CAFC Rejects Patent Claim Construction That Excludes All Embodiments of Invention
The CAFC relied on a technical dictionary and the patent specification to conclude that the Central District of California erred in adopting a claim construction which effectively excluded all of the described embodiments. The CAFC also rejected the district court’s addition of a limitation that was not present in the claim, specification, or prosecution history. As a result, the CAFC vacated the district court’s grant of summary judgment which had found no infringement of Nellcor’s patent for pulse oximeters by Masimo’s competing product. See Nellcor Puritan Bennett, Inc. v. Masimo Corp., No. 04-1247 (Fed Cir. April 08, 2005).
A pulse oximeter measures blood oxygen and pulse rate. A familiar form is a device clipped to a patient’s fingertip which beams both red and infrared light through the fingertip, and measures the difference in the absorption of the two beams. The absorption rates of red and infrared light are dependent on the amount of oxygen in the blood, and the amount of blood in the fingertip. Thus, differences in the amount of light measured at the red and infrared wavelength can be used to calculate blood oxygen and pulse rate.
The CAFC first considered whether the phrase “attenuated and filtered” meant “reduced and removed,” as held by the district court, or whether it meant simply reducing in comparison to the desired periodic signals.
To decide, the CAFC used a standard technical dictionary, which had 8 definitions for “filter.” The definitions included a device that “separates” data or signals, and a device that “eliminates” certain portions of a signal, both of which were consistent with Nellcor’s argued claim construction. The court then looked to the specification to determine the meaning intended by the patentee. The specification described a method for collecting data and then effectively filtering out the aperiodic signal waveforms through a relative reduction of aperiodic data (not its total elimination). This was consistent with the most detailed description of the invention, provided in the Summary of the Invention portion of the specification. In fact, the disclosed process did not actually remove data, but suppressed aperiodic noise relative to the periodic signal. Based on this, the CAFC concluded that “attenuate and filter” meant effective removal, not absolute removal, of unwanted data.
A further factor in Nellcor’s favor was that a construction that defined “filter” to mean the absolute rather than relative removal of aperiodic noise would eliminate all of the embodiments described in the specifications, because none of them actually removed the aperiodic noise. As found in Vitronics Corp. v. Conceptronics, Inc., 90 F.3d 1576, 1583 (Fed. Cir. 1996), a construction that excludes all the embodiments is “rarely, if ever, correct.”
The district court had relied heavily on the prosecution history, in which Nellcor had distinguished its method from prior art by describing it as one in which “aperiodic data was removed.” That language led the district court to conclude that Nellcor’s method completely removed the aperiodic noise. In contrast, the CAFC took that language in context, and concluded that it did not support the district court’s interpretation.
The CAFC also found the district court erred in holding that the phrase “calculating . . . from the relative maximum and minimum” required that any calculation using the relative minimum must be made only after formation of a composite signal (accomplished through use of a Fourier transformation). The relative minimum components of the signal, or zero frequency components, of the signal were used to normalize the relative maximum values of the infrared and red signals – that calculation produced the same result regardless of whether performed before or after the Fourier transformation was applied to the data. In effect, the district court improperly added a limitation that was not in the claim language, nor supported by the specification or prosecution history.
While expressing no opinion on the ultimate issue of infringement, on remand the CAFC ordered the district court to adopt its construction with respect to the critical claim terms “attenuated and filtered” and “relative minimum” value.
A pulse oximeter measures blood oxygen and pulse rate. A familiar form is a device clipped to a patient’s fingertip which beams both red and infrared light through the fingertip, and measures the difference in the absorption of the two beams. The absorption rates of red and infrared light are dependent on the amount of oxygen in the blood, and the amount of blood in the fingertip. Thus, differences in the amount of light measured at the red and infrared wavelength can be used to calculate blood oxygen and pulse rate.
The CAFC first considered whether the phrase “attenuated and filtered” meant “reduced and removed,” as held by the district court, or whether it meant simply reducing in comparison to the desired periodic signals.
To decide, the CAFC used a standard technical dictionary, which had 8 definitions for “filter.” The definitions included a device that “separates” data or signals, and a device that “eliminates” certain portions of a signal, both of which were consistent with Nellcor’s argued claim construction. The court then looked to the specification to determine the meaning intended by the patentee. The specification described a method for collecting data and then effectively filtering out the aperiodic signal waveforms through a relative reduction of aperiodic data (not its total elimination). This was consistent with the most detailed description of the invention, provided in the Summary of the Invention portion of the specification. In fact, the disclosed process did not actually remove data, but suppressed aperiodic noise relative to the periodic signal. Based on this, the CAFC concluded that “attenuate and filter” meant effective removal, not absolute removal, of unwanted data.
A further factor in Nellcor’s favor was that a construction that defined “filter” to mean the absolute rather than relative removal of aperiodic noise would eliminate all of the embodiments described in the specifications, because none of them actually removed the aperiodic noise. As found in Vitronics Corp. v. Conceptronics, Inc., 90 F.3d 1576, 1583 (Fed. Cir. 1996), a construction that excludes all the embodiments is “rarely, if ever, correct.”
The district court had relied heavily on the prosecution history, in which Nellcor had distinguished its method from prior art by describing it as one in which “aperiodic data was removed.” That language led the district court to conclude that Nellcor’s method completely removed the aperiodic noise. In contrast, the CAFC took that language in context, and concluded that it did not support the district court’s interpretation.
The CAFC also found the district court erred in holding that the phrase “calculating . . . from the relative maximum and minimum” required that any calculation using the relative minimum must be made only after formation of a composite signal (accomplished through use of a Fourier transformation). The relative minimum components of the signal, or zero frequency components, of the signal were used to normalize the relative maximum values of the infrared and red signals – that calculation produced the same result regardless of whether performed before or after the Fourier transformation was applied to the data. In effect, the district court improperly added a limitation that was not in the claim language, nor supported by the specification or prosecution history.
While expressing no opinion on the ultimate issue of infringement, on remand the CAFC ordered the district court to adopt its construction with respect to the critical claim terms “attenuated and filtered” and “relative minimum” value.
posted by at 4/11/2005 09:19:00 AM
0 Comments:
Post a Comment
<< Home