Convert 2D CT scan images into 3D models using Hounsfield Units

Abstract

Computed tomography scans, which provide accurate information on the internal organs of the human body, are frequently used in medical imaging. The goal of the proposed study is to create a system that uses Hounsfield units to transform scan data into extremely precise 3D models. This method first segments the CT scans, assigns Hounsfield Units values to the segmented regions, and then builds a 3D model using the values of these units.  The tool will enable healthcare professionals to visualize and look at anatomical structures in three dimensions by utilizing cutting-edge methods and algorithms. In the 3D models as are generated, this method also efficiently recognizes and represents soft tissues. Two models have been studied for this paper.  This proposed work tries to develop 3D models from easy greyscale digital CT scan images of various anatomical structures, and further perform analysis in terms of fracture patterns/fracture anatomy along with soft tissue detection. The primary model is a review of the literature on 3D modelling of x-ray images. To create 3D models from CT scan image, create a machine learning or deep learning model and train it. This model should to be capable of accurately restoring the anatomical structures from the Hounsfield unit values. The process of converting of a CT scan to a 3D model is implemented in the second model. This proposed effort has the potential to revolutionize medical imaging and offer insightful information for surgical planning, therapy, and diagnosis.