Cerebral vasospasm is a narrowing of the blood vessels in the brain caused by certain conditions, such as a stroke or a traumatic brain injury. Vasospasms can be easily overlooked when the symptoms are not severe or when the patient does not know to look for them. They can also be overlooked if the patient's medical history and physical examination do not point to the possibility of a vasospasm. If the vasospasm is not severe, it may be difficult to diagnose it without the use of imaging tests such as angiography or Doppler ultrasound. Since radiologists tend to be very busy, and finding and contacting one can be time consuming, the goal of this project was to create a vasospasm detection algorithm using a 2D convolutional neural network (CNN) to predict positive or negative presence of vasospasm for a given CT angiogram with the goal of outperforming radiologists and saving crucial time for both patients and neurosurgeons.