innovatewebbanner2

Innovate Magazine

INNOVATE is the online magazine by and for AIPLA members from IP law students all the way through retired practitioners. Designed as an online publication, INNOVATE features magazine-like articles on a wide variety of topics in IP law.

Publishing an article to INNOVATE is a great way for AIPLA members to build their brand by increasing recognition among peers and setting themselves apart as thought leaders in the IP industry.

Any current AIPLA member in good standing may submit an article for consideration in INNOVATE throughout the year. IP law students are especially encouraged to submit articles for publication.

Articles submitted to innovate@aipla.org are reviewed by an ad-hoc sub-committee of volunteers from AIPLA's Special Committee on Publications, the Fellows Committee, and other AIPLA peers. Approved articles are published quarterly. 

Don’t miss your chance to be published with AIPLA’s INNOVATE! Email your article submission to innovate@aipla.org to be considered for the next edition.

For more information please review the Guidelines for Article Submission and the Author Acknowledgement Letter for guidelines and terms of article submission and publication.

 

 

 

Trends, Developments, and Filing Strategies of AI Medical Patents in Taiwan


George Jui-Hsien  Huang

The use of artificial intelligence (AI) technology has been booming all around the world. The increased use of AI is reflected in an exponentially growing number of AI related patent applications according to the World Intellectual Property Organization (WIPO)[1]. The interest on medical related inventions has also been on the rise due to COVID-19[2], resulting in medical related patent applications that involve AI, one of the hottest trends in technological development. Statistics[3] show that in Taiwan, the number of AI related patent applications filed in the medical field comes in second of all fields, right behind patents applications related to industry and manufacturing. This article provides an analysis of filing trends for AI related medical patent applications, discusses problems faced in patentability and eligibility, and possible future amendments to related examination guidelines in Taiwan, and provides strategies for filing such patent applications in Taiwan.

 

Part I. Filing trends of AI medical patents

A. Applicants

A list of the 10 applicants filing the most AI medical patent applications in the world[4] includes five electronics technology companies (Siemens, Philips, GE, Samsung, and Fujifilm), two information and software companies (IBM and Microsoft), two medical equipment companies (Heartflow and Medtronic), and one academic and research institute (The University of California in the US). The companies on the list often partner on research and development projects as well, including partnerships involving Siemens and IBM, GE Healthcare and Heartflow, to name a few[5].

In Taiwan, academic and research institutes account for over 50% of the total AI medical patent applications, followed by 34% by enterprises, the top two being HTC and Foxconn Technology Group, and 13% by natural persons [6]. This is quite different from the proportions worldwide, and also different from the filing trends of the IP5[19]. For instance, applicants for AI medical patent applications filed with the United States Patent and Trademark Office (USPTO), European Patent Office (EPO) and Japan Patent Office (JPO) are mostly electronics technology companies, and the China National Intellectual Property Administration (CNIPA) also receives a large number of AI medical applications from information and software companies[7]. We believe this shows that AI medical technology is still under development in Taiwan, and that in the future, we can expect further advances in the field of AI in medicine with the help of academia-industry collaborations, technology transfer, and even licensing.

 

B. Technical Fields

In terms of technical fields in which AI medical patents are applied, Taiwan shares similar trends with the global patent landscape. Most AI medical patents in Taiwan are applied in physiological diagnoses and measurements, followed by data calculation and medical data management. There is also a great number of applications within the field of data processing with machine learning and image recognition technology. The majority of the applications focus on data processing, physiological measurements, image processing and recognition, genetic testing, and medical and biological information, etc., replacing conventional medical testing with novel AI technologies. What is also noteworthy is that in patents involving computer-aided diagnosis and treatment, most are directed toward breasts, e.g., targeting tumors or calcified lesions, followed by those directed toward lungs. Diagnostic and therapeutic technology directed toward brains and intestines have also shown a surge in development along with the advent of image recognition and segmentation technology[8].

 

Part II. Current Filing Practice and Recommended Strategies of AI Medical Patents in Taiwan

A. Subject Matter Eligibility

AI technologies often involve pure algorithms or model improvements, which are easily deemed abstract ideas. Diagnostic and therapeutic methods, which are commonly utilized in the medical field, might be found to be directed to judicial exceptions, and hence ineligible for patent protection according to Article 24 of the Taiwan Patent Act[20]. AI medical patent applications must simultaneously deal with these two aspects of subject matter eligibility in Taiwan.

According to Article 24 of the Taiwan Patent Act, diagnostic, therapeutic, and surgical methods for the treatment of humans or animals shall not be granted. As for the definition of diagnostic methods, the Taiwan Patent Examination Guidelines further stipulates that[9]:

A diagnostic method shall not be granted only if it meets the three criteria:

  1. The method is performed on animate human or animal bodies;
  2. The diagnosis relates to a disease;
  3. The direct objective is to acquire the diagnosis of the disease.

Therefore, other inventions such as in vitro testing or processing of samples taken from human or animal bodies, and methods irrelevant to the diagnosis of a disease or mere measurements of human or animal properties do not fall within the judicial exceptions and are not eligible. The same applies if the information obtained by the invention is only intermediate results and cannot be used to directly acquire diagnostic results.

Current criteria for subject matter eligibility of AI-related patent applications in Taiwan are based on the Examination Guidelines of Computer Software-Related Invention[10], which focus on whether the invention involves “technical character” and “technical effects.” Where there is further technical effect going beyond the normal physical interactions between the program and the computer in the execution of a program, the technical means as a whole has technical character, and the invention is in accordance with Article 21 of the Taiwan Patent Act. It can be inferred that currently the Taiwan Intellectual Property Office (Taiwan IPO)’s criteria for determining the patentability of AI inventions are akin to those of the EPO, also stipulating that a computer program is patentable when having a technical character, which may be defined if the computer program produces a "further technical effect" when run on a computer[11]. The criteria differ in that the EPO only requires further discussion of technical effects when the claimed invention is a “computer program (product),” while the Taiwan IPO has yet to provide a set of consistent principles on this matter.

For now, the majority of the rejected AI-related invention patent applications in Taiwan are rejected for the lack of novelty or obviousness, or for claims lacking clarity, and rarely are such applications rejected solely based on subject matter ineligibility. To ensure that the subject matter is considered eligible, it is recommended to highlight the outstanding medical and therapeutic effects in the specification. Furthermore, comparing with other countries, such as China[12], subject matter eligibility of diagnostic methods is also generally judged less strictly in Taiwan. Nonetheless, patentees are advised to frame “articles” as the claimed subject matter in their patent applications, such as the equipment or apparatus used to implement the medical diagnostic methods, so as to prevent the method patent from falling within the scope of “diagnostic methods.”

 

B. Non-obviousness of AI Technologies

A computer software-related invention is considered obvious if its technical character is merely a “conversion between technical fields,” “addition or replacement of publicly known technical features,” “systemization of human operation methods,” or “a software achieving the effects of prior hardware technology.”[13] Further, the Taiwan Examination Guidelines[14] stipulate that the following aspects shall be considered: (1) whether the technical feature contributes “no technical character to the invention”; and (2) whether the technical feature contributing no technical character cooperates with the technical feature having technical characters, and therefore contribute a technical character to the invention. If not, the invention will be deemed as a mere combination of known art, and thus considered obvious.

When a technical feature is considered to make no contribution to the technical character of an invention, it should be further inspected if the technical problem can still be solved without the said technical feature. If not, said technical feature is considered to contribute to the technical character of the invention, and shall not be deemed as known art.

 

Part III. Possible Amendments to the related Taiwan Patent Examination Guidelines in the Future

In terms of subject matter eligibility, the current Examination Guidelines of Computer Software-Related Invention in Taiwan opt for a combination of the criteria adopted in the US, Europe, and Japan. In 2004, the Taiwan Patent Act underwent a significant amendment, introducing the concept of “technical character” stipulated in the EPO regulations. The term “technical effect” was further introduced in 2008. In 2014, the law adopted the US opinion that an invention lacks technical idea if it is “nothing more than the idea” of doing that thing on a computer. However, the US subsequently amended the criteria for determining subject matter eligibility based on the decision rendered in Alice Corp. v. CLS Bank International[15]. The Taiwan Patent Examination Guidelines of Computer Software-Related Invention therefore did not adopt the latest US regulations, while also only adopting part of the EPO’s Examination Guidelines[16]. As a result, the current criteria for determining subject matter eligibility in Taiwan has lost clarity in an excess of elements, rendering the standards different from case to case.

Consequently, professionals have expressed concerns about the confusing standards of subject matter eligibility of AI-related patents[17]. Some argue that the criteria are too strict, while others believe too much emphasis is put on “technical effect,” which is an element of non-obviousness that should be considered separately. The Taiwan IPO therefore plans to announce an amendment next year[18] which adopts regulations similar to the JPO, determining whether an invention “substantially realizes information processing of the computer software with hardware resources.” The amendment will include a set of revised steps for determining non-obviousness, an addition of examination examples and clarifications of new technologies such as Internet of Things (IoT) or AI, and important factors to be noticed for the examination on AI medical patents.

 

Part IV. Conclusion and Suggestions

The rapid growth of novel AI medical technologies is reflected on the number of patent applications filed with each major patent office. Such technologies are also showing a steady development in Taiwan. However, in view of the controversy surrounding the subject matter eligibility of AI patents, one should take notice of the different regulations in each country upon filing. Although diagnostic methods in Taiwan are judged less strictly than in other countries, whether they meet the definition of an invention remains ambiguous. Applicants should also pay attention to the possible significant amendments to the related examination guidelines in the future to develop a successful patent portfolio.



[1] WIPO Technology Trends 2019 – Artificial Intelligence (2019). World Intellectual Property Organization

[2] the Taiwan Intellectual Property Office, Taiwan Patent Information Related to International Potential Drugs for COVID-19 Treatment. https://www.tipo.gov.tw/tw/cp-85-873260-f17e4-1.html.

[3] Li, C.-C. (2019) Practical development of Examination and Response Strategies of Taiwan AI-related patent applications. The Taiwan Intellectual Property Office.

[4] Xue, L.-H., Zhong, P.-X., Huang, Y.-L. (2020) Patent Analysis of eHealth Technology. Intellectual Property Right Journal VOL.258.

[5] Xue, L.-H., Zhong, P.-X., Huang, Y.-L., op. cit.

[6] Xue, L.-H., Zhong, P.-X., Huang, Y.-L., op. cit.

[7] Xue, L.-H., Zhong, P.-X., Huang, Y.-L., op. cit.

[8] Feng, Y.-H., Chen, H.-M. (2020) Patent Analysis for Smart Diagnosis Based on Computer Aided Diagnosis. Intellectual Property Right Journal VOL.258.

[9] Chapter 2, Part 2: Definition of an Invention. (2013) Examination Guidelines for Patents. Taiwan Intellectual Property Office.

[10] Chapter 12, Part 2: Computer Software-Related Invention. (2014) Examination Guidelines for Patents. Taiwan Intellectual Property Office.

[11] Section 3.6, Chapter II, Part G: Programs for computers. (2019) Guidelines for Examination. European Patent Office.

[12] Chapter 1, Part 2: Invention-Creations for Which No Patent Right Shall Be Granted. (2010) Guidelines for Patent Examination. the National Intellectual Property Administration of the People's Republic of China.

[13] Chapter 12, Part 2: Computer Software-Related Invention. (2014) Examination Guidelines for Patents. Taiwan Intellectual Property Office.

[14] Chapter 12, Part 2: Computer Software-Related Invention. (2014) Examination Guidelines for Patents. Taiwan Intellectual Property Office.

[15] Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014).

[16] Section 3.6, Chapter II, Part G: Programs for computers. (2019) Guidelines for Examination. European Patent Office.

[17] Public Hearing of Draft Amendments to the Examination Guidelines of Computer Software-Related Invention. Taiwan Intellectual Property Office. (2021).

[18] Zhu, H.-Y. (2020) Cong geguo ruanti zhuanli shencha shiwu fansi woguo ruanti zhuanli. (Reflection on Possible Amendments to the Taiwan Examination Guidelines for Software Related Patents based on Worldwide Examination Practice) The Taiwan Intellectual Property Office.

[19] the European Patent Office (EPO), the Japan Patent Office (JPO), the Korean Intellectual Property Office (KIPO), the National Intellectual Property Administration of the People's Republic of China (CNIPA), and the United States Patent and Trademark Office (USPTO)

[20] Article 24 Judicial Exceptions

An invention patent shall not be granted in respect of any of the following:

1. animals, plants, and essential biological processes for the production of animals or plants, except for processes for producing microorganisms;

2. diagnostic, therapeutic and surgical methods for the treatment of humans or animals; or

3. inventions contrary to public order or morality.


George is the Director of Wisdom International Patent & Law Office. He is the first lawyer having a solid technical background (biotechnology) to be qualified as a patent attorney in Taiwan. Having over 27 years' experiences in the intellectual property field, George is skilled in handling patent, trademark and design matters, including prosecution, infringement, litigations, cancellations and oppositions.

George is fluent in English and Japanese and he demonstrates his strengths by assisting notable overseas and Japanese corporate clients from diverse industries to obtain worldwide IP protections, including pharmaceutical, mechanical, chemical, cosmetics, manufacturing, retail, luxury goods, financial service and software industries.

George actively participates in professional associations and is a member of AIPLA, Councillor & Supervisor of Asian Patent Attorneys Association (APAA) Taiwan Group, Enforcement Committee of International Trademark Association (INTA), Trademarks Committee of APAA, and International Federation of Intellectual Property Attorneys (FICPI). George was also recognized as a leading professional in the 2021 Edition of World Trademark Review 1000.

 

 

Please login to see the content.
Register now