Image Analysis

overview

Invicro’s medical image analysis team develops robust algorithms and pipelines to solve complex image science problems. Our team is responsible for extracting scientific quantitative data in support of Invicro’s studies from discovery research through late-phase clinical trials.

The medical image analysis team blends cutting edge image science with proven image engineering capabilities to continually create novel algorithms and quantitative image analysis pipelines while also developing those methods into robust, efficient automated tools. Machine learning and deep learning are two powerful tools routinely used to drive these solutions. These methods are coupled with Invicro’s deep scientific domain knowledge to produce unique, custom solutions to a variety of tasks, spanning therapeutic area and modality.

Capabilities

We support a wide-range of quantitative image analysis tasks and the execution of these tasks is considered as a function of the trial phase, subjects and species of interest, imaging modalities, and therapeutic areas. We then utilize or develop the tools needed to support these tasks.

Task(s)	Phase	Species	Modality	Disease Area	Tools
Quantitation	Discovery	Human	PET / SPECT / Quanticell	Neurology	Image Processing
Classification	Translational	NHP	MRI / MRS	Oncology	Modeling
Estimation	Early (0, 1, 2)	Sheep/Pig	CT / uCT	Cardio	Machine Learning
Automation	Late (2, 3, 4)	Dog/Cat	BLI / Epi-FL / CFT	Pulmonary	Simulation
Visualization		Rabbit/Rat	CIQA / ARG / MARG	Musculo-Skeletal	Registration
Segmentation		Mouse	US	Infectious Disease	Filtering
SW Deployment		Other			Statistics
					Atlases

Invicro has also applied machine learning and deep learning methods to projects across various therapeutic areas and imaging modalities, from basic discovery research through late-phase clinical applications. By continuously building upon our team’s broad range of skill sets and experience, new problems may be addressed using pre-existing trained networks through creation of de novo analysis pipelines or via a combination of existing architecture and study-specific feature extraction.

Focus Areas

Neuro PET Analysis & SPECT Analysis

The image analysis team provides industry-leading tools for Neuro PET analysis and SPECT analysis, supporting quantification projects from discovery research through late phase trials. Invicro’s teams have analyzed thousands of clinical scans for Alzheimer’s Disease studies, utilizing validated SUVR pipelines, Centiloids and the state-of-the-art Amyloid^IQ and Tau^IQ algorithms for Amyloid imaging and Tau imaging, respectively. Some examples include:

The novel IQ Analytics suite for Tau Imaging, Amyloid imaging and DaT Imaging
A range of flexible Biomathematical Modeling tools
Validated pipelines across multiple quantitative outcomes
Preclinical and clinical support for Tracer Characterization
Sophisticated batch Image Triage and preprocessing methods

For more information, check out some of our most popular case studies. A full list is also available in our case studies section!

Amyloid Load

Dosimetry Analysis

Tracer Kinetic Analysis

Quantitative Oncology

The medical image analysis team supports standard analysis of FDG PET imaging, including PERCIST and other SUV-based metrics, metabolic tumor volume and total lesion glycolysis. The image analysis team also specializes in developing and applying a broad range of tools to support advanced quantitative oncology tracer characterization and analytics across a broad range of modalities and studies. Some examples include:

Powerful, multi-modal tumor and organ Segmentation tools
A range of flexible Biomathematical Modeling tools
Dosimetry calculations across radionuclide, administration route, and species
Integrated Radiomics pipelines using standard and custom feature vectors
Preclinical and clinical support for Tracer Characterization
Sophisticated batch Image Triage and preprocessing methods

For more information, check out some of our most popular case studies. A full list is also available in our case studies section!

Dosimetry Analysis

Radiomics

Biodistribution

Multi-Modality – Oncology

MRI Analysis

The medical image analysis team’s multidisciplinary MRI core builds MRI image processing workflows to support brain and whole-body applications across a huge array of studies. Examples include:

Preclinical and clinical support for Functional MRI (fMRI), Pharmacologic MRI (phMRI) and Resting State fMRI (rsfMRI) studies
Validated Pipelines across multiple sequences and analysis methods
Analysis of body MRI such as LipoQuant for NASH
Powerful, validated and exploratory tools for Volumetrics and Cortical Thickness
Flexible DCE-MRI Analysis pipelines across species
Neuro study expertise spanning Arterial Spin Labeling (ASL MRI), Diffusion Weighted Imaging (DWI)/Diffusion Tensor Imaging (DTI), MR Spectroscopy (MRS), QSM, and many more…

For more information, check out some of our most popular case studies. A full list is also available in our case studies section!

Volumetrics Analysis

Functional MRI

Lipid Quantification Tool Development Using Multi-Echo MRI

Preclinical Imaging

The image analysis team has extensive experience supporting preclinical data and data analysis across a wide range of species and imaging modalities.

Species data analyzed include: mice, rats, canines, felines, non-human primates, rabbits, ferrets, pigs, and more – contact Invicro for a comprehensive list.
Modalities analyzed include: PET, SPECT, MRI, CT, US, CFT, CIQA, ARG, X-Ray, MSOT and more – contact Invicro for a comprehensive list.
~3 Million anatomical ROIs generated
Proven, Automated Workflows supporting in-house and externally acquired data
Support for any preclinical imaging Data Format

For more information, check out some of our most popular case studies. A full list is also available in our case studies section!

Biodistribution

Rhesus Teeth Segmentation by Use of Multi-Atlas Library

Combined in vivo and ex vivo Study of Canine Pancarpal Arthrodesis

Big Data & Artificial Intelligence (AI)

The medical image analysis team’s machine learning core manages data organization and structuring as well as machine learning and deep learning for medical image analysis development for a large variety of studies and applications.

Development and deployment of Machine Learning (ML) algorithms
Development and deployment of Deep Learning (DL) for medical image analysis algorithms
Management and structuring of large Multi-Modal Data Repositories
Integration of ML and DL tools with Invicro’s iPACS data management system
Customized pipelines for Image Triage across many applications
Combining AI and Domain Knowledge to build powerful, custom tools

For more information, check out some of our most popular case studies. A full list is also available in our case studies section!

Image Triage

Machine Learning

Radiomics

Modalities

Invicro’s image analysis, medical image analysis, medical image processing and quantitative image analysis workflows are modality agnostic. We specialize in analysis for radiology and pathology applications. Modalities supported by our image analysis workflows include: