Project and Research Descriptions
Scientific & Medical Image Processing

Project and Research Descriptions: | Applied Optics | Microscope | Telescope | Scanning Mechanisms | Target Acquisition |
| Acoustics | Luggage Scanning | Feedback Controls | Image Processing | 3D Virtualization | Holographic Microscope |

I have devised a cancerous and diseased tissue tracking system design using ultra-sound or x-ray beams, laser alignment, remote cameras, and miniature MEMS (Micro-Electronic-Mechanical Systems) IMUs (Inertial Measuring Units).

The proposed method for medical tracking & pointing of diseased and cancerous tissue has the following features:

1. Tracking of lesions and cancerous tissue in the presence of patient motion
2. Pointing treatment with crossing ultra-sound beams or a scanning x-ray beam

The configuration for imaging and tracking a lesion or cancerous ROI (Region-of-Interest) elliptical volume or ellipsoid is shown in the figure below.

The treatment region is at the intersection of two ultrasound beams given a known statistical miss distance error. The miss distance is smaller than the size of the cancerous tissue. Crossing beam intersection at the region of beam overlap raises the temperature at a specific location within the prostate gland for treatment. Pointing a high energy x-ray and delivering a treatment dose to the precise region of interest is also possible through accurate coordinate frame referencing and laser tracking across the patient. Coordinate frame referencing error is shown as an "Additional Equipment Coordinate Frame" which has positional offset and orientation bias. The inertial sensors on the ultrasound probes provide attitude or rotation data and position data depending on the number of degrees-of-freedom of the IMU (Inertial Measuring Unit). Gyro angular rate IMUs provide 3-DOF (Degrees-of-Freedom) attitude determination using quaternion based rate integration algorithms. The 6-DOF IMUs provide both attitude and position measurement.

The LOS (Line-of-Sight) pointing direction error is based on the coordinate frame alignment of the equipment to the reference coordinate frame and the calibrated ultrasound treatment beam direction with respect to the ultrasound imager or emitter body. The LOS direction of the emitted treatment beam is calibrated to the body frame of the passive ultrasound sensors. If the gyro IMUs are 3-DOF then two fixed cameras are used to correlate images and track the motion of the two ultrasound transceivers. The two fixed cameras provide 3D position measurement when moving the ultrasound transceivers. If the gyro IMUs are 6-DOF then the fixed cameras are not necessary. The prostate moves with patient breathing which can be compensated with tracking from ultra-sound imaging or x-ray imaging and the proposed inertrial measurement method.

• Developed an ultra-sound holographic scanning device to detect cancer cells using a spatially coherent beam interference method to determine acoustic impedance and wave propagation velocity instead of traditional temporal methods (see ACHM (Acoustic Confocal Holography Microscope) publications)
• Derived a narrow viewing angle 3D tomographic reconstruction algorithm for the detection of cancer cells and diagnostic measurements
• Developed pattern recognition tracking algorithms for laser probes in the presence of patient movement
• Derived unique holographic image processing algorithms for OCT (Optical Coherence Tomography) retinal scanning microscopes
• Target signature conversion from infrared to hardbody and aimpoint determination (AIAA TMD Conference)
• Medical image processing as applied to Optical Coherence Tomography (OCT) for retinal scanning (proposal to NSERC)
• OCT interferometer images are processed with specially developed digital filters that feature extended spatial frequency bandwidth for greater sharpness and resolution
• Tomography algorithms for CAT scanning and ultra-sound imaging given a limited viewing angle with no rotational scanning as developed for the CSLH microscope
• Optimization of the digital filter bilinear z-transform for higher spatial frequency bandwidth out to the Nyquist limit
• Data mirroring, data windowing, and frequency domain spectral filtering using the Fourier transform
• Data windowing at low sampling resolution and high spatial bandwidth
• Finite difference solutions and state-space integration methods to filter function differential equations
• State-space digital filters for medical image processing