This Phase II proposal is a continuation of our successful Phase I grant whose overarching goal is to develop amore robust and accurate method for the deformable fusion of pre-operative magnetic resonance images (MRI)and intra-procedure trans-rectal ultrasound (TRUS) images to guide prostate biopsy. Prostate cancer is a majorcause of death in the U.S. and biopsy is the gold-standard for its diagnosis. Each year over 1.1 million biopsiesare performed in the U.S. alone at a cost of well over $1 billion. Most of these biopsies are performed under TRUSguidance in a systematic fashion but blinded to potential tumor locations because many tumors are invisible onultrasound. This procedure is highly inef?cient and as many as 30% of serious tumors may be missed on a ?rst-time biopsy. Multi-parametric MRI (mpMRI) acquired before the biopsy can be used to de?ne potential targetsand then used in conjunction with live TRUS imaging to perform targeted image-guided biopsy. A key challengeis the accurate registration (fusion) of the intra-procedure TRUS with the pre-acquired MRI. This fusion procedureis often performed using rigid registration which does not account for prostate deformation between the MRI andTRUS acquisitions. Eigen's Artemis system is one of a newer generation of devices that performs deformablemapping between MRI and TRUS. Current fusion methods (such as in Artemis), however, are highly operatordependent as they rely on semi-automated TRUS segmentation by the clinician during the procedure which iserror prone, time-consuming and critically reliant on operator skill. We propose to extend the development ofour strategy for deformable registration (from Phase I) based on a statistical deformation model learned froman existing, large, multi-institutional database (N=200) of MRI-TRUS prostate image pairs. We will learn thestatistics of both biopsy deformation and prostate shape clusters. The model will reduce the dimensionality ofthe search as well as constrain the deformation. The proposed registration method will use a regional con?denceestimate for the intra-procedure interactive TRUS segmentation in order to enhance robustness by focusing theregistration on the more con?dent portions of the bounding surface. We will also generate regional anisotropicestimates of registration uncertainty. These methods will streamline the clinician's work?ow by providing a moreaccurate and robust registration with less dependence on segmentation errors. This project is signi?cant in thatit has the potential to improve the utilization of MRI-TRUS fusion guided prostate biopsy (especially outside ofmajor academic hospitals) by alleviating the dependence of the work?ow on exact prostate segmentation fromoften poor quality TRUS images.
In Aim 1, we will further improve our fusion algorithm, integrate it into theArtemis system and test for improved accuracy.
In Aim 2, we will create an innovative research interface to ournew algorithm to allow for clinical testing and then perform a two-site clinical validation in N=40 patients. We willcompare our new method to the current non-rigid MRI-TRUS fusion method used in our FDA-cleared Artemissystem to demonstrate improved accuracy.
Prostate cancer is the second most common form of cancer (behind skin cancer) and biopsy is the clinical goldstandard for diagnosis. This grant is aimed at enhancing the accuracy and robustness of state-of-the-art MRI-ultrasound fusion-guided prostate biopsies with an expected improved detection of disease resulting in earlierdiagnosis and improved survival. At the core of the effort is the development and validation of novel imageanalysis methods for MRI-ultrasound registration using a statistical model of deformation and shape based onpopulation statistics of prostate deformation.
| Onofrey, John A; Staib, Lawrence H; Sarkar, Saradwata et al. (2017) Learning Non-rigid Deformations for Robust, Constrained Point-based Registration in Image-Guided MR-TRUS Prostate Intervention. Med Image Anal 39:29-43 |
