Art reference

O prior art, prior art, why are you prior art?

by Dennis Crouch

In a new opinion, the court asked and answered an interesting question: what if most of the prior art on the point was created accidentally due to a typographical error? In LG Electronics, Inc. c. ImmerVision, Inc., — F.4th — (Fed. Cir. 2022), the Court sided with the patentee in concluding that a person skilled in the art would have “ignored or corrected” the error rather than relying on the error as a basis of an inventive project. A key to the analysis was the finding that the error would have been obvious to a person skilled in the art. See In re Yale434 F.2d 666 (CCPA 1970).

ImmerVision’s patent claims an improved method for panoramic images. The basic idea is to use a nonlinear objective lens captured and then digitally mapped onto a linear pixel grid. The claims require that the lens compresses the image at its center and edges and expands the areas in between. You can compare the results of the prior art linear lens (Fig. 4A) with those of the proposed panoramic lens (Fig. 7A).

Due to lens curvature, the center and edges of a super-wide-angle lens generally produce lower image quality. The approach here improves the quality of these areas but at the same time creates weird distortions in the image which then need to be corrected using digital image processing. There are other solutions that would also work. An example would be a higher density of sensors in these particular areas.

Many smartphones now have multiple lenses, including panoramic lenses. ImmerVision sued LG (and others) for patent infringement and LG responded with two IPR claims. The PTAB instituted the IPR here, but ultimately concluded that LG had failed to prove that the claims should be overturned. On appeal, the Federal Circuit upheld. Judge Stoll wrote the majority opinion joined by Judge Cunningham. Justice Newman wrote in partial dissent – arguing that the error in the prior art was not a sufficiently obvious error.

Typo in the prior art: Key prior art (Tada) included an array of aspherical data that appeared to show a nonlinear lens similar to ImmerVision. The problem – these numbers do not match the other descriptions of the embodiment in the Tada spec and there is no discussion in Tada about having such a non-linear lens. Tada is a US patent, and the decisive point for the court was that the original Japanese priority document included different numbers that actually matched the other parts of the specification. Finally, LG did not dispute that “the aspherical coefficients in Tada’s Table 5 were wrong”.

Of course, patent attorneys usually don’t get a balcony scene asking: O prior art, prior art, why are you prior art? Instead, we take the printed publication as is and compare the published words and numbers to the patent document. But for the sake of obviousness, we ask an additional WWPD question: What would POSITA do? And here we have precedent with the court announcing that a person skilled in the art (POSITA) would ignore obvious typography in the prior art and instead correct the typo before trying to change the reference. In re Yale434 F.2d 666 (CCPA 1970).

In Yale, a patentee sought protection for a specific anesthetic inhalation compound—CF3CF2CHClBr. The USPTO rejected Yale’s claims as obvious based on a prior art publication that listed this same compound (CF3CF2CHClBr) as a known important analgesic compound. The compound was not actually known at the time and it was clear from the context that the author of this prior art publication meant CF3CHClBr which was in fact well known at the time. In correspondence, the author of the prior art admitted that this was “of course a mistake”. Yet the PTO dismissed the claims as obvious since the compound was in fact disclosed in the prior art. On appeal, the CCPA overruled – finding that such POSITA would “mentally ignore” such an “obvious error” in the prior art.

He would certainly not be led by the typographical error to use the erroneous compound as an anesthetic even if, as a chemist of ordinary skill, he would know how to prepare the compound. He just wouldn’t go that far in the thought process to figure out if he knew how to make CF(3)CF(2)CHClBr, because it would have long since been dismissed by him as an obvious typographical error.

Yale. An aside on Yale. Yale was decided by the Federal Circuit’s predecessor court, the Court of Customs and Patent Appeals (CCPA), and remains good law and binding precedent in the Federal Circuit. In its first ruling in 1982, the Federal Circuit adopted all previous CCPA holdings as binding precedent. South Corp. vs. United States690 F.2d 1368 (Fed. Cir. 1982):

In LG, the Board found that the Tada error was an obvious error consistent with the decision of Yale. On appeal, the Federal Circuit asserted — finding that substantial evidence supported the Board’s findings.

This is where Judge Newman’s dissent comes in. She notes that the error was not discovered “until an expert witness had conducted a dozen hours of experimentation and calculation” . Additionally, the owners of Tada had filed a correction certificate to correct certain aspects of the patent, but had not corrected the table. Judge Newman also notes that the only way for the expert witness to firmly conclude that the error existed was to translate the Japanese priority document – ​​something even optical experts are unlikely to do.

I agree with the majority of the panel that Yale sets the correct standard for determining whether an error would be obvious to a person of ordinary skill in the field. However, I disagree with the majority’s application of this standard to the facts of this case. An “obvious error” should be obvious at first glance and should not require the conduct of experiments or the search for potentially conflicting information to determine whether an error exists. When a reference contains an erroneous teaching, its value as prior art must be determined.

The error in the Tada reference is clearly not a “typographical or similar error”, as the error is not apparent at first glance and the correct information is not readily apparent. It should not be necessary to search for a foreign document in a foreign language to determine if there is an inconsistency in a US patent. The basis of the “typographical or similar” standard is that the error is easily recognized as an error. I’m concerned that we’re upsetting a long-standing law, so I respectfully disagree in part.

Drag Operation. (Newman, J. dissenting). The majority ignored Judge Newman’s arguments: “The distinction between the typographical error in Yale and the copy-and-paste error here is a distinction without difference.” ID. (majority opinion).

LG had asked the court to change its standard – only consider typographical errors that are “immediately” obvious. The court rejected this temporal urgency as not being required by Yale.