Real-Time 3D Ultrasound Image-Guidance for Breast Surgery

Institution: Stanford University
Investigator(s): Michael Bax, M.S. -
Award Cycle: 2006 (Cycle 12) Grant #: 12GB-0148 Award: $66,141
Award Type: Dissertation Award
Research Priorities
Detection, Prognosis and Treatment>Innovative Treatment Modalities: search for a cure



Initial Award Abstract (2006)

During breast cancer surgery the primary goal of the surgeon is to remove the entire cancerous tumor as well as a small margin of surrounding healthy tissue to ensure that no cancerous cells are left behind. If a sufficient margin is not removed during the initial surgery, there is a substantially increased chance of recurrence, resulting in additional surgeries and frequently the need to remove the entire breast. In general, the surgeon operates without any image feedback during breast cancer surgery. The purpose of this research is to provide surgeons with an ultrasound-based, three-dimensional (3D) visualization and navigation tool that can be used prior to, during, and after the procedure to ensure successful removal of the cancer. Such an advanced imaging system will increase the likelihood of success by assisting with the surgeons' primary tasks: (1) locating the target tumor prior to the operation, (2) planning and executing an intervention along a minimally-invasive path, (3) removal of the cancerous tissue, and (4) evaluation of the region after surgery to ensure that no cancerous tissue remains.

The first stage of the project will consist of system development. We have already developed a prototype freehand 3D ultrasound system that uses a readily-available commercial 2D ultrasound system for image acquisition, an optical tracking system for probe and surgical instrument tracking, and a personal computer for 3D image reconstruction and visualization in real time. The system will first be prepared for use in the operating room. A collection of preliminary 3D images will then be acquired prior to, during, and after a series of routine cancer removal procedures. In order to be useful during the surgery, the system must be straightforward and easy-to-use. A significant portion of this project will consist of developing an intuitive and useful method of presenting the 3D image of the tumor and surrounding regions to the surgeon during the procedure, as well as allowing the surgeon to easily interact with the data. The preliminary images will be used to develop this intuitive method for image presentation and interaction. The second stage of the project will be to validate the system in the operating room.

The development of an innovative 3D visualization and navigation tool for use before, during, and after the procedure will provide surgeons with a new tool to provide additional insight into the physical configuration of the surgical site and the requirements of the procedure, in turn enabling more effective surgery with fewer negative side effects.




Final Report (2009)

Note: this project was extended by one year to allow the completion of the aims.

The goal of this project is to improve the results of surgery by developing a system which can help the surgeon use three-dimensional ultrasound to observe the area under operation and to plan the procedure. Specific tasks are to help locate a tumor, to enable a minimally-invasive operation, to remove all the malignant tissue, and to measure whether that the operation was successful. In particular it is expected to assist by presenting the patientís anatomy in 3D instead of a set of 2D slices.

An improved device was developed to calibrate the raw ultrasound probe data, using a pre-designed layout of a fixed geometry, including helper points to facilitate location of the device with a 3D tracking system. Software was written to automatically process the images from the calibration device so that manual intervention could be limited to positioning the ultrasound probe and initiating calibration.

The many variable options for the software processing the incoming stream of digital ultrasound images were tuned for optimal performance so that the ultrasound data is handled at maximum speed yet with maximum resolvable detail. The tracking subsystem was extended to allow the synthesis of continuous position and orientation tracking data from the individual snapshots provided, so that an accurate reading could be obtained even when the ultrasound and tracking snapshots were not made simultaneously.

A display system was designed for 3D display of the ultrasound data using the internal graphics hardware to substantially speed up display, and the performance was shown to be faster than the required threshold while freeing the computer for other tasks. Integrating this with the display of clinical tools used in the target area and displaying 3D and cross-sectional images was evaluated and shown to be practical at full speed. A variety of performance characteristics were evaluated and found to be within the required ranges, and exercising the system allowed the identification and correction of outdated assumptions regarding system operation.

Instead of going immediately to clinical testing as was originally planned, it was decided to add a hands-free interface to accommodate clinical constraints. A subsystem that can track where on the screen the operator is looking was obtained, customized, and improved to suit this system, and an appropriate screen interface was designed for its use using large buttons in one corner of the screen.

Future steps will be to evaluate the system in the clinical environment, taking feedback from these tests to further refine and improve the system for practical use




Symposium Abstract (2007)

The primary goal of breast cancer surgery is to remove all of a cancerous tumor together with a sufficient margin of surrounding healthy tissue to ensure that no cancerous cells remain. The likelihood of recurrence if a sufficient margin is not removed is significantly higher; this may result in additional surgeries and often removal of the whole breast.

Surgeons generally operate without any image feedback during breast cancer surgery. The goal of this research is to provide these surgeons with an ultrasound-based 3D visualization and surgical navigation tool for use before, during, and after surgery to ensure success. Such an advanced imaging system will improve the likelihood of success by assisting in the surgeonís primary tasks: (1) locating the tumor before to the operation; (2) planning and executing an operation using a minimally-invasive route; (3) removal of the tumor; and (4) evaluation after surgery to verify that no cancerous tissue was left behind.

This project will begin with system development. We have a prototype freehand 3D ultrasound system using a commercially-available 2D ultrasound system to acquire images, an optical tracking system to track the ultrasound probe and surgical instruments, and a computer workstation to reconstruct and visualize the 3D image in real time. First the system will be further developed for operating room use, then preliminary 3D data will obtained before, during, and after a series of routine cancer removal surgeries. The system must be straightforward and easy-to-use to be useful during surgery. A substantial element of this research will be the development of a useful, intuitive approach to displaying 3D imagery of the tumor and its surroundings to the surgeon during the operation, while giving the surgeon a simple interface for interacting with the system. The preliminary data will be used to develop the image presentation and interaction components. The complete system will then be validated the operating room.

The development of an innovative 3D visualization and navigation tool for use before, during, and after a procedure will give surgeons a new tool to provide additional insight into the physical configuration of the surgical site and the requirements of the procedure, in turn enabling more effective surgery with fewer negative side effects.