Electric Skateboard | Justin Hua

01

SEEING THE VISION

Using machine learning to interpret medical images

The goal was to combine classification and segmentation into a workflow that could organize medical images, distinguish relevant patterns, and visualize tumor regions more clearly

Develop a machine learning workflow that combines image classification with tumor segmentation for medical imaging analysis

OBJECTIVE

Develop a segmentation workflow capable of identifying and visualizing tumor regions from medical imaging data

02

FORMING A PLAN

From image data to prediction and segmentation

The workflow was structured around two main tasks: classifying medical images using machine learning and then analyzing relevant images through segmentation

This required preparing the image data, defining the classification process, and designing a pipeline that could move from prediction to visual segmentation results

Image preprocessing

Machine learning classification

Tumor segmentation

Result visualization

03

MAKING IT HAPPEN

Processing first Learning second

Before applying machine learning, the medical images were processed to prepare them for analysis. Image processing helped transform the raw imaging data into a more consistent and useful input for the classification model

The processed images were then used in the machine learning stage, where the classification model learned patterns from labeled data to distinguish between different image categories

04

THE OUTCOME

From classification to visual insight

The completed project demonstrated how machine learning classification and image segmentation can work together to analyze medical imaging data

It strengthened my understanding of how machine learning can be applied in biomedical engineering, from preparing image datasets and training classification models to interpreting and visualizing computational results

WHAT I LEARNED

Machine learning classification, medical image preprocessing, model evaluation, tumor segmentation, data analysis, and result visualization

TECH STACK

Machine Learning Classification Segmentation Medical Imaging