Diagnosis of Osteosarcoma Through CNN
Developed and trained a convolutional neural network to classify osteosarcoma from histopathology images, reaching roughly 82% accuracy.
- Role
- UCSB SRA Researcher
- Dates
- June 2023 — July 2023
- Capabilities
- Convolutional neural networks · Image classification · Research methodology · Model evaluation · Healthcare AI applications
01
Problem / Opportunity
Osteosarcoma diagnosis from histopathology images is traditionally manual and time-intensive — a well-trained model could help flag classifications faster and support pathologists.
02
Role & Team
Research Assistant (UCSB Summer Research Academy) — model development, training, and evaluation.
EDIT: Add this detail once available.
03
Process & Timeline
Built and trained a convolutional neural network on histopathology image data, iterating on architecture and training parameters to improve classification performance, then evaluated the model against held-out data.
04
Technical Approach
Convolutional neural network trained for binary/multiclass histopathology image classification, evaluated using standard accuracy metrics.
05
Key Decisions
EDIT: Add this detail once available.
06
Challenges & Iteration
EDIT: Add this detail once available.
07
Results & Outcomes
~82% classification accuracy on histopathology image data
Reached approximately 82% classification accuracy.
08
Reflection
EDIT: Add this detail once available.