No pending appeal to the Supreme Court has filled inventors and innovators with more dread than Bilski v. Kappos. Before the court was nothing less than the question of what can and cannot be patented.

Some feared a sweeping decision that would wipe away entire categories of patents, including those covering software, business methods, and many types of medical diagnostics. On June 28, 2010, the Supreme Court issued its opinion and inventors heaved a sigh of relief.

The majority opinion confirmed that abstract ideas, laws of nature and physical phenomena are not patentable, but declined to conclude that software or business methods are inherently unpatentable abstract ideas, or that medical diagnostics involve only unpatentable physical phenomena or laws of nature.

Indeed, the court rejected as overly rigid the Federal Circuit’s approach for determining what methods can and cannot be patented.

As we have reported, Bernard Bilski filed a patent related to hedging the seasonal risks of buying energy. The patent claims had very general language. Mr. Bilski was rebuffed by the patent examiner, who found that the application described an abstract idea, not an invention. This view was upheld by the Board of Patent Appeals.

The Federal Circuit likewise found the application ineligible for patent protection, but its opinion was based upon a newly articulated “machine or transformation” test. To pass this test for patentability, a process must either be tied to a special-purpose machine or apparatus or transform an article into “a different state or thing.” (The meaning of “special-purpose machine” is not clear, but might include a general purpose computer running a program that causes the computer to perform specialized functions.)

The majority and concurring opinions in Bilski reveal a sharp debate among the justices on many issues of patentability. As a guide to practice, it is useful to identify the areas where at least five justices agree, for it is on those points that the Federal Circuit is likely to focus its future efforts to define the boundaries of patentability. The following points drew majority support:

• Abstract ideas, laws of nature and physical phenomena are not patentable. (Bilski’s patent application falls victim to this holding.)
• The idea expressed by the Federal Circuit in State Street Bank & Trust Co. v. Signature Financial Group Inc. – that patent coverage can be obtained for anything that produces a “useful, concrete and tangible result” – is disavowed.
• While the “machine or transformation” test cannot be an exclusive test for determining whether a process is patentable, it is nonetheless quite useful.
• A business method may be eligible for patent protection if it satisfies the machine-or-transformation test or otherwise can be shown to be within the category of patent-eligible processes.

On the remaining issues, the Court was fragmented. Four of the conservative justices (Kennedy, Roberts, Thomas and Alito) suggested that the “Information Age” requires new paradigms regarding patent-eligible subject matter, while four liberals (Stevens, Breyer, Ginsburg and Sotomayor) argued for strict construction based upon a review of centuries of precedent. Justice Scalia divided his vote between these two groups, depending on the issue. For example, he seems to favor business method patents, while shying away from the “Information Age” paradigms.

What then can we conclude about patents previously believed to be at risk–those based upon software, business methods, medical diagnostics and other fields that might be seen as involving only abstract ideas or phenomena occurring in nature?

As a first principle, patent applications for processes in these fields should, if possible, include claims anchored to tools, disposables, special-purpose computers or machines, and/or to the transformation of physical articles in order to satisfy the machine-or-transformation test.

While this test sometimes seems difficult to satisfy, it can be interpreted with some latitude. The medical diagnostics field is currently the beneficiary of such an interpretation. As we discussed in an earlier article, the Federal Circuit held in Prometheus Laboratories, Inc. v. Mayo Collaborative Services (2009), that a diagnostic test that administered a compound to mammals and measured the level of a metabolite into which the digestive system broke the compound was patentable, and was not merely a natural phenomenon.

The minority in Bilski made favorable reference to the dissent in Laboratory Corp. of America Holdings v. Metabolite Laboratories, Inc. (2006), which would have denied patent eligibility to such a diagnostic method. In their view such a test merely reveals a natural phenomenon. Thus, the patent eligibility of diagnostics is controversial among the justices of the Supreme Court, but for now its critics are in the minority.

The notion that business methods may be patented was confirmed by a majority inBilski, notwithstanding an eloquent essay to the contrary authored by Justice Stevens and joined by three other justices. While business methods are not ineligible, claims addressed to business methods should be drafted to satisfy the machine-or-transformation test.

As for software, the outlook is murkier. Justice Kennedy, writing for himself and three other justices, made the only mention of software in Bilski, noting that the machine-or-transformation test creates uncertainty as to the patentability of software. No resolution of this uncertainty can be gleaned from the opinion. However, it is instructive that a portion of the opinion joined by the majority reviewed prior Supreme Court opinions in which software played a role.

In Gottschalk v. Benson (1972), the court held that an algorithm to convert decimal numbers into binary code was an unpatentable abstract idea. Later, in Diamond v. Diehr (1981), the court upheld the patentability of an invention for molding uncured synthetic rubber into cured precision products in a process that included having a computer follow a mathematical formula.

Diehr explained that while the formula could not be patented, the application of the formula to a known structure or process could be. The court noted that the claim was directed to an industrial process for the molding of rubber products.

In light of Bilski’s affirmation that business methods may be patented, it seems unlikely that software claims need to be tied to an industrial process to be valid. Nonetheless, it will be much easier for the courts to uphold such claims if the software plays a role in a process that satisfies the machine-or-transformation test.

While the majority opinion in Bilski said that the machine-or-transformation test is not exclusive, we suspect that the Federal Circuit will be reluctant to expand the contours of patent eligibility beyond that test for fear of reversal. Despite that limitation, and the rejection of the notion that patents can cover “anything under the sun,” inventors can breathe a little easier as a result of the Supreme Court’s opinion.

For a number of years the court has been critical of the patent system. In the Supreme Court’s eBay decision (2006), which made it harder for patent holders to get injunctions, Justice Kennedy wrote a concurring opinion, which three other justices joined, in which he complained of the “vagueness and suspect validity” of patents being asserted by firms using “patents not as a basis for producing and selling goods but, instead, primarily for obtaining licensing fees.”

Now, in 2010, when the United States economy has been licking its wounds from a worldwide recession, the tone of the court is different. For once, the Court chides the Federal Circuit for being too narrow in its view of patent eligibility. Perhaps the pendulum has finally begun to swing the other way.