Top 3 ways AI/ML impacts Drug Delivery Systems – A Strategic Outlook on Precision Medicine
Artificial Intelligence, Machine Learning, and Healthcare are becoming more and more intertwined as of late. The rapid development of AI/ML has begun to take root in the medical device/drug delivery space during the last few years. The idea and possibility of a deeply integrated, artificial intelligence driven healthcare engine capable of quickly assessing human health, may not just be fiction found in movies anymore. As more advances are made in machine and deep learning, the more regulations are deemed necessary. The general gist was that AI/ML could serve as an accessory used to help monitor patient progress, diagnose diseases, or simply help schedule appointments. However, with the ability to evolve over time via machine/deep learning, it looks like artificial intelligence will be playing a more integrated role than first foretold. It is widely known that the most advancement of artificial intelligence in healthcare is within the practice of radiology, however the niche industry of drug delivery systems is quietly and quickly developing.
1. SaMD
To shed light on AI/ML in the medical device industry, we turn to software as a medical device. The primary vehicle for artificial intelligence/machine learning in the medical device industry. Software as a medical device, as defined by the FDA, is “software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device”, also referred to as “health software”. The main focal points assessed in determining SaMD functionality/usefulness are the following:
· The ability to communicate the significance of the information provided by the SaMD to end users
· The ability to treat or diagnose.
· The ability to manage/drive clinical decisions (inputs)
· A definition statement accurately describing the use of the software in an organized framework
2. AI in Drug Delivery Devices
Drug delivery devices are pumps, injectables, or medical machines that facilitate the delivery of medicine/drugs through the body in order to achieve a desired therapeutic outcome. The application of artificial intelligence and machine learning within medical devices, is most found in the form of artificial neural networks (ANNs). ANNs are computational models that are made up of several processing elements that intake inputs and distribute outputs based on predetermined parameters. Artificial neural networks (neural nets) also strive to imitate the brain, as they are structured as a collection of connected units (nodes). The activity of having neural nets process large data in a configuration like the brain is a machine learning process (subset) known as deep learning. As AI continues to learn and advance through the processing of large data sets, the possibilities for applications seem to become boundless.
3. New AI Applications
One such example is the use of RAVEN, an automated contextualization and OEE software that has partnered with Evaxion and ExpreS2ion to co-develop a new cytomegalovirus vaccine candidate. This specific AI will aid in the designing of the antigen constructs that will be used to induce cellular and humoral/antibody responses. Factors such as rapid drug action and absorption, low drug degradation by gastric secretion, and less concentration are all attributes that can be significantly improved with AI. Tools like Quantitative structure-activity relationship (QSAR)- based computational model are still facing challenges in correctly deciphering complex biological properties such as the efficacy of compounds.
Nanorobots and advanced microrobots, which consist of integrated circuits, sensors, and power supplies are maintained using advanced artificial based technologies. These micro-sized drug delivery systems can be administered via oral, parenteral, nasal, or topical routes. In 2021, Hanlu Gao et al were able to use a random forest algorithm to generate a classification model to accurately distinguish between two dissolution profiles of a drug (spring-and-parachute vs maintain supersaturation) at 85% sensitivity, 85% accuracy, and 85% specificity in 5-fold cross validation. The random forest algorithm was able to predict the time-dependent total drug release within a mean absolute error of 7.78 in 5-fold cross validation as well. This achievement using AI/ML provides an opportunity to overcome the challenge of targeting multiple receptors using nanocarriers as drug delivery systems.
AI/ML is also present in what are now called “smart infusion pumps”. Baxter International Inc. has recently announced an innovative study finding that machine learning shows an enormous amount of potential in enhancing patient safety when using infusion pumps. Using the software’s ability to build, maintain, and process drug libraries will allow the smart infusion pump to review infusion sessions and call out possible errors through a system called Dose Error Reduction Systems. Although this technology is still under development, it still invokes excitement when thinking of the heights that AI/ML can bring healthcare to.
It’s thrilling to see the advancement of healthcare come this far in just four years; and as these technologies advance, so will the regulation and quality requirements. Quality Means Business will be here to help facilitate the launch of these critical advancements, helping to ensure a better quality of life for all.
Endnotes
1. Alshawwa, S.Z.; Kassem, A.A.; Farid, R.M.; Mostafa, S.K.; Labib, G.S. Nanocarrier Drug Delivery Systems: Characterization, Limitations, Future Perspectives and Implementation of Artificial Intelligence. Pharmaceutics 2022, 14, 883. https://doi.org/10.3390/pharmaceutics14040883
2. “Evaxion and ExpreS²ion Partner to Develop New CMV Vaccine.” Pharmaceutical Technology, 6 Dec. 2022, https://www.pharmaceutical-technology.com/news/evaxion-expres²ion-cmv-vaccine/. Accessed 2 Feb. 2023.
3. Jiang, J.; Ma, X.; Ouyang, D.; Williams, R.O., III. Emerging Artificial Intelligence (AI) Technologies Used in the Development of Solid Dosage Forms. Pharmaceutics 2022, 14, 2257. https://doi.org/10.3390/pharmaceutics14112257 https://www.mdpi.com/1999-4923/14/11/2257
4. Jonathan Zaslavsky, Pauric Bannigan & Christine Allen (2023) Re-envisioning the design of nanomedicines: harnessing automation and artificial intelligence, Expert Opinion on Drug Delivery,20:2, 241-257, DOI: 10.1080/17425247.2023.2167978
5. “PART I: AI and Machine Learning in Drug Delivery - Next Big Thing in Industry?” AAPS Magazine, Oct. 2022, https://www.aapsnewsmagazine.org/aapsnewsmagazine/articles/2022/oct22/cover-story-oct22b. Accessed 2 Feb. 2023.
6. Paul D, Sanap G, Shenoy S, Kalyane D, Kalia K, Tekade RK. Artificial intelligence in drug discovery and development. Drug Discov Today. 2021 Jan;26(1):80-93. doi: 10.1016/j.drudis.2020.10.010. Epub 2020 Oct 21. PMID: 33099022; PMCID: PMC7577280.