Introduction
Rohan K -
Hello! My name is Rohan Kshatriya. Welcome to my Senior Project blog. After working on several AI + Biology projects last summer, I discovered the field of computational biology. For my project, I will be researching computational methods for studying laryngeal dystonia (LD), a rare neurological speech disorder. On Friday, February 9th, I will be flying to Boston for 2 months to research LD at the Dystonia and Speech Motor Control Lab at Mass General Brigham.
Laryngeal dystonia is when the laryngeal muscles in the neck spasm, causing a patient’s voice to sound strangled, shaky, or forced. Dr. Kristina Simonyan’s lab at Mass Eye and Ear specializes in researching diagnostic tools and treatments for LD by using several methods: neuroimaging, clinical-behavioral testing, and machine learning. I will be focusing specifically on machine learning and AI as a tool to study speech improvement.
Training an AI to classify a piece of data requires a vast amount of labeled data to learn from. This process is similar to how infants learn the name of an object. Let’s take, for example, dogs. Let’s say you are a toddler seeing a golden retriever for the first time, and someone tells you, “This is a dog.” Immediately, your brain pairs the image (golden retriever) with the label (dog). This would work every time you see a golden retriever, however, dogs are different shapes, sizes, and colors. Your brain needs to see a large variety of animals labeled “dog” or “not dog” before you can learn effective patterns to classify what a dog is by yourself. The same goes for an AI using computer vision. It needs to see thousands of labeled pictures before it can find the important features to classify something as a sitting dog, a moving car, or a stop sign. For my project, I will be using AI to help facilitate the development of treatment options for patients. I am doing some exploratory work with speech recordings in Audacity and Python in the weeks leading up to my site placement.
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.