CAS at University of Basel

Fluoroscopy is an imaging technique commonly used by physicians to obtain real time images of the internal structures of a patient through the use of a fluoroscope. It is used to visualise:

• organ motion
• ingested or injected contrast agent
• catherised interventions (inserting stents, RF blation,...)
• Computer-assisted Surgery
  

CAS supports the surgeon in performing surgicasl intervention by providing additional visual, ... feedback about the surgical procedure.

• Imaging Modality (CT, Fluo, US, MR)
• Navigation Support (AR, Screen, ...)
• Tracking Modality (Optical, magnetic,..)
• 
  

• Articulated Arm technology
• Acoustic Tracking
• Optical Tracking
• Electromagnetic Tracking
• Fibre Shape Sensing
• Inertial Measurement Unit
• Machine Vision

Historically the first navigation support.

Calibrated angle and length of segments helps to determine the position of the end tip of the arm's probe.

It can only track one object, so the patient had to be rigidly fixed.

Not used alot anymore, as the cons outweigh the pros

• Detects infrared-emitting or retro-reflective markers in space
• Sensor calculates the position and orientation of the tool based on the information the position sensor receives from these markers
• Two types of markers: Active and passive

Basically a Gyroscope, accelerometer, and magnetometer integrated into one sensor.

Accelerometer: Mass spring system, measures acceleration

Gyroscope: measures movement/rotation

Magnetometer: Measures the direction of the earth magnetic field

IMU require initializing (alignment) at a known point in order to provide a reference from which the system measures.

Machine vision uses a computer that attempts to directly understand the content of images sent to it by means of a video camera. Basically tries to measure the environment. Simplified as it is a hard task.

Triangulation to determine position, and multiple markers to determine orientation.

Accuracy is dependent on the target size and baseline of the triangulation.

In medical applications, the environment e.g room lighting can be difficult to control and may reduce the effectiveness of this technology.

A projectivity is an invertible mapping h from P^2 to itself such that three points x1, x2 and x3 lie on the same line if and if only h(x1), h(x2 and h(x3) are on one line too.

-> Collineation / projective transformation / homography

-> Projectivities from a group called the projective linear group

A mapping h: P^2 -> P^2 is a projectivity if and only if there exists a non singular 3x3 matrix H such that for any point in P^2 represented by a vector x it is true that h(x) = Hx

Exactly at the corner, the gradient is ill-defined. However, a corner is normally made of two edges, so in the region around the corner, the gradient has two or more values. This means intensity changes in all directions.

In dense matching no real key points are determined. Instead, a dense grid of sample points is taken. Of each point, the local descriptor is calculated and we try to match them. This approach might be fast, but the problem is, that there will also be many points with similar local descriptors.

Exhaustive search is a multi-scale procedure: we compare the descriptors while varying the patch size. It is computationally inefficient but possible for matching.

Prohibitive for retrieval in large databases and recognition.