Research Consent: Bachelor’s Thesis on Machine Learning for Women’s Health
Please read the following information carefully before participating in this study.
1 About the Thesis
I am conducting research for my Bachelor’s Thesis on how to use Machine Learning to analyze images in real-world environments. Specifically, this project focuses on classifying images of vaginal discharge.
Because these images are user-generated (taken at home, not in a lab), my goal is to write a programm that also uses machine learning to identify "useful" data versus poor-quality data.
2 Goal of the Data Collection Period
- Timeframe: from December 1st until January 15th
- Goal: 800 Images in total
- Approx. 50+ Images per Person
- Upload approx. 50% Good Quality Images of vaginal discharge on brown toilet paper or your finger
- Also Upload approx. 50% Blurry, Dark, Bright, Out of Focus and out of context Images
- You can upload multiple Images at once to reach the goal faster
3 How your Identity is protected
- Decoupled Data: Although you must log in to upload, your images and tags are not stored with your username or real name.
- Randomized ID: Upon upload, the system assigns an ID to your data. This allows me to still group your images (to ensure the dataset isn't biased by one user) without attaching your real identity to the medical images when creating the final dataset.
4 What happens to the data?
For the Thesis: I (the researcher) will analyze the images to also train the ML model. My university supervisors will see only the pseudonymized (random ID) data for grading purposes.
Future Research Goal: My long-term hope is to publish an pseudonymized open-source dataset to help other scientists research this understudied topic.
Note: You have the absolute right to request that your data is excluded from any future public release. Simply contacte me directly or via hello@dischargediary.com.
5 Your Rights & Withdrawal
Participation is 100% voluntary.
- Confidentiality: No real names or personal info will ever appear in the thesis or publications.
- Withdrawal: You can stop at any time. If you wish to have your previously uploaded data removed from the study (or excluded from the future public dataset), simply contact me.
6 Consent Agreement
By clicking below and continuing to the app, you confirm that I (the resercher) am allowed to send you a consent document via the email you register your account with.