Improving nvAMD treatment course by predicting drug switches

Via Zoom click here

Abstract:

Age-related macular degeneration (AMD) is a multifactorial disease and is the leading cause of visual loss in the developed world. There are four FDA-approved drugs for treating the more severe form of the disease (nvAMD) and a fifth, much less expensive drug that is FDA-approved for treating cancer but is frequently used for treating nvAMD. In many countries, including Israel, it is chosen as a first-line treatment, and a switch to the other more expensive drugs is considered only when it fails. Knowing in advance whether a switch between the inexpensive drug to other more expensive drugs will be required, is essential information for the physician in order to prioritize patients’ care, especially since patients in the first year and a half can still improve their vision quality if given the right treatment. Also, it might help patients set the right expectations for the treatment efficiency. In this work, we predict drugs switches in the first 1.5 years, using several machine learning methods. The prediction is based on the medical textual data of nvAMD patients of Assuta Medical Center. To ease the use of the model as a decision support tool, we applied the SHAP method on the results for local explainability.

Bio:

Rinat Levinstein Gross is an M.Sc. student at the department of Industrial Engineering in Tel Aviv University, specializing in Business Analytics. Rinat holds a B.Sc. degree in Industrial Engineering from Tel Aviv University. Rinat is working as a Product Analyst at Intel.

Her research is supervised by Prof. Irad E. Ben-Gal, Tel Aviv University & Dr. Alon Sela, Ariel University.

Contact:

• E-Mail: Rinat.levinstein@gmail.com