Search for: All records

IntroductionQuantum computing is increasingly being investigated for integration into medical radiology and healthcare applications worldwide. Given its potential to enhance clinical care and medical research, there is growing interest in evaluating its practical applications in clinical workflows. MethodsWe developed an evaluation of quantum computing-based auto-contouring methods to introduce medical physicists to this emerging technology. We implemented existing quantum algorithms as prototypes tailored for specific quantum hardware, focusing on their application to auto-contouring in medical imaging. The evaluation was performed using a medical resonance imaging (MRI) abdominal dataset, comprising 102 patient scans. ResultsThe quantum algorithms were applied to the dataset and assessed for their potential in auto-contouring tasks. One of the quantum-based auto contouring methods demonstrated conceptual feasibility, practical performance is still limited by current available quantum hardware and scalability constraints. DiscussionOur findings suggest that while quantum computing for auto-contouring shows promise, it remains in its early stages. At present, artificial intelligence-based algorithms continue to be the preferred choice for auto-contouring in treatment planning due to their greater efficiency and accuracy. As quantum hardware and algorithms mature, their integration into clinical workflows may become more viable. 

Abstract Image segmentation of the liver is an important step in treatment planning for liver cancer. However, manual segmentation at a large scale is not practical, leading to increasing reliance on deep learning models to automatically segment the liver. This manuscript develops a generalizable deep learning model to segment the liver on T1-weighted MR images. In particular, three distinct deep learning architectures (nnUNet, PocketNet, Swin UNETR) were considered using data gathered from six geographically different institutions. A total of 819 T1-weighted MR images were gathered from both public and internal sources. Our experiments compared each architecture’s testing performance when trained both intra-institutionally and inter-institutionally. Models trained using nnUNet and its PocketNet variant achieved mean Dice-Sorensen similarity coefficients>0.9 on both intra- and inter-institutional test set data. The performance of these models suggests that nnUNet and PocketNet liver segmentation models trained on a large and diverse collection of T1-weighted MR images would on average achieve good intra-institutional segmentation performance.