Radiology

ICAMP: Imaging-Centric Abdominal Multiomics Program

Overview

The Imaging-Centric Abdominal Multiomics Program (ICAMP) Shared Resource generates, maintains and facilitates access to longitudinal, highly curated radiology-pathology data linked to clinical, molecular and outcomes information. Our purpose is to catalyze innovative research towards improving the imaging-based detection, diagnosis and treatment of cancer and advancing our understanding of the tumor microenvironment.

Our Vision

To develop an imaging-based multiomic model to understand and predict cancer biology and image-guided therapeutic outcomes.

Our Mission

Catalyze innovative research and tool development through data integration and curation to improve early detection, diagnosis and treatment of cancer.

Acquire high-resolution, spatially registered radiology and pathology images.
Capture high-value clinical and outcomes data.
Spatially track and annotate findings across all observations at the lesion level.
Collect biospecimens (blood, tissue) for multiomic analysis.
Link with institutional resources such as the Institute of Precision Health, internal and external EMR and patient outcomes.
Establish a web of multidisciplinary collaboration across radiology, pathology, genomics, and clinical care to accelerate precision medicine for abdominal cancers.

Explore

Regulation and Oversight

The Research Operations team oversees prospective consenting of participants, manages institutional review board protocols, and serves as an honest broker for data access.

Database and Integration

The Infrastructure team develops and maintains the database, data entry interface, dashboard, integration services and analytics services.

Image Archiving and Annotations

The Image Archiving & Annotations team is responsible for the banking, deidentification and annotation of radiological images for each domain.

Data Management

The Data Management team oversees data quality control activities and fulfillment of data access requests.

Type of Data Available

Clinical and demographical information

Radiological imaging (MRI, CT, ultrasound)

Digital pathology images

Image annotations (organ and lesion segmentation)

Expert-provided image interpretations (semantic features)

Measurements derived from images (radiomic features)

Outcomes information (disease-free survival)

Domains

Prostate

Matched MRI and whole-mount resection cases
Ex vivo imaging (a subset of resection cases)
Prospective MR-guided biopsy cohort

Kidney

Prospective biopsy cohort with prior multiphase contrast-enhanced CT and MRI scans
CT or MRI cases with radiology pathology correlation

Liver

US/CEUS/CT/MRI/quantitation of diffuse liver disease and liver neoplasms, with banked tissue and blood

ICAMP Prostate

ICAMP Prostate is a multidisciplinary cross-collaborative platform created to continuously improve MR/PET/Micro-US imaging with high-resolution histopathological and molecular correlations. Our vision is to unlock the potential of diagnostic imaging, enable detection of enable better patient triage and image-guided therapy, predict cancer biology, and allow for precise tissue acquisition and therapy.

Research Objectives

MR, PET: Validate existing techniques and interventions, and develop and adopt new technologies with a focus on the correlations with pathology and microstructure
AI: Deep learning, machine learning, and federated learning for prostate cancer diagnosis on imaging and digital pathology.
Radiogenomics: Correlation with commercial tests and gene panels, correlation with general genomic panels, and differential gene expression based on PIRADS Score

Highlights and Key Publications

New Techniques for Quantitative T2 and Diffusion MRI of Prostate Cancer
Integrating MRI and Pathology to Characterize Prostate Tissue Microstructure and Improve Prostate Cancer Classification
Personalized MRI analysis methods for accurate and precise DCE-MRI parameter estimation
Artificial intelligence for improved diagnosis of prostate cancer
Multimodal analysis of clinical and quantitative imaging data to improve prediction of biochemical recurrence

Recent Grants and Awards

NIH/NCI R01 (PI: Sung/Wu) Integrating Quantitative MRI and Artificial Intelligence to Improve Prostate Cancer Classification
UCLA Prostate Cancer SPORE (PI: Hsu) Computational Framework for Discovering and Validating Imaging Endotypes to Predict Clinically Significant Prostate Cancer Aggressiveness
2020 SAR Morton A. Bosniak Research Award
2020 SAR Trainee Scholarship Award

Overview

The Prostate ICAMP program has now migrated from a developmental phase to a mature phase where multiple databases have been built with clinical, pathological and MRI data with a mature interface, integration with the imaging biomarkers, and a mechanism for data and imaging retrieval. It is now a template for other programs including liver and kidney. We have finished the initial validation projects and are now implementing ongoing technical development as well as multiple artificial intelligence, federated learning and radiogenomic phases. We are looking forward to integrating the genomic and other biological data for expanded discovery in this new phase.

Available Prostate Data

scroll to right ->
Cohort	Total Cases	Cases Annotated	Cases w/Clinical Data	Cases w/Tissue	Cases w/Blood
mpMRI Cases (incl Resection & Bxs)	30,062	2,042	2,097 (all MRI data avail)	-	-
IPH Overlap	6,695	-	-	-	-
Resection Cases (MR & Path matches: micrographs available)	1,649	1,604	1,649	-	95
3D Mold	663	-	-	-	-
Ex-Vivo	111	110	49	-	-
PSMA-PET (matched to Path, some MR)	520	479	520	-	-
microUS	94	-	-	-	-
HD WMHP Slides	395	-	-	-	-
Biobank MR-Guided Biopsies	448 (20 re-consents)	438	448	433 (18 w/tissue twice)	126

ICAMP Kidney

The goal of Kidney ICAMP is to understand the imaging of small renal masses, explore the use of imaging for the differentiation of kidney masses, including benign and malignant tumors, and predict the biological behavior of renal cancers undergoing image-guided therapy.

Research Objectives

Our research objectives include reviewing the efficacy of the current imaging modalities available today in detecting and helping to characterize small renal masses and using AI to diagnose and report these lesions. We are evaluating outcomes in patients who were managed with surgery and ablative techniques like cryoablation, radiofrequency ablation and so on. We are also improving the prognostic determination of renal cell carcinoma (RCC) based on imaging features by building models or computer-based artificial intelligence diagnostic algorithms to enable CAD-based diagnosis of these lesions on the unique appearance of different types of tumors.

Highlights and Key Publications

UCLA CT & MR Score for Classification of Solid Renal Masses
Machine Learning-Based Quantitative CT Texture Analysis for Differentiation of Benign and Malignant Renal Masses
Using Aorta-Lesion Attenuation to Differentiate Between Malignant and Benign Renal Lesions
Validation of An Automated Software to Measure Total Kidney Volume in Patients with Autosomal Dominant Polycystic Kidney Disease
Long-term outcomes for primary and metastatic renal cell cancers

Overview

Kidney ICAMP includes BioBank, Radpath and CT Perfusion projects, all of which are seeking to establish high-resolution, accurate spatial registration between radiology and pathology. The development of Kidney ICAMP has expedited over the past year. Researchers and staff have defined the scope of data collection and annotations, implemented new database initiatives and streamlined existing workflows.

Data Available

Cohort	Total Cases	Cases Annotated	Cases w/Clinical Data	Cases w/Tissue	Cases w/Blood
Restrospective RCC Cases	1,449	289	1,449	-	-
Biobank CT/US Guided Biopsies (all consented)	213	164	213	199	97
CT Perfusions	109	-	109	-	-

ICAMP Liver

The goal of Liver ICAMP is to integrate imaging, pathology, biospecimens, and molecular profiling to advance precision-medicine approaches for liver malignancies and metabolic liver diseases.

Research Objectives

Our research objectives include developing MR, MRI elastography, US, Contrast US, CT, and other image-derived biomarkers that predict tumor progression risk in hepatocellular carcinoma (HCC), correlating annotated imaging with clinical and pathologic data to support individualized risk assessment. We aim to evaluate histopathologic, genomic, and molecular correlates of tumor biology and treatment response following locoregional liver therapies, including thermal ablation, irreversible electroporation, and histotripsy. We also aim to develop non-invasive, blood-based biomarkers to assess progression risk in metabolic dysfunction-associated steatotic liver disease (MASLD) and its advancement to MASH and HCC, and build a validated patient-derived 3D liver organoid platform from cryopreserved biopsy tissue to enable individualized drug-response screening.

Highlights and Key Publications

NIH/NCI U01CA230705 — UCLA Center for Early Detection of Liver Cancer (Zhou; Lu, co-I)
NIH P30 Cancer Center Support Grant — Liver Disease Biobank collaboration (Han, Agopian)
Multiple peer-reviewed publications, with additional abstracts, posters, and oral presentations at national meetings

Overview

Liver ICAMP integrates imaging, pathology, biospecimens, and molecular profiling in partnership with the Division of Hepatology & the Liver Cancer Center. The program collects imaging, biopsy-derived tissue, blood, and clinical data from patients with MASLD, MASH, hepatitis B/C, cirrhosis, and pre- and post-treatment liver malignancies, including hepatocellular carcinoma and cholangiocarcinoma. These specimens serve as a core resource for multimodal research to identify novel biomarkers and other factors associated with disease onset, progression, and outcomes, ultimately contributing to improved diagnostics, personalized therapies, and enhanced patient care.