Skip to Content

Programs & Resources

The University of Iowa

Quantitative Imaging to Assess Response in Cancer Therapy Trials
John M. Buatti, (john-buatti@uiowa.edu)
U01-CA140206
University of Iowa

The University of lowa has assembled a strong multidisciplinary team of oncologists, clinical imaging specialists, computer scientists, imaging and therapy physicists, bioinformatics specialists, and statisticians, as well as collaboration with NIST and a public/private partnership with Siemens as part of the Quantitative Imaging Network. They plan to achieve the goals through four broad specific aims with subaim. They plan to create an accessible environment for quantitative image analysis tool development and testing using an existing large H&N cancer dataset and caBIG. They plan to design and deploy broadly connected informatics system architecture to support a locally networked relational database for development, implementation and validation of shared tools to the QIN through caBIG and the National Cancer Imaging Archive (NCIA). They will Interface their large PET/CT imaging and clinical outcomes database of pre and post therapy H&N cancer patients as a universally accessible test database that will be shared with the QIN. In the second aim, they will develop new semi-automated tools for quantitative image-based response assessment. They will design, implement, and validate open source 2D and 3D semiautomated quantitative image analysis tools for tumor definition and quantitative metrics applicable to response assessment. They will also develop and apply advanced decision-support software as well as measurement variability tools so that image analysis tools can be compared and optimally selected for clinical cancer trials. They will test these image analysis tools on the imaging and clinical data to assess applicability and accessibility by QIN. In a third aim, they will establish robust quality procedures and standards for quantitative imaging, working with NIST and the QIN to enhance the quality and reliability of quantitative imaging for clinical decision-making. They will define and validate calibration procedures, tools and standards for PET imaging using a Ge-68 NIST-traceable phantom. They will define and validate robust quality assurance tools and standards for 4D PET/CT. They will define similar tools and standards for quantitative imaging response assessment using ACR and GE phantoms for MRI, including MRSI. In the fourth broad aim, they will adapt and enhance quantitative image-based response assessment for clinical trials decision-support: In an existing, NCI-funded study of FDG and F-18 fuorothymidine in head and neck cancer patients. In an existing, NCI-funded study of FDG and C-11 acetate imaging in lung cancer, and in a planned clinical trial of MRSI and F-18 fluorodopa in high-grade gliomas.

This work can positively impact public heath by determining the effectiveness of cancer treatment more objectively as well as earlier after treatment using detailed analysis of imaging. This may enable better decision making for cancer treatments. Results will be shared so that work among groups will help apply these new imaging approaches more effectively.

The Specific Aims of the program will include:

Aim 1. Create an openly accessible environment for quantitative image analysis tool development and testing using an existing large head and neck cancer (H&N) dataset and caBIG infrastructure.
Aim 2. Develop novel semi-automated tools for reproducible tumor definitions applicable to quantitative image-based response assessment that will be compared with manual methods.
Aim 3. Establish robust quality assurance procedures and standards for quantitative imaging working with NIST and the QIN to enhance the quality and reliability of quantitative imaging for clinical decision-making.
Aim 4. Adapt, enhance and extend quantitative image-based response assessment for clinical trials decision-support.