University of Washington

Cordis Corporation

Gregory L Zick

Hain-Ching H Liu

Stanley H Saulson

Florence H Sheehan

Lawrence Yapp

A technique to perform speech recognition directly from audio files compressed using the MPEG/Audio coding standard. The technique works in the compressed domain and does not require the MPEG/Audio file to be decompressed. Only the encoded subband signals are extracted and processed for training and recognition. The underlying speech recognition engine is based on the Hidden Markov model. The technique is applicable to layers I and II of MPEG/Audio and training under one layer can be used to recognize the other.

Speech recognition on MPEG/Audio encoded files

Analysis of video data in an MPEG compressed format to identify scene changes to facilitate indexing to access specific video frames and to improve data compression and/or quality of the compressed video. MPEG compressed video data include I-pictures, P-pictures, and B-pictures. Scene changes are identified by analyzing P-pictures using average error power, based upon direct cosine transform components in the compressed data, and/or using an interframe correlation between a P-picture and its past reference frame. When a scene change occurs, average error power becomes large and interframe correlation between frames becomes substantially zero. Alternatively (or in addition), B-pictures in a video slot can be analyzed, in terms of interframe correlation between past and future reference frames, using either an AND analysis technique or an OR analysis technique. The AND analysis identifies a scene change when two or more B-pictures between two reference frames defining the video slot all imply a scene change, while the OR analysis indicates that a scene change has not occurred in the video slot if one B-picture implies that a scene change has not occurred. Once the scene changes have been identified, all I-pictures not occurring at a scene change are converted to P-pictures, and all P-pictures immediately following a scene change are converted to I-pictures. This step improves compression and/or quality of the compressed video, since P-pictures require less data than do I-pictures, and fewer I-pictures will be used in the compressed data following this step.

Automatic scene decomposition and optimization of MPEG compressed video

A method and system for identifying end systole and end diastole frames within an angiography sequence. A plurality of images produced during an angiography sequence are digitized, producing digital image data in which gray scale values for each of the pixels in the images are represented. The digital image data are input to a computer (48) to determine the frames in which the coronary arteries are most visible. The coronary arteries are made visible in the images by injecting a radio-opaque contrast substance into the arteries. The frames that occur a end diastole are preferred for diagnostic analysis because the arteries are distended, spread apart from each other, and moving very slowly. To identify such frames for further analysis, the total length of edges within a centered window covering approximately one-fourth of each image is determined. The edges represent spatial transitions between relatively light and dark areas in the image that occur across the borders of the coronary arteries. At end diastole during the cardiac cycle, the total length of the coronary arteries within the window is substantially less than at end systole, when the heart has contracted to a minimum volume. Frames in which a local maximum total edge length is observed thus depict the arteries at end systole. Similarly, frames having a local minimum total edge length show the arteries at end diastole. Either a spatial method or a discrete cosine transform (DCT) method is used for determining the total edge length in each frame.

Automatic indexing of cine-angiograms

In the reference electrode construction disclosed herein, silver and silver chloride particles are incorporated in a mixture including a fusible glass frit. After being printed on a suitable substrate, the mixture is fired to bond the silver/silver chloride composite into a conductive film on the surface of the substrate.

Solid state reference electrode

Oxygen sensing as disclosed herein is performed by applying a solid state, self-heating sensor to the patient's skin. A noble metal cathode and a thick-film reference electrode are provided on the front face of a thin flat substrate employed in the sensor. A thick-film resistive heater is provided on the back surface of the substrate for heating the sensor to produce hyperemia in the adjacent tissue so that a polarographic measurement of oxygen concentration may be obtained by means of the electrodes on the front face.

Oxygen sensing

Disclosed is apparatus and method for continuously compensating for electrode drift during the measurement of the partial pressure of oxygen by the net charge transport technique. The apparatus derives a correction factor from variations in waveforms representing charge returned from an electrochemical cell after successive interrogating voltage pulses. These wave-forms are independent of oxygen partial pressure but dependent on electrode parameters thereby permitting drift to be monitored and oxygen partial pressure measurements to be corrected.

Growing community of inventors

Kirkland, WA, United States of America

Gregory L Zick