SBIR-STTR Award

The proposed research effort is aimed at advancing the state-of-the-art in high-frequency (200 MHz), broadband ultrasound to provide images of malignant tumor cells and surrounding tissue. The principal types of cancer that can readily be diagnosed include oral cancer and cancer of the oropharynx; esophageal cancer and other tumors of the gastrointestinal tract (endoscopic ultrasound); skin cancer, in particular melanomas; and cancers originating within the eye (mainly intraocular melanomas). As ultrasound frequencies increase, one must adopt new transducer construction techniques to accommodate the very small spatial dimensions involved in the arrays. The microfabrication techniques proposed give rise to low-cost transducer arrays because many (100 or more) can be grown on a single silicon wafer. The baseline Phase 1 design entails the construction of two different types of lead-zirconate-titanate (PZT) transducer arrays on silicon wafers. The silicon wafer is used as one of the ultrasound backing layers. A modified sol-gel process is used to deposit a PZT thick film. In Phase I, a 32-element linear array and a 4x4 two-dimensional array will be constructed and tested to determine the feasibility of the proposed fabrication process. As part of Phase I, we will also determine the lowest transducer frequency that can take advantage of our low-cost microfabrication techniques. In one fabrication scenario, it is possible to operate the ultrasound transducer non-simultaneously at two different frequencies that have a ratio of 2 (e.g., 100 MHz and 200 MHz). In Phase 2, a full-size one-dimensional phased array with an elevation lens will be constructed as well as a two-dimensional phased array. Proper sizing of the aperture will yield lateral spatial resolutions of the order of 10-15 jim. This combined with 6 .tm axial resolution will allow us to resolve gross cellular structure; The shrinking of the transducer array to dimensions of roughly I mm x 1 mm allows the sensor to be easily placed at the tip of an endoscope. The commercial products that will emerge from the proposed high-resolution ultrasound system include production-quality transducer assemblies, beam steering software for field programmable gate arrays, signal processing software for DSP hardware, and a comprehensive system design that can be licensed

The proposed effort focuses on the final development of a high-frequency (100 MHz), broadband ultrasound microsystem designed to image cellular structure/tissue along the gastrointestinal (GI) tract. The unit fits within a standard GI endoscope. It consists of a paraboloid transmitter, a 16 x 16 ultrasound receiver array, a read-out integrated circuit (ROIC), and an interposer layer that links the latter two items. Fourteen-bit digitized data flow from the ROIC to an external computer where three-dimensional images are formed. The nominal sampled volume extends below the GI tract surface to depths in the range 1-3 mm, depending on the proximity of the sensor to the GI wall. The diameter of the probed volume ranges from 0.6 mm to 0.9 mm. The main functions of the imager include: 1) the imaging of pre-cancerous dysplastic mucosa, polyps, and adenomas, 2) real-time grading of dysplasia (pre-cancerous tissue), 3) immediate viewing of cellular structure in tumors (e.g., squamous cell carcinoma and adenocarcinoma, benign growths), 4) as required, guidance for directing fine-needle aspiration biopsies to regions that pose the greatest threat, and 5) the system serves as a patient friendly, pre-cancer diagnostic for severe gastroesophageal reflux disease (Barrett's esophagus). Overall, the suggested imager provides a new opportunity for the detection of pre-cancerous tissue along the GI tract, which at present is usually a chance finding and cannot be recognized endoscopically. The developmental prototype has a novel bistatic architecture aimed at increasing system sensitivity, reducing the time required to obtain an image, and simplifying the receiver system. A new sol-gel PZT (PbZr0.6Ti0.4O3) thick film, invented as part of the preceding NIH program, is used as the transducer material. It is much improved over previous efforts in this area. The PZT thick film density is 89-90% of the full thin film density and the thick film piezoelectric coefficients are very close to those of thin films. The advantage of the new thick film is 1) its patterning and fabrication process is well established and many methods are available, 2) it supports a very large etch selectivity of 15:1, 3) the fabrication/assembly process is straightforward and amenable to large-scale production, and 4) the fabrication process yields very inexpensive transducer arrays. The latter is very important because the transducer array/interposer/ROIC will be used only once and discarded