Automated real time fetal ultrasound image classification and fetal biometry

This is a research project which predicts the diseases and other abnormalities of a fetus, but when the state-of-the-art is considered, the improvement of the accuracy is preferred. And since there is only a model is made but no interface, so our focus is to build such an interface in which the imag

Project Title

Automated real time fetal ultrasound image classification and fetal biometry

Project Area of Specialization

Artificial Intelligence

Project Summary

This is a research project which predicts the diseases and other abnormalities of a fetus, but when the state-of-the-art is considered, the improvement of the accuracy is preferred. And since there is only a model is made but no interface, so our focus is to build such an interface in which the image or video will be uploaded and the desired output would be displayed.

Project Objectives

Ultrasound-based fetal head biometrics measurement is a key indicator in monitoring the conditions of fetuses. Since manual measurement of relevant anatomical structures of fetal head is time consuming and subject to inter-observer variability, there has been strong interest in finding automated, robust accurate and reliable method. Here, we propose a deep learning based method to segment fetal head from ultrasound images. The proposed method formulates the detection of fetal head boundary as a combined object localization and segmentation problem based on deep learning model. Incorporating an object localization in a framework developed for segmentation purpose aims to improve the segmentation accuracy achieved by fully convolutional network. Finally, ellipse is fitted on the contour to the segmented fetal head using hough ellipse fitting method.

Project Implementation Method

The process will comprise of around 20-25 training videos, and 7-9 testing videos of ultrasound. Videos recorded would be in 1st trimester phase. The first task would be to split the video in such a way so that only the head section is retrieved. Then the video would be split into its frames. Then the frames would be classified as either they are head or not, and since there would be more than enough frames, no augmentation of the data would be needed. After that, localization (bounding boxes around the head) would be applied on the head section and segmentation with ellipse fitting on the head would be performed, which would yield the radius of the head which will help us to calculate/predict the gestational age of the baby.

Benefits of the Project

It will help to predict the gestational age of the fetus which is not so accuract if calculated manually. So the automated system will be built for it. And in manual process, there is a high probability from sonographers to make human-errors which could easily predicted the wrong gestational age. And when the ellipse has to be drawn around the head of the fetus, sonographers have to stay still and draw the ellipse manually very carefull, which is highly prone to erros, but the system we will built will do such tasks automated. This will also help the sonographers to refrain from back strain. 

Technical Details of Final Deliverable

We will build the full pledged software which will predict the gestational age of a fetus on the given input in the form of video or image. The system will be build on Python and will be downloadable, and can be supplied to the hospitals which urge to such automated tasks.

Final Deliverable of the Project

Software System

Core Industry

Medical

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com