Lexion Medical, LLC v. Northgate Technologies, Inc., 2009-1494 (Fed. Cir. Apr. 22, 2011).

Writing for the court, Chief Judge Rader handed patent prosecutors a checklist for getting the most out of a numerical range.  First, “[t]his court prefers a claim interpretation that harmonizes the various elements of the claim to define a workable invention.”  A claim term that inherently provides flexibility in the endpoints of the range supports a broader interpretation.  Second, when the specification is consulted, it should support that flexibility. This one “clarified” the endpoint in a process range as “about 2°C” and described the temperature fluctuating outside the range in the method.  The court shanghaied “about” from the specification to modify an unqualified endpoint.  As a result, the defendant literally infringed the claim even with a temperature that was not always within the specified range.

The patent claimed a method for introducing heated, humidified gas “having a temperature within 2°C of the predetermined temperature” into a patient’s body.  Infringement depended on whether the temperature could “at times, fluctuate outside the four-degree range.”  In the method, the temperature of the gas as it leaves a chamber with heating means is sensed to see if it is in a predetermined range; if not, the heating means is actuated.  This language shows that the temperature can fluctuate outside the range because the method provides for bringing it back to the range.  The specification also informs one that the method allows the temperature to fluctuate outside the range by describing such fluctuations in normal start-up and use, by modifying the numerical value with “about,” and by noting a lag time to achieve the target temperature range during the method.  Therefore, the defendant’s device could provide gas at temperatures sometimes outside the “within 2°C” range and still literally infringe.

Lexion calls to mind the finding of direct infringement in Cohesive Technologies v. Waters Corporation, 543 F.3d 1351 (Fed. Cir. 2008), in which chromatography columns packed with particles having measured average diameter of either 29.01 μm (plaintiff’s measurement) or 25.16 μm (defendant’s measurement) were in the literal scope of a claim reciting “particles having average diameters of greater than about 30 μm.”  The court took two key disclosures into account.  First, the court looked to the purpose for having the low-end cut-off of about 30 μm, which was to generate sufficient turbulence in liquid flow through the columns for them to function.  Specification examples showed particles of “nominal average diameters” of 50 μm worked well while those of 20 μm did not.  Second, the specification disclosed that a batch of particles with a nominal diameter of 50 μm had an actual measured mean diameter of 42.39 μm, a variance of 15.22%.  The claim term did not specify if the diameter was a nominal diameter or a measured diameter, but the court interpreted “about 30 μm” to encompass a 15.22% variation, from 25.434 to 34.566 μm.  The specification example of the non-working, nominal 20 μm columns encompassed 16.956 μm to 23.044 μm.  In between, where the defendant’s measured diameter of 25.16 μm lay, the court held the claim literally covered the diameter if the column worked.  (

The court concluded that Cohesive Technologies could not have resorted to the doctrine of equivalents, as particle sizes that worked in the same way with the same results as the 30 μm particles of the claim were already in the literal scope of the claim—but at that point, why would Cohesive have cared:  It was only being denied columns with particle sizes its own comparative testing showed would not work.

Lexion and Cohesive teach us to describe the function a range serves and give examples.  In Lexion, the range of ±2° was a target for the process, not an absolute it could not violate.  In Cohesive, the average particle diameter was linked with its purpose of creating the necessary turbulence in flow through the column.  The examples in Lexion of the temperature fluctuating outside of the range and the comparative examples in Cohesive Technologies of average particle diameters that did and did not work supported liberal interpretations of the numerical ranges in their claims.

Two more notes.  First, despite what the Lexion court did, the presence of the modifier “about” in the claim was also critical in Cohesive.  Second, the Cohesive court noted as dicta, “Although ‘greater than’ and ‘between’ 30 μm and a larger number would not encompass exactly 30 μm, while ‘not less than’ 30 μm would, this distinction is not relevant for our analysis in this case.”  Id. at 1358 n.1.  If Cohesive’s claim had been written as “particles having average diameters of greater than 30 μm” so that it did not encompass “exactly 30 μm,” the dicta would not have been mere dicta, and the whole claim scope below 30 μm would have been lost.