Personal Project Update #1

Problem:

Patients with eczema suffer from it on a daily basis. Most importantly, they do not know how to effectively treat it. My project aims to incorporate both Machine Learning (ML) and Artificial Intelligence (AI) Models to help patients better diagnose and manage AD on a daily basis.

Research Completed:

Articles Read:

1. https://my.clevelandclinic.org/health/diseases/9998-eczema
2. https://pmc.ncbi.nlm.nih.gov/articles/PMC11983576/
3. https://pmc.ncbi.nlm.nih.gov/articles/PMC4773205/

Projected Solution:

I will build a machine learning model to evaluate images to diagnose them with either “eczema” or “normal”. This will be done in Jupyter Notebook, Tensor flow. I will feed it extensive amounts of data that will act as a “train and test” set. After that, I will evaluate the machine against my definition statement by providing a different set of images called the validation set.

Progress Made:

1. Finished code for 3 models to compare and contrast results: Tensorflow, Random Forest, and SVM (Support Vector Machine)
2. Finished overall code for validation set (will print images and the accuracy of the machine)
3. Gathered a dataset of approximately 300 images.
4. Effectively utilized a large, pre built dataset to increase accuracy

Challenges:

1. Code not working as intended
2. Lack of data, decreases accuracy overall
3. Ethical concerns with pictures involved

Next Steps:

1. Gather more data (pictures)
2. Decide on a final model to use for production
3. Learn more about eczema by interacting with authentic dermatologists and specialists

AI Statement:

No AI was used in the production of this blog post, everything here was produced by myself.