4 minute read

AI in Precision Medicine

In recent years, precision medicine has started to become the new standard when it comes to healthcare. This is moving us from a one size fits all approach to a more personal, data-driven approach that allows hospitals and treatment centers to spend more efficiently and have a higher patient outcome. Precision medicine is using knowledge that is specific to one patient, such as biomarkers, rather than the generic approach to an issue. The overall goal is to "design and optimize the pathway for diagnosis and prognosis through the use of large multidimensional biological datasets that capture different variables such as genes" [1].

Artificial intelligence (AI) has been increasingly growing in business, society and now is emerging in healthcare. The potential that AI has can completely transform patient care. These technologies can perform to or exceed human capability when it comes to different medical tasks such as cancer diagnosis or disease diagnosis as well as patient engagement and administration tasks. AI has the potential to offer automated care to individuals by providing precision medicine.

Precision medicine enables patients to not only recover from illnesses faster but to also stay healthy longer. However, with the increased use of precision medicine new challenges arise such as the increasing amount of data, a lack of specialists and ever increasing drug development costs. "Healthcare data is projected to grow by 43 percent by 2020, to roughly 2.3 zettabytes. The size of the data is not the only problem; it's the kind of data as well. Eighty percent of it is unstructured and mostly unlabeled, making it hard to extract value from the datasets" [2].

Artificial intelligence (AI) has helped reshape how precision medicine is distributed. AI is able to solve many of the problems that have arisen. For big data challenges, AI methods are able to clear up obstacles that large and unstructured data present. In medical imaging, machine learning can be introduced to help classify what type of issue is present by training a model over thousands of images and predicting on the patient's image. Neural networks have also been able to make predictions when it comes to precision medicine.

Neural networks are a more advanced form of AI. The uses in precision medicine is for categorisation applications such as the likelihood of a patient developing a disease. Neural networks look at problems from inputs, outputs, and weights of features to try and associate inputs with the corresponding outputs. "It has been likened to the way that neurons process signals, but the analogy to the brain's function is relatively weak" [3].

Deep learning is one of the most complex forms of AI. This involves hundreds or thousands of models with numerous levels of features that are needed to predict the outcomes. Precision medicine takes advantage of this technology through the "recognition of potentially cancerous lesions in radiology images" [4]. Deep learning is able to be applied to fields such as radiomics. This is the practice of detecting features in image data that cannot be detected with the human eye. "Their combination appears to promise greater accuracy in diagnosis than the previous generation of automated tools for image analysis, known as computer-aided detection or CAD" [4].

AI plays a pivotal role in the future of healthcare. In the development of precision medicine, it is one of the primary components in order to advance care for patients. Efforts to help classify medical imagery more quickly and accurately have proven more effective with the amount of data used to train such models. A big challenge that AI is facing in precision medicine is whether or not this technology will be widely adopted. These systems will need to have some regulations in order to have a universal standard. This will allow doctors and medical personnel to train with this technology so they will be able to provide the care their patients deserve. AI will never replace the human aspect of precision medicine but over time AI will be able to make the jobs and lives of the doctors and patients better and healthier.

References

[1] M. Uddin, Y. Wang, and M. Woodbury-Smith, "Artificial intelligence for precision medicine in neurodevelopmental disorders," Nature News, 21-Nov-2019. [Online]. Available: https://www.nature.com/articles/s41746-019-0191-0. [Accessed: 11-Oct-2020].

[2] H. Chamraj, "Powering Precision Medicine with Artificial Intelligence," Intel. [Online]. Available: https://www.intel.com/content/www/us/en/artificial-intelligence/posts/powering-precision-medicine-artificial-intelligence.html. [Accessed: 12-Oct-2020].

[3] T. Davenport and R. Kalakota, "The potential for artificial intelligence in healthcare," Future healthcare journal, Jun-2019. [Online]. Available: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6616181/. [Accessed: 12-Oct-2020].