About

This is the website for the STAT390 C-MIL Classification project! STAT390 is the Data Science Project course, the last required course for Data Science majors at Northwestern University. Students of this course are currently working on developing a classification model to classify a C-MIL lesion as benign, low-grade or high-grade.

People

Stakeholders:

Dr. Yamini Krishna
  Consultant Ophthalmic Pathologist,
  Honorary Senior Clinical Lecturer (Eye & Vision Sciences),
  Royal Liverpool University Hospital & University of Liverpool,
  Liverpool, UK
Dr. He Zhao
  Lecturer (Eye & Vision Sciences),
  University of Liverpool,
  Liverpool, UK.

Project Co-ordinator (STAT390):

Dr. Arvind Krishna (Krish)
 Assistant Professor of Instruction,
 Department of Statistics & Data Science,
 Northwestern University,
 Evanston, IL, USA.

Collaborator:

Prashant Kumar
 Staff Engineer,
 Qualcomm,
 San Diego, CA, USA.

Student Teams (STAT390):

  • Thinkers: John Olsen, Kota Suzuki, Yaelle Pierre, Jake Mead, Ryan Yi, Anna Roney
  • Coders: Sari Eisen, Sharon Lin, Ryan Lach, Nathan Jung, Hannah Ma
  • Literature Surveyors: William Wang, Bennett Markinson, Allen Zhang, Alex Zhou, Jacob Muriel, Patrick Schmid, Lainey Neild
  • Consultants/Organizers/Testers: Olivia Joung, Anna Deka, Margaret Pirozzolo, Jack Ouyang, Haneef Usmani, Radhika Todi
  • Health Informaticians: Lucinda Hu, Dila Bitlis, Julia Nelson, Walker Frisbie, Kayla Terrelonge

GitHub and OneDrive

This GitHub contains all the necessary code and algorithms from our progress:

This OneDrive contains all data and processed folders from our progress:

Data Overview

The Data consists of 105 cases from 97 patients. Below are some statistics of the data:

  • 60 women (age range: 23-93 years; median 65; mean 61.5) and 37 men (age range: 31-91 years; median 68; mean 66.5)

  • The ethnic group mix comprised of the following: 88 White Caucasian, 4 Black, 2 South Asian, 1 Inuit, 1 Mixed race, and 1 Unspecified

The data corresponds to patients in the following 3 ocular oncology/pathology centers:

  • Liverpool University Hospitals NHS Foundation Trust (Liverpool; cases from 2018 to 2021),
  • Royal Hallamshire Hospital (Sheffield; from 2011 to 2021), and
  • Rigshospitalet (Copenhagen; from 1996 to 2021)

For each patient, a tissue is taken from their conjunctiva. The tissue is sliced into multiple slices. Multiple slices are used so that each slice can be analyzed separately and there is more evidence to support the conclusion, i.e., the classification of the C-MIL.