Tak's Home Page

Takashi Buma

Assistant Professor

Department of Electrical and Computer Engineering

104 Evans Hall

Newark, DE 19716-3130

Teaching

· CPEG 202 - Introduction to Digital Systems (S08, S09, S10, S11)

· ELEG 240 - Physical Electronics (First half) (S08)

· ELEG 479/679 - Introduction to Medical Imaging Systems (F06, F07, S09, S10, S11)

· ELEG 661 - Nanoelectronics, Electromagnetics, & Photonics Seminar Schedule (S07)

· ELEG 664 - Bioengineering Seminar Schedule (S07, F08, F10)

· ELEG 823 - Ultrafast Optics (F05, S07, F10)

Optics and Ultrasonics Research Laboratory

Research Areas:

1) Ultrasound Biomicroscopy (UBM)

UBM employs high frequency ultrasound to produce high resolution images of tissue microstructure. However, the lack of suitable sensor arrays has prevented UBM from making the leap to widespread clinical use. We are developing UBM systems based on “optoacoustic” sensor technology. The basic idea is to detect ultrasound with optical interferometric techniques instead of conventional piezoelectric technology. This project has two major research aims: (1) develop broadband, large aperture, and highly populated optoacoustic sensor arrays (2) integrate these optoacoustic arrays into real-time UBM imaging systems. (Current funding: NSF)

We are also developing a fiber optic-based hydrophone capable of characterizing transducers beyond 100 MHz. Considering that ultrasound imaging involves transmitting and receiving ultrasound pulses, accurately mapping the transmitted ultrasound field is clearly an important measurement. Unfortunately, such basic information is difficult to obtain due to the lack of suitable point-like receivers operating at frequencies higher than 40 MHz. No existing hydrophone combines large bandwidth, sufficient sensitivity, fine spatial resolution, and ease of use. (Past funding: NIH)

Ultrasound field measured by a 256 element optoacoustic array

Hydrophone based on a fiber-optic Sagnac interferometer

2) Terahertz Time-Domain Imaging (THz-TDI)

Imaging with electromagnetic terahertz (THz) pulses has received considerable interest for applications in security, package inspection, and nondestructive testing. Two-dimensional (2-D) phased arrays of transmitter and receiver elements can produce fully focused volumetric THz images over a large field of view. We are exploring adaptive reconstruction methods to improve the three-dimensional imaging capabilities of THz arrays.

Conventional THz time-domain imaging systems scan a focused THz beam across the region of interest. Finest spatial resolution requires sharp focusing, but this produces an extremely short depth-of-focus. Consequently, 3-D imaging requires scanning over all three dimensions, which is extremely time-consuming. We are exploring a virtual transceiver approach with an adaptive reconstruction algorithm to overcome this limitation. A 2-D synthetic aperture array is produced by laterally scanning a virtual transceiver over the object of interest. Images are formed by synthetic aperture reconstruction, where the focusing quality is improved with an adaptive weighting factor based on the spatial coherence of the recorded signals. (Past funding: UDRF)

Sparse arrays use widely separated elements to reduce array complexity. However, their image quality is severely degraded by grating lobe artifacts. This is because propagation path differences between neighboring elements can exceed the THz pulse duration. The non-interfering pulses form ripple-shaped artifacts (grating lobes) in the image. We demonstrate artifact suppression by introducing an “interference factor” that adaptively measures the degree of interference between signals of nearest neighbor elements. This adaptive factor is close to one when nearest neighbor signals overlap (e.g. at the array focus) but is nearly zero when there is little overlap (e.g. far from the array focus).

Time-domain terahertz imaging system

THz images with conventional and adaptive reconstruction

Conventional and adaptive sparse THz array images

3) Photoacoustic Microscopy (PAM)

PAM provides high resolution images with excellent image contrast based on optical absorption. A laser pulse illuminates tissue, where optically absorbing regions emit ultrasound by thermoelastic expansion. The detected ultrasound waves are processed to reconstruct the location of the optically absorbing regions. We are developing a high repetition rate and tunable optical source that is suitable for spectroscopic PAM. Pulses from a high repetition rate Q-switched Nd:YAG microchip laser are sent through a photonic crystal fiber. Highly nonlinear optical propagation produces a supercontinuum spectrum spanning 500 to1300 nm. A tunable band pass filter selects the desired wavelength band from the supercontinuum. Our PAM system employs optical focusing and a 25 MHz spherically focused detection transducer. (Past funding: UD Undergraduate Research Program)

Photoacoustic microscopy system employing optical focusing

Image of USAF target shows a resolution of 18 um

Multispectral processing distinguishes the different spots

4) Optical Coherence Tomography (OCT)

OCT is an emerging technique for high resolution biomedical imaging. Advantages include fine spatial resolution (better than 10 um laterally and 7 um axially), portability, cost effectiveness, and miniaturization into arthroscopic devices for minimally invasive imaging. Unlike histology, OCT can produce thinly sliced cross-sectional images without physically cutting tissue. OCT typically has a penetration depth of about 1 mm in tissue, which is limited by optical scattering. We have developed a spectral domain OCT (SD-OCT) system with the eventual goal of integrating it with our PAM system. This multimodality system will combine the structural imaging capabilities of OCT with the functional imaging capabilities of PAM. (Past funding: UDRF)