Analysis Physics World  June 2017

Are your US patents safe?

Inventors work hard to prepare patent applications that will stand up against the competition in court, but several precedent-setting US cases indicate that many of these patents may not be valid. US patent attorneys Curtis Vock and Douglas Link explain how the dynamic has changed

Legal landmarks Recent decisions have reshaped the landscape for US patent claims. (iStock/padnpen)

In 2005 Quad/Tech received a patent (US patent 6,867,423) for a visual inspection system that uses a CMOS detector to inspect images on a substrate illuminated by LEDs. Unfortunately, because this patent failed to illustrate both the LEDs and the substrate in a single drawing, a federal appeals court found some of the patent’s claims invalid in 2014.

That same year a similar fate befell TSI, which owned US patent 6,831,279 for a laser diode that illuminates biomolecules with radiation above a certain wavelength λ. In this patent the illumination causes the mole­cules to fluoresce; a photon counter senses this fluorescence; and a microprocessor evaluates the signal’s validity. Nonetheless, because it was not practical to obtain diode emission wavelengths at λ at the time the patent was filed, an Arizona district court found the patent invalid.

A third case concerned US patent 6,799,864, which was issued in 2004 for a spot module illuminator that controls thermal loading by employing a thermal spreader and an elongated thermally conductive core connected to heat-dissipating fins. Because “elongated” is a term of degree and the patent did not sufficiently differentiate “elongated” from “non-elongated” cores, a federal circuit court found the patent invalid in 2015.

Although these three cases revolve around photonics and optics technologies, we are seeing similar situations across a whole range of hi-tech industries. The most important common thread between these three cases isn’t their link to photonics; it’s that their patents ran afoul of rigorous new US standards on several important aspects of patent law.

Your patent is obvious

The trend started around 10 years ago when the US Supreme Court made it much easier to label an invention “obvious” and thus not patentable. According to the decision in KSR International v. Teleflex, Inc. (2007), an invention is “obvious” when prior art – that is, anything publicly disclosed and available in journal articles, Internet postings, patents or even products – can be reasonably combined into something that is the same as your invention by a person “skilled in the art”. Here, the term “skilled in the art” denotes a hypothetical person with skills in the relevant industry; for a photonics invention, a physicist may easily qualify as someone skilled in the art. So if this hypothetical person could argue, for example, that the synthesis of two (or more) pieces of prior art is the same as your invention, then your invention is obvious and you are not entitled to a patent.

The US patent office (USPTO) sometimes uses “obviousness” arguments to deny patent applications, but a competitor can also make such arguments, either in court or at the USPTO – even after the patent is issued. For example, a patent (US patent 7,934,041) assigned to Crossroads Systems was invalidated under KSR on a petition by Cisco Systems to the US Patent Trial and Appeal Board (a panel of judges at the USPTO) because evidence of non-obviousness was not commensurate in scope with the invention.

In this case, the patent described a storage router including a controller that provided virtual local storage between multiple devices of different protocols by connecting remote small computer system interface (SCSI) devices with fibre channel (FC) devices. In an attempt to avoid prior art cited by Cisco, Crossroads stated that its invention was non-obvious because it resolved the long-felt problem of “bottlenecks in network file system performance”. The appeal board was not persuaded, however, because Crossroads’ arguments did not establish how the invention (for example, the constraints of the controller) resolved these problems. In other words, the invention was not sufficiently tied to the reasons that Crossroads used to argue against obviousness, and thus the combination of prior art references invalidated Crossroads’ patent.

Possessing the invention

In 2010 another US court made a similarly far-reaching decision. In Ariad Pharmaceuticals, Inc. v. Eli Lilly and Co., the federal circuit court (which sits below the US Supreme Court, but above the USPTO and US district court) required patent descriptions to demonstrate that the inventor had “possession” of the invention claimed at the time of filing. Unlike most federal circuit decisions, this one was made by all of the circuit’s judges (an en banc decision), so it carried a lot of weight. Following Ariad, many other federal court decisions have invalidated patents on similar bases.

But what does the term “possession” mean in this context? Well, the en banc judges in Ariad stated that the written description of a patent must demonstrate reasonable “structure-function” correlation for the claims it makes. The claims of a patent are sentences that describe the scope of the invention (for example, “I claim a new widget that creates noise”), while the specification refers to the text and figures describing those claims (for example, describing and showing the claimed widget and how it creates the noise). The problem occurs when the text and drawings do not sufficiently support the words in the claims. When that occurs, the patent is invalid. Sometimes this failure is characterized as failing the “written description” requirement of a patent.

A real example of an invalid patent that did not demonstrate possession concerned a prosthetic disc implant used to replace a spinal intervertebral disc. US patent 7,429,270 was assigned to a company called Synthes, and the description of this implant mentioned a “plurality of openings” on the upper and lower surfaces to promote bone growth when implanted in the spine. In a lawsuit filed by a competing company, Spinal Kinetics, this patent was found to be invalid because the term “openings” was not reasonably ­supported by the figure in Synthes’ patent, which showed only narrower mechanical grooves. Spinal Kinetics’ rival spinal implant product did not employ grooves but instead used trapezoidal and circular slots. Following the precedent set in Ariad, the court determined that Synthes had not shown possession of the broad “openings”; instead, it only possessed intervertebral implants
having narrower “grooves”.

Ineligible subject matter

A third series of decisions concerns a topic dear to the hearts of many physicists: natural law. Calculating how long it takes for an apple to reach the ground after it falls from a tree is not really an invention at all, and patent law recognizes this. However, a trio of US Supreme Court cases beginning in 2012 (Mayo Collaborative Services v. Prometheus Laboratories, Inc., Association for Molecular Pathology v. Myriad Genetics, Inc., and Alice Corp. v. CLS Bank International) have made it much easier to label an invention “natural law” and thus ineligible for patenting. They also made it easier to label a patent “abstract” – another ineligible category. While there is no clear definition of an “abstract” idea, representative examples might include such things as fundamental financial/economic business practices; processes that can be implemented in the human mind; mathematical algorithms not tied to a specific machine; or the mere collection, display and manipulation of data.

Spaced out A drawing of Biosig Instruments’ heart-rate monitor, taken from the patent application. The application stated that the electrodes on the bar at the top were “in a spaced relationship”, but the US Supreme Court found this phrase “indefinite” and invalidated the patent. (Gregory Lekhtman/Biosig Instruments/Patent US5337753 A)

This series of important cases immediately called into question all types of medical diagnostic and software inventions, and even hardware inventions employing off-the-shelf components stitched together by algorithmic or proprietary logic. As an example, consider a software patent assigned to an intellectual-property firm, Intellectual Ventures. This particular patent (US patent 7,984,081) related to systems and methods for dynamically managing extensible markup language (XML) data, and it included behind-the-scenes software that manipulated XML documents in different XML formats. This software meant that the user could view and update the multi-format XML documents without a need for programming skills.

The USPTO granted this patent in July 2011, but in March 2017 the federal circuit court invalidated it on the grounds that modifying XML was merely an “abstract” idea of collecting, displaying and manipulating data – that is, collecting data from different XML formats, manipulating it into a consolidated XML format and displaying it according to the user’s modifications. In particular, the court invalidated the patent under Alice because it found that claims limited to a particular type of computer document (XML), while using conventional hardware, did not differentiate the claims from an “abstract” idea (collecting, displaying and manipulating data) and thus the claims were ineligible for patent protection.


In 2014 the US Supreme Court clarified the circumstances in which a patent is invalid because its claims are “indefinite”. The Nautilus, Inc. v. Biosig Instruments, Inc. case centred on Biosig’s patent (US patent 5,337,753) for a heart-rate monitor. The description of this patent mentions a hollow cylindrical bar that a user grips with both hands, such that each hand comes in contact with two electrodes (one “live”, one “common”) that are “in a spaced relationship”. After an appeal questioned whether the phrase “in a spaced relationship” was indefinite, the Supreme Court decided that a patent is indefinite (and thus invalid) if the claims, read in light of the patent text as well as prosecution history, fail to inform with reasonable certainty those skilled in the art about the scope of the invention.

One patent that fell afoul of this decision concerned a multiple-sclerosis drug developed by Teva Pharmaceuticals. The patent (US patent 5,800,008) described the drug as including “copolymer-1 having a molecular weight of about 5 to 9 kilodaltons”. Following Nautilus precedent, the court found that since there are multiple ways of determining molecular weight, and the patent did not clarify how “molecular weight” was measured, the invention was indefinite and the patent is invalid.


These cases (plus many others we could mention) place many existing patents at risk, but that is far from their only impact. Such rulings may also prove fatal to future patents if applications are not properly drafted or argued with the USPTO. The solution is for inventors and companies to endeavour to become much more diligent when filing new patent applications. The written description is the foundation for “fallback” positions (that is, narrower patentable scope for your inventions than originally envisioned at filing) to counter challenges to validity based on anticipation (essentially the exact invention is disclosed in a single piece of prior art), obviousness, indefiniteness or possession. Hence, applicants need to consider including more and better examples, as well as fuller descriptions that delineate the scope of the inventions. Before filing, inventors may wish to contemplate and describe not only current commercial implementations, but also foreseeable ones in as much detail as possible. In short, now is a good time for companies to think about revisiting their patent portfolio and examining their strategy for filing stronger and better patents.