Medical Applications of Artificial Intelligence
Dec. 2023
By Mahathi Karthikeyan
Edited by Mackenzie Twomey
How have you used AI? Nowadays, artificial Intelligence (AI) is used for many purposes including familiar applications such as Alexa or Siri as well as other forms such as facial recognition on iPhones. Surprisingly, AI has also been used in pharmaceutical discovery and journalism. Advanced technology such as this can be applied to make completing certain tasks more efficient. Specifically, an important part of AI called “machine learning” has aided the medical field in an immense nature.
What is “machine learning”?
One of the most common uses for algorithm-specific AI is “machine learning” (Cui, 2022). Machine Learning is a type of artificial intelligence method that relies on algorithms to predict certain results that help in completing certain tasks in a faster manner. Since Machine learning can easily identify patterns in data, it can help predict certain healthcare outcomes using previously determined data. For example, when a patient has a specific disease, it can be difficult for a doctor to predict exactly what type of treatment to use. Using known attributes of the patient and the disease, a more precise treatment protocol can be suggested using an algorithm which will train the dataset to recognize patterns for specific diseases. This has the potential to greatly improve patient diagnosis and shorten the time it takes for the patient to get diagnosed as the algorithm can help narrow down potential treatments based on patient variables.
AI in Image Analysis
Another way that AI can assist is through image analysis, using a subset of machine learning called “deep learning.” Deep learning is similar to machine learning; however, it uses artificial neural networks instead of previously determined data, which analyze patterns similar to the human brain to come up with solutions to common problems. Neural networks can identify patterns faster than other forms of AI which help identify certain key characteristics in images.
Such a method is very useful in the context of image processing. Since deep learning can easily identify patterns in neural networks, certain elements of these images, such as tumors, can be detected and potentially extracted to be viewed separately and analyzed through a process called “image segmentation.” Image segmentation helps doctors focus on one area of an image and ignore the parts of the image that are not important or do not hold potential harm to the body. This form of analysis can help doctors more efficiently figure out patient diagnosis since the identifiable part of the image is extracted through AI. A good example of how this method is used is in stroke diagnosis. A stroke is a disease that must be diagnosed very quickly, as early diagnosis is key to making at least partial recovery. Deep learning algorithms can help isolate the part of the brain that is most affected by stroke, called lesions, in CT and MRI images which can aid doctors in identifying damage and assessing the best treatment plans for the patient. These methods can also be used to help with image denoising. Oftentimes, MRI images can be very “noisy” (described as images that are too grainy) and any attempt to denoise them while the patient is inside the machine can be uncomfortable. Using AI, the noisy images can be denoised in order to increase the quality and clear up the MRI image.
In essence, AI in medicine has been shown to be a massive benefit to physicians and patients. Using algorithms, tedious and time-consuming procedures, and practices are shortened while also increasing physician accuracy in patient diagnosis. Even though AI in medicine may not be a solution to every healthcare problem, as there are many “ethics, privacy and reliability concerns” it is a massive help to the field (Harvard Medical School, 2023).
References:
Cui, L., Fan, Z., Yang, Y., Liu, R., Wang, D., Feng, Y., Lu, J., & Fan, Y. (2022, November 14). Deep learning in ischemic stroke imaging analysis: A comprehensive review. BioMed research international. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9678444/. Accessed 10/27/23.
Ai in Medicine. NEJM- New England Journal of Medicine. https://www.nejm.org/ai-in-medicine. Accessed 10/28/23.
“How Artificial Intelligence Is Disrupting Medicine and What It Means for Physicians.” HMS Postgraduate Education, 13 Apr. 2023, https://www.postgraduateeducation.hms.harvard.edu/trends-medicine/how-artificial-intelligence-disrupting-medicine-what-means-physicians.
