The impact of Artificial Intelligence on patent system has raised a unique opportunity globally in inviting several debates on issues relating to strategies to be adopted in protecting AI based inventions as well as drafting and prosecution of patent applications in different geographies. The accumulation of big data in the past couple of decades, flooding of algorithms and better processing capacity of computers has led to the explosion of AI in every other walks of life in the forms of broad spectrum of technologies, inventions and companies. In India, over 5000 AI patent applications were filed over the last decade, out of which 94% were filed in the last 5 years. The methods of AI in healthcare industry includes a gamut of applications such as drug design, radiology, imaging, prediction of a disease, medical diagnosis for example cancer detection, genomics, proteomics, pharmacology and so on. One of the major milestones of AI implemented methods in healthcare industry is enhancing the accuracy and efficacy of diagnosis. The following article throws light on the plethora of business opportunities that has manifested in the form of startups in India at the same some of the practical roadblocks encountered by inventors of healthcare related AI inventions in patenting their inventions in India.
AI based Healthcare Startups in India
With strong backing from software services, India is strategically placed to build a robust AI system as it seeks to establish itself as a hub of AI development. According to a report issued by NITI AAYOG titled “RESPONSIBLE AI” (Part 2), “During COVID-19 AI image recognition solutions and ML- based resource allocation disaster platforms greatly enhanced state’s capability to deliver services bridging gaps of limited access, resources, healthcare delivery and knowledge.” AI has also been recommended by the Indian Judiciary in various instances to uphold the fundamental rights of citizens and improve efficiency. Over the last two decades around 324 AI based healthcare startups were started in India. Some of them are as follows;
Innovaccer – Founded in 2014, a provider of a cloud and AI-based patient health data analytics platform that aids in virtual care networks, remote patient monitoring, Social determinants of health (SDOH)management, surgery optimization, CRM and referral management and so on.
SigTuple- Founded in 2015, provides a data-driven, machine-learned, cloud-based solution for the detection of anomalies and trends in medical testing to improve the accuracy and efficiency of disease diagnosis
Perfint Healthcare- Founded in 2005, provide image-guided equipment uses medical image processing and robotics to help clinicians plan and execute CT-guided cancer biopsies, drug delivery and ablation of small tumors. Maxio is an FDA 510(k) approved device.
Healthplix- Founded in 2014, provides AI and cloud based electronic medical record management which includes features such as e-prescription generation, lab management, billing, and dashboards providing AI & machine learning-based insights related to finance, marketing, test clinical hypothesis, and treatment outcomes.
Niramai Health Analytx- Founded in 2016 a developer of AI-based early-stage breast cancer screening devices and products like Thermalytix which is an AI-based diagnostic tool using a cloud-based platform for analyzing thermal images. The other product is “Smile Tool” which is a real-time cloud-based diagnostics tool for breast cancer screening. Niramai Health Analytx have been granted 4 patents in India and 10 in the US.
Training AI and ML models in the field of healthcare and medical diagnosis
In a scenario before AI, patients can present a huge amount of medical data, such as X-Ray images, CT Scans, histology reports and biochemical test reports. For a human to assimilate such vast amount of data might take years in order to provide a diagnosis and recommend a treatment for each individual case taking into account the numerous permutations and combinations of symptoms and signs. After the advent of AI, wherein AI and ML models are trained to look for specific patterns in the dataset, for example a set of cancer biopsies and take new biopsy data from a patient and classify it into a likely diagnosis quickly and efficiently. The classification hence obtained can be used by the medical professionals in assessing a patient’s clinical needs.
Though the application of AI to healthcare is plentiful, several roadblocks are encountered in patenting AI based healthcare inventions in India. From the perspective of Indian Law, the challenges posed are exclusion of patentability of mathematical methods, computer programs and algorithms as per Section 3(k) of the Indian Patents Act, 1970. Another major challenge is patentability of methods of treatment and diagnosis which is a murkier and contested subject in India.
Patentability of computer related inventions in India
As per Section 3(k), a mathematical or business method or a computer programme per se or algorithms related invention are not patentable matter unless the invention shows technical advancement over the existing prior art(s). In the case of Ferid Allani Vs Union of India and Ors., the Intellectual Property Appellate Board (IPAB) set aside the refusal order of the Indian Patent Office (IPO) and allowed the appellant’s patent application. By setting aside IPO’s refusal order, IPAB concluded that IPO has not given due weightage to ‘technical effect’ (reduction in bandwidth usage and search mean-time duration) and the ‘technical contribution’ (generation of more precise search query locally in the client device and delay the hit of the single search query to the web) for the purpose of assessment of patentability of computer related inventions. The IPAB thus reiterated that the invention must be examined holistically and the ‘technical effect’ and ‘technical contribution’ are essential factors in deciding the patentability of computer related inventions.
Patentability of diagnostic methods in India
While it is very obvious that inventions ranging from mechanical medical devices, Medical electrical gadgets, synthetic medicinal chemicals, electronic medical devices fall within the ambit of patentable inventions, patentability of diagnostic methods fall under exclusion from patentability if it is practiced on a human or animal body. However, with regards to patentability of diagnostic methods the said provision shall not apply to any of products, particularly substances or compositions used in any of these methods.
Sec 3(i) of Indian Patents Act 1970, reads “( any process for the medicinal, surgical, curative, prophylactic diagnostic, therapeutic or other treatment of human beings or any process for a similar treatment of animals to render them free of disease or to increase their economic value or that of their products.”
Diagnostic methods are defined Manual of Patent Office Practice and Procedure, as “Diagnosis is the identification of the nature of a medical illness, usually by investigating its history and symptoms and by applying tests. Determination of the general physical state of an individual (e.g. a fitness test) is considered to be diagnostic.”
It is imperative to note that a method for detecting an analyte in a biological sample removed from a body may be construed as an in-vitro method if it is supported by the specification and not as a diagnostic method of treatment. Therefore, methods of treatment as recited in Section 3(i) serves as a blanket provision without throwing clarity on screening methods, probability methods or assay of analytes without any connect with diagnosis or treatment. Further, the interpretation of diagnostic methods performed ‘in vivo’ or ‘in vitro’ varies from case to case. The following table gives a glimpse of rejected claims and granted claims by IPO with regards to Section 3(i);
|S.No||Patent Number||Rejected Claims||Granted Claims|
|1||IN257811||A method of treating a neurological disease or disorder in a mammal, which method comprises administering to said mammal in need of such treatment a therapeutically effective amount of a proteasome-based composition||A composition in the form of a medicament for a neurological disease or disorder comprising a proteasome-based composition|
|2||IN279948||A process for isolation of mesenchymal stem cells (MSCs) from a sample||An in-vitro process of isolation of mesenchymal stem cells (MSCs) from a heterogenous cell population|
|3||IN299426||A method for treating a Plasmodium related disease in a subject to prevent, inhibit orameliorate the pathology and/or symptamology of the Plasmodium related disease||A compound as and when used in a medicament for treatment of Plasmodium related disease|
While the AI based technologies are spreading like wildfire, it is very important for India to have a set Guidelines for AI related inventions. In absence of such a policy, it is problematic for inventors for to draft a well-rounded patent specification for grant. As of now, AI based healthcare patent applications are sure to face the bottlenecks as stated above. Hence, keeping in mind the legal requirements, patent specifications should be drafted carefully so as to ensure adequate emphasis of technical effect of the invention and/or technical contribution resulting from the invention.