In Korea, using a clinical decision support system (CDSS) has many limitations:
① The treatment that CDSS recommended was not suitable for Korea's national health insurance system.
② Black-box systems made clinicians hard to believe the results of CDSS's recommendation.
Therefore, this project aims to develop a CDSS based on eXplainable AI using Gill medical center's clinical data.
Duties:
• Analyzed clinical data of 3381 colorectal cancer patients, and proposed a way to improve CDSS.
• Discovered the prognostic factors of colorectal cancer patients through medical statistics.
• Proved important factors that affect the selection of chemotherapy through machine learning models.
Mental illness such as ADHD, Depression, and Alzheimer's is not easy to detect before become serious.
This project aims to develop a brain disease prediction model by using medical data.
Duties:
• Supervised an experiment for extracting 20 subjects’ brainwaves during meditation and concentration.
• Served as a research assistant for calculating concentration from brain waves.
• Investigated the impact between brain waves and Binaural beats, providing different frequency sound to each ear.
• Designed a classification model between patients diagnosed with depression and a healthy control group.
Before recruiting participants for clinical trials, it is essential to check the medication taken by the patient.
This project aims to develop a pill recognition system from photos and creates a chatbot that evaluates whether patients are appropriate for participating in clinical trials.
Duties:
• Collected images of high blood pressure & diabetes pills for developing a pill-recognition system.
• Assisted in developing a chatbot with the pill-recognition system.