In re Kelkar — Half Right for the Wrong Reason?

In re Kelkar involves a patent application owned by the IBM Corporation that claims methods and computer program articles for processing gene expression profiles to find clusters of similar profiles.  The Board of Patent Appeals and Interferences found the following claim 1 of the application non-statutory as attempting to claim a prohibited abstract idea under Bilski v. Kappos, 561 U.S. ___ (2010):

 1.  A method for determining similarity between portions of gene expression profiles in a computer comprising the steps of:

processing a number of gene expression profiles with a similar sequences algorithm that is a time and intensity invariant correlation function to obtain a data set of gene expression profile pairs and a match fraction for each gene expression profile pair;

listing gene expression profile pairs in clusters by their match fractions;

removing a first gene expression profile from a cluster when another cluster has another gene expression profile with a higher match fraction with the first gene expression profile, unless the another gene expression profile requires a larger number of subsequences to achieve similarity with the first gene expression profile;

repeating the removing step until all gene expression profiles are listed in only one cluster;

providing output of the listing of clusters of gene expression profiles. (emphasis added)

The Board also found a Beauregard claim to a computer program article non-statutory as well, on the grounds that the claim was broad enough to encompass transitory signals, a category of subject matter found outside of Section 101 by the Federal Circuit in In re Nuijten (Fed. Cir. 2007).  This posting does not address that rejection, instead focusing on the rejection of claim 1.

The three key holdings/conclusions of the Board that allowed it to find claim 1 non-statutory under 35 U.S.C. Sec. 101, are as follows:

  • That “as in Flook, “[t]he application’s only innovation … [is its] reliance on a mathematical algorithm.” Bilski, 130 S.Ct. at 3230. We therefore conclude that claim 1 is directed to an abstract idea – the mathematical algorithm – as applied to a defined type of data.”   
  • That “[t]he method of claim 1 is not limited to a particular machine.  At best, the preamble identifies the environment in which the method is performed and represents nothing more than a field-of-use limitation that is insufficient to render the otherwise ineligible process of claim 1 patent eligible.”   Thus claim 1 fails the machine prong of the MOT test.
  • That claim 1 “does not recite a process that “transforms an article to a particular state or thing” and therefore fails the transformation prong of the machine or transformation test.”

In addition, the Board rejected the Appellants’ contention that the process of claim 1 “is limited to a particular practical application” and “does not convert or process just any data but is limited to useful concrete and non-abstract gene expression profiles in a data base of such profiles.”

I think the Board either got this decision wrong, or at best half right as discussed further below, for the following reasons:

  • The Board found that Claim 1 was limited to a computer implementation, and therefore the claim should have satisfied the machine prong of the MOT test.
  • Claim 1 is not merely limited to a “field-of-use”,  as was the case in Parker v. Flook, but rather to the very specific task of processing data representative of gene expression, which itself is representative of physical properties of biological material.  Accordingly, the invention is not “abstract” but clearly applied to the processing, in a machine, data representing biological properties of biological material.  Claim 1 does not merely calculate a “number” but rather produces “clusters of gene expression profiles” that are not mathematical in nature, but rather reorganized or correlated data representative of biological properties of biological material.
  • Claim 1 does transform “articles” to a different state or thing, wherein the articles in question are data items representative of biological properties of biological materials, and wherein the articles are transformed by various processing steps to a different organization.

Let’s start with my first point.  It is unquestionable that claim 1 is limited to implementation in a computer:  it expressly requires that the invention be performed “in a computer” in the preamble, and it also has a number of limitations that are consistent with computer implementation – “processing”, “data set”, “repeating .. step”, “providing output” and others.  If the Board’s only complaint is that the claim does not explicitly recite a computer implementation in the body of the claim, then this rejection will be easily avoided by bringing that limitation into the body of the claim.  However, the Board’s finding of fact seems to acknowledge that claim 1 is limited to being performed in a computer:  “The method of Appellants’ claim 1 is performed “in a computer.” (FF1), but in its reasoning appears to take the opposite tack:  “At best, the preamble identifies the environment in which the method is performed and represents nothing more than a field-of-use limitation that is insufficient to render the otherwise ineligible process of claim 1 patent eligible.” 

Less ambivalently, the Board also asserts that reciting a “computer” does not constitute a “particular machine”:  “The method of claim 1 is not limited to a particular machine (FF 1-2)”.  One must conclude therefore that the Board does not regard a computer as a “particular machine” for the purposes of the MOT test.  This later rejection seems to be gaining in popularity, and one worth taking a closer look at.  First off, if this rejection means to take the position that limiting a claim to a computer implementation, without more, potentially encompasses more than machines, then perhaps it is a valid objection – one could argue that the human mind is a computer of sorts, and therefore a claim limited to a computer, broadly defined, is not limited to a machine.  If this is the case, then this rejection should be easily circumvented by defining, for example, a “computer” to require some form of machine-based processing of data using electronic circuits wherein the machine performs various data processing functions.

On the other hand, if the Board’s position suggests or requires that “particular machine” requires more than limiting the claim generically to an electronic computing device, then I think the rejection is logically indefensible for several reasons.  First, this approach to the requirement seems to imply that there is some level of specificity of structure that separates a “particular” computer from a “non-particular” (i.e. general purpose) computer.  There are several problems with this.  First, this is probably better stated as an enablement rejection — i.e. the specification does not support claiming all possible kinds of computers.  Second, if the issue is that a computer has general and not specific utlity, just because a programmable computer has “general purpose” does not make it any less of a particular machine compared to the universe of all machines.  Another is that this requirement for “particularity” is arguably based on the state of the art at the time the invention is examined, and therefore could illogically yield different results at different points in time.  For example, the very first general purpose computer would certainly qualify as a “particular machine” in that no other such machines would exist at the time, but fifty years later that once unique computer design may be ubiquitous, and therefore no longer “particular”, at least as compared to other computers.  

What I don’t think “particular” can possibly mean is that a “machine” be novel in and of itself, without consideration of the algorithm or formula or abstract idea it applies or implements.  The proof of this comes from Diamond v. Diehr, 450 U.S. 175 (1981), where the rubber molding press itself was not novel, but an old and conventional machine:

“The patent examiner rejected the respondents’ claims on the sole ground that they were drawn to nonstatutory subject matter under 35 U.S.C. 101. He determined that those steps in respondents’ claims that are carried out by a computer under control of a stored program constituted nonstatutory subject matter under this Court’s decision in Gottschalk v. Benson, 409 U.S. 63 (1972). The remaining steps – installing rubber in the press and the subsequent closing of the press – were “conventional and necessary to the process and cannot be the basis of patentability.” The examiner concluded that respondents’ claims defined and sought protection of a computer program for operating a rubber-molding press.”  450 U.S. 175, 180. 

Accordingly, it cannot be argued that the term “particular machine” underpinning the MOT test  requires that the machine in question be novel.  More logically, the requirement for a “particular” machine probably comes from the requirement of 35 USC Section 112, second paragraph, which states that “[t]he specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.”  (emphasis added).

Even if the Board gives no weight to the computer-implementation requirement expressly in the preamble, is it still true that the limitations in the body of the claim merely limit “the use of the formula to a particular technological environment”, and that therefore that the claim fails for the same reason as the claim at issue in Parker v. Flook , 437 U.S. 584 (1978).  Here is that claim:  

1. A method for updating the value of at least one alarm limit on at least one process variable involved in a process comprising the catalytic chemical conversion of hydrocarbons wherein said alarm limit has a current value of

Bo + K

wherein Bo is the current alarm base and K is a predetermined alarm offset which comprises:

 (1) determining the present value of said process variable, said present value being defined as PVL;

(2) determining a new alarm base B1, using the following equation:

B1=B o(1.0-F) + PVL(F)

where F is a predetermined number greater than zero and less than 1.0;

(3) determining an updated alarm limit which is defined as B1 + K; and thereafter

(4) adjusting said alarm limit to said updated alarm limit value.

Before I compare the inventions in Flook and Diehr to that in In re Kelkar, it is interesting to note that three of the justices in the majority in the Diehr decision,  Justice Burger, Justice Stewart, and Justice Rehnquist, dissented in the Flook decision, and would have found the claim in Flook patent eligible under Section 101.  So, to even try to square the two decisions based on a meaningful principle of law is perhaps a fool’s errand, particularly since the two cases present substantively the same question of law on the basis of only a seemingly minor difference in form of the claim, and reach opposite results.   Nonetheless, if we are to make anything out of the difference between the claims and outcomes of the Flook and Diehr decisions, it must be that just a slight difference in form of a claim can make a world of difference between patent-eligible and patent-ineligible subject matter.

So why wouldn’t the Board more fairly equate the claimed invention in In re Kelkar to the invention in Diehr as opposed to Flook?  It appears to me that if anything the claimed invention in In re Kelkar is more parallel to the claimed invention in Diehr.  The Kelkar claim recites, in the body of the claim, steps that are limited to being performed in a machine, i.e., a computer — “processing”, “data set”, “repeating … step”, “providing output”, while the body of the claim in Flook recites no such machine-implemented steps.  On the other hand, the algorithm in both Diehr and Flook  was clearly used in an “otherwise statutory” invention – rubber curing and catalytic conversion, respectively, whereas the algorithm in In re Kelkar is applied in the field of data processing – a class of invention that has never been found expressly within patent eligibility by the Supreme Court, notwithstanding the fact that  hundreds of thousands of patents have now been issued for inventions within that field.   

As to this last point, it is of considerable importance I believe to focus on the fact that the invention in In re Kelkar is not directed to the mere solution of an abstract mathematically algorithm, but rather to processing, organizing and correlating data related to properties of physical, biological material .  Strictly speaking, the “algorithm” is at least as much a data processing procedure as it is a procedure for solving a mathematical problem, which is the definition given to algorithm by the Court in Diehr – “We defined “algorithm” as a “procedure for solving a given type of mathematical problem,” and we concluded that such an algorithm, or mathematical formula, is like a law of nature, which cannot be the subject of a patent.”  437 U.S., at 586.  As opposed to solving a mathematical problem per se, the  invention in In re Kelkar provides for clustering genes having potential functional similarity by a comparison of their time varying gene expression profiles.  Since genes by themselves are not mathematical constructs, the “solution” provided by the procedure in In re Kelkar cannot be fairly said to be mathematical in nature.  Accordingly, the algorithm in In re Kelkar is in the nature of a data processing invention that uses a data processing machine to achieve clustering of genes having functional similarity.

The last question to address is whether the invention in In re Kelkar is an “abstract idea” even if not a prohibited mathematical algorithm.  This boils down to whether or not the use of data processing technology to correlate genetic data belongs to the universe of abstract ideas or to the universe of computer-enabled technology that consumes electricity, generates heat and manipulates and reorganizes and analyzes machine-readable data using electronic data registers and functional units.   The only sensible answer is that it belongs to the latter, in which case the invention in In re Kelkar, at least if more explicitly limited to the processing of data, is not an abstract idea but patent eligible subject matter.  Further, even if In re Kelkar involves a mathematical algorithm, that algorithm is not wholly preempted since it has preexisting application to the processing of business information, as noted by the Appellant in its patent specification.

So, what do I mean about the In re Kelkar invention of claim 1 being more explicitly limited to processing data?  This is a point that was not addressed by the Board but I think is pretty fundamental.  If you look closely at the claim, you will take note that it does not differentiate between the data being processed and the information that the data represents.  More particularly, the claim conflates those two notions, which are fundamental to differentiating abstract information from the processing of data representing abstract information.  Abstract information cannot be processed by a computer.  Only data can be processed, and by proxy the abstract information it represents.  Referring to claim 1 again, I note that it only once uses the term “data”, and for the most part makes no clear distinction between “disembodied information” that is abstract in nature and cannot be processed with a computer, and machine-readable data that can be processed using the data processing facilities of the computer.  In this regard, there is little or no attempt made in In re Kelkar’s claim 1 to explicitly recite data, and the electronic processing of that data, such that the method steps of that claim arguably call for a computer to manipulate disembodied genetic information in the abstract, as opposed to manipulating data embodying genetic information.  To the extent that this is the case, then there is some logic in the Board alleging that the recitation of a computer in the preamble of the claim is merely an “environment” for manipulating abstract information, and that the recited method is not tied to the machine by virtue of being embodied in machine readable form that can read and manipulated by the computer.  However, I am not sure that the Board was consciously thinking along these lines when it affirmed the rejection of claim 1 as being directed to an abstract idea, although it may have been thinking along these lines.

In summary, I believe that In re Kelkar was wrongly decided if the Board was treating the method steps as if they recited computer processing of machine-readable data representing genetic information, and got it right, albeit without adequate explanation, if it was thinking that the method steps of claim 1 merely claimed the manipulation of abstract, disembodied data that was incapable of being processed in a computer that operates on machine-readable data, such that the recitation of “computer” in the claim was superfluous or at least irrelevant.

Tags: , , , , ,

Comments are closed.