science
- Neuroscience
- Imaging Techniques
- Software Development
Background
- Preclinical imaging is essential for understanding physiological and pathological changes.
- Manual analysis of imaging data is often time-consuming and prone to errors.
- Automation can improve the efficiency and reproducibility of image analysis.
- Integrated multimodal devices combine strengths of various imaging modalities.
Purpose of Study
- To develop user-friendly software for automated image fusion and analysis.
- To eliminate the need for fiducials in image registration.
- To enable reproducible quantitative analysis of multimodal imaging data.
Methods Used
- Development of software tools for fusion, segmentation, and quantification of imaging data.
- Automation of image fusion using transformation matrices derived from calibration steps.
- Integration of imaging data from devices like micro CT and PET.
- Implementation of user-friendly interfaces for loading and analyzing images.
Main Results
- Automated fusion of imaging modalities was successfully achieved without fiducials.
- Software tools demonstrated efficiency in handling volumetric data from small animals.
- New file formats were defined for better data curation and storage.
- Retrospective respiratory gating and high throughput settings were validated.
Conclusions
- The developed software enhances the analysis of multimodal imaging data.
- Automation significantly reduces the time and errors associated with manual analysis.
- This approach can facilitate more accurate assessments in preclinical research.
Frequently Asked Questions
Executive Industry Relevance
Robust, marker-free co-registration of preclinical imaging modalities addresses a critical bottleneck in translational research by enabling reproducible, quantitative integration of multimodal datasets. This capability enhances predictive confidence in disease models and supports more reliable target validation and mechanistic de-risking across the discovery pipeline. Streamlined, automation-ready workflows reduce manual variability and facilitate scalable, cross-site imaging studies for enterprise R&D portfolios.
Strategic Applications in Biopharma R&D
Early Discovery & Target Validation
- Enables quantitative, non-invasive assessment of biological changes in small animal models.
- Supports functional target validation by integrating anatomical and molecular imaging data.
- Reduces mechanistic ambiguity through precise spatial correlation of multimodal signals.
- Facilitates portfolio triage by improving predictive value of preclinical findings.
Screening & Assay Development
- Prepares validated imaging datasets for downstream quantitative analysis and screening workflows.
- Standardizes image fusion and co-registration, improving reproducibility across studies.
- Enables scalable, automation-compatible workflows for high-throughput imaging assays.
- Supports reliable evaluation of compound effects in disease-relevant models.
Translational & Preclinical Research
- Aligns imaging outputs with translational biomarker strategies by integrating functional and structural data.
- Ensures continuity from discovery through preclinical validation by supporting cross-modality data integration.
- De-risks advancement decisions by providing robust, quantitative imaging endpoints.
- Improves disease model characterization for translational relevance.
Pipeline & Workflow Integration
This marker-free co-registration workflow bridges early discovery, lead identification, and preclinical validation by enabling seamless integration of imaging data from multiple modalities and devices.
- Discovery Biology: Supports hypothesis testing and pathway clarification through integrated imaging analysis.
- Screening: Delivers reproducible, quantitative imaging outputs for compound evaluation.
- Analytics: Provides transformation matrices and curated file formats for robust statistical comparison.
- Translational Research: Facilitates biomarker alignment and disease model validation across imaging platforms.
- Enterprise Reuse: Offers a scalable, device-agnostic solution for multimodal imaging integration across R&D sites.
Operational & Enterprise Impact
- Scientific Value: Increases predictive confidence and reduces biological risk in preclinical imaging studies.
- Operational Value: Standardizes and automates image fusion, minimizing manual error and variability.
- Strategic Value: Enables more informed go/no-go decisions and capital-efficient portfolio management.
- Portfolio Impact: Supports risk-adjusted prioritization and advancement of discovery programs.
Implementation Considerations
- Requires expertise in preclinical imaging and quantitative analysis.
- Depends on compatible imaging devices and software infrastructure for data integration.
- Necessitates cross-team standardization of imaging protocols and metadata handling.
- Adaptable to various imaging modality combinations and site-specific workflows.
- Phantom-based calibration and transformation matrix generation are essential for initial setup.
Biopharma FAQs
Chapters in this video
0:00
Introduction
2:01
Calculation of the Differential Transformation for Cross‐Device Co‐Registration of Two Preclinical Imaging Modalities
6:10
Application of the Differential Transformation for Cross‐Device Co‐Registration of Two Preclinical Imaging Modalities
