Machine Learning-based Fertility Prediction Tool

2022 COE Engineering Design Project (RP01)

---

Faculty Lab Coordinator

Robnier Reyes Perez

Topic Category

Software Systems

Preamble

In-vitro fertilization (IVF) has become very popular since the first successful case of a live birth was announced in the late 70’s. Subsequent improvements in assisted reproduction technology (ART) led to the use of follicle-stimulating hormone and the development of techniques such as intracytoplasmic sperm injection (ICSI). However, the female reproductive cycle is sensitive to external factors and patient-specific fertility is difficult to evaluate. A robust fertility prediction tool can serve both the patient and the specialist by performing a fast and accurate assessment of the probabilities of having a live birth over a number of cycles.

Objective

The goal of this project is to use an open-source IVF database to develop a machine learning-based fertility prediction tool. This tool will take the form of a smartphone application that can receive patient data and output live birth success rate as a function of IVF cycles.

Partial Specifications

- The patient’s and partner’s data must both be used in the fertility evaluation.
- Open-source fertilization databases must be used for the machine learning prediction tool.
- The fertility prediction tool must be in the form of a smartphone application.

Suggested Approach

- Find available open-source fertility database
- Use a software package such as MATLAB for the machine learning component.
- Develop an Android App with an intuitive graphical user interface (GUI) to query the patient and display results

Group Responsibilities

All team members are responsible for performing a literature review of the current state of machine learning based fertility prediction. This includes but it is not limited to finding peer-reviewed literature, patented solutions, and available solutions. The team is responsible for creating a block level design of the complete system and assigning specific tasks to each of the blocks. Every member is responsible for documenting the design of their subsystem, explaining design choices, and description of the performed tasks.

Student A Responsibilities

This team member will be responsible for finding open-source fertility databases and selecting the relevant variables to use for prediction. Additionally, this student will work in close collaboration with Student D to develop the Android App. Additional responsibilities as assigned by the FLC.

Student B Responsibilities

This team member will be responsible for using software packages such as MATLAB Statistical and Machine learning Toolbox and MATLAB Neural Network Toolbox to develop the prediction tool. Additional responsibilities as assigned by the FLC.

Student C Responsibilities

This team member will be responsible for using software packages such as MATLAB Statistical and Machine learning Toolbox and MATLAB Neural Network Toolbox to develop the prediction tool. Student B and C will work in close collaboration to achieve this goal. Additional responsibilities as assigned by the FLC.

Student D Responsibilities

This student will develop the graphical user interface (GUI) for the Android App which the patient will use to answer medical questions and obtain their fertility prediction as a function of IVF cycles. Additional responsibilities as assigned by the FLC.

Course Co-requisites

COE318, COE428, COE528

To ALL EDP Students

Due to COVID-19 pandemic, in the event University is not open for in-class/in-lab activities during the Winter term, your EDP topic specifications, requirements, implementations, and assessment methods will be adjusted by your FLCs at their discretion.