ESE 499 Capstone Design Project (Spring 2018)

Classification of Prostate Cancer Based on Diffusion MRI Histology Using Machine Learning Algorithms

Summary

Efficient detection and risk assessments of prostate cancer (PCa) are important for the diagnosis and treatment of prostate cancer. With data obtained from patient samples using diffusion [MRI] histology (D-Histo) technique, an advanced MRI technique developed in the Biomedical Magnetic Resonance Laboratory at Mallinckrodt Institute of Radiology, Washington University, I have studied two classification tasks,  differentiating cancerous prostate tissues from benign ones and differentiating high-risk cancer from low-risk ones, and have evaluated their performances.

Main Steps

  • Process and analyze imaging data.

  • Attempt to address two classification problems:

    • Prostate cancer detection:
      classifying cancerous prostate tissues and benign prostate tissues

    • Risk stratification:
      differentiating low-risk and high-risk prostate cancer tissues

  • Evaluate classification performances, analyze results and summarize possible causes of problems encountered.
No page found

Useful Links

No page found
No page found
No page found
No page found