In DIP flap reconstruction, a flap of abdominal tissue is taken from the patient to reconstruct

the breast after a mastectomy.

For the tissue to be connected to the circulatory system again, it is vital to prevent damaging

the eponymous perforator vessels when taking out the flap.

The vessel's course is therefore analyzed in advance to determine special landmark points,

for example, the location where a perforator crosses the abdominal muscles.

This 3D point is normally found by inspecting the stack of preoperative CT images on the

computer.

The estimated location on the skin is afterwards annotated on the patient according to visual

judgment.

To improve on accuracy in this procedure, we project a three-dimensional volumetric rendering

of the preoperative data directly onto the patient using a standard projector and provide

the physician with a method to intuitively choose the 3D point of interest using an HTC

Vive controller.

The viewer is tracked and the projection is updated each frame to create the illusion

of a semi-transparent 3D object.

Using the trackpad on the controller, the transfer function can be changed to see different

types of tissue.

The selection of a 3D point is realized in two stages.

First, a plane is selected using the controller.

Second, a point on this plane is chosen in a laser pointer fashion.

There are three plane modes with different advantages.

The semi-transparent mode, the clip plane mode and the MPR mode.

They are toggled by clicking the trackpad on the controller.

The semi-transparent plane is best for the overall understanding of depth as it allows

the user to see both the plane and the dataset at the same time.

In clip plane mode, the part of the dataset which lies in front of the plane is clipped,

providing the user a better view of the selected structures.

The last mode shows the multiplanar reconstruction along the plane to bridge the gap between

the volumetric view and the underlying CT data.

The plane is moved by moving the controller.

By pressing the trigger button on the controller, the user locks the plane and employs the controller

as a laser pointer to select the desired point on the locked plane.

Once this point is confirmed by pressing the trigger button once again, both the direct

path to the skin and the corresponding point on the surface are visualized.