Index

A disentangled representation strategy to enhance multi-organ segmentation in CT using multiple datasets

Medical image segmentation is important for identifying human organs, essential in clinical diagnosis and treatment planning.However, the accuracy of segmentation results is often compromised due to the limited quality and completeness of medical imaging data. In practical applications, deep learning has become a key method for multiorgan segmentation[1, 3], but it struggles with challenges related to the amount and quality of data.Deep learning segmentation models typically require numerous paired images and annotations for training[2]. However, fully annotated multi-organ CT datasets are rare, while those annotating only a few organs are more frequent. The variation in annotations restricts the efficient utilization of numerous public segmentation datasets. Inspired by disentangled learning’s ability to share knowledge across tasks[4, 5, 6], we’ve developed a method that allows models to learn and incorporate features from different datasets. We attempt to combine two types of datasets: one fully annotated for multiple organs but with a small amount of data, and another larger dataset annotated only for certain organs.This method is designed to improve the model’s capability in segmenting multiple organs.Using disentangled learning, the model is able to extract and combine crucial features from various datasets, thus overcoming the challenge of inconsistent annotations. This method aims to enhance the model’s adaptability and precision. We assess its performance by comparing the model’s predicted segmentations with actual annotations, allowing for a detailed evaluation of using the disentangled learning approach versus models trained with only a single dataset in multi-organ segmentation tasks. To summarize, the thesis will cover the following aspects:

Design a multi-organ segmentation model using disentangled learning methods.
Investigate the influence of the quantity of fused datasets on the multiorgan segmentation model.
Investigate the influence of the proportion of data quantity from different datasets on the multi-organ segmentation model.
Investigate the influence of feature weights from different datasets on the multi-organ segmentation model.

References
[1] Yabo Fu, Yang Lei, TongheWang, Walter J. Curran, Tian Liu, and Xiaofeng Yang. A review of deep learning based methods for medical image multiorgan segmentation. Physica Medica, 85:107–122, 2021.
[2] Tianxing He, Shengcheng Yu, Ziyuan Wang, Jieqiong Li, and Zhenyu Chen. From data quality to model quality: an exploratory study on deep learning, 2019.
[3] Yang Lei, Yabo Fu, Tonghe Wang, Richard L. J. Qiu, Walter J. Curran, Tian Liu, and Xiaofeng Yang. Deep learning in multi-organ segmentation, 2020.
[4] Yuanyuan Lyu, Haofu Liao, Heqin Zhu, and S. Kevin Zhou. A3dsegnet: Anatomy-aware artifact disentanglement and segmentation network for unpaired segmentation, artifact reduction, and modality translation, 2021.
[5] Qiushi Yang, Xiaoqing Guo, Zhen Chen, Peter Y. M. Woo, and Yixuan Yuan. D2-net: Dual disentanglement network for brain tumor segmentation with missing modalities. IEEE Transactions on Medical Imaging, 41(10):2953–2964, 2022.
[6] Tongxue Zhou, Su Ruan, and St´ephane Canu. A review: Deep learning for medical image segmentation using multi-modality fusion. Array, 3-4:100004, 2019.

Defect Detection Probability as a Metric for CT Image Quality Assessment

This project focuses on using defect detection probability within CT (Computed Tomography) images as a metric for assessing image quality. Key steps include:

Establishing a data preparation pipeline to insert defects into CT volumes sourced from CAD files.
Simulating CT scans to replicate imaging processes.
Developing a defect detection neural network to analyze CT images and determine the probability of defect presence.
Utilizing the defect detection probability as a quantitative metric for evaluating the quality of CT images, with potential integration of trajectory optimization techniques.

Automated ONNX2TikZ: Generating LaTeX-TikZ Diagrams of Neural Networks

This project aims to automate the conversion of ONNX models into TikZ code, facilitating the creation of visually appealing diagrams in LaTeX documents. Leveraging Python for ONNX parsing and manipulation, alongside LaTeX and TikZ for rendering, this tool streamlines the process of visualizing neural network architectures for academic papers, presentations, and educational materials

Contrastive Learning of 3D Objects via Patch-Level Point Cloud Encoding for Similarity Matching

Review of Zero-shot, Few-shot classification, detection and segmentation methods in Medical Imaging

Review of Zero-shot, Few shot classification, detection and segmentation methods in medical imaging.

Evaluation of MedKLIP for Zero-shot and Fine-tuned classification of CXRs

Zero-shot scores on NIH and RSNA Pneumonia datasets. Analysis of attention maps and point score on VinDR-CXR dataset. Analysis of performance improvement from zero-shot to fine-tuned classification performance for various findings.

Center-to-Peer Federated Learning Research

Generating High-Resolution CT Images via Score-Based Diffusion and Super-Resolution Techniques

Partial Convolution for CT Field of View Extension

Diffusion Model-Enabled Energy Level Transformation in Photon Counting Computed Tomography (PCCT)

Introduction:

Photon counting computed tomography (PCCT) marks a new era in medical imaging, offering an unprecedented ability to discriminate between different photon energy levels. This feature of PCCT is crucial for enhancing image contrast and specificity, allowing for more accurate tissue characterization. However, efficiently managing and converting between these diverse energy levels in a clinically practical manner remains a significant challenge.

This project aims to utilize diffusion model to streamline and optimize the energy level conversion process in PCCT. By leveraging the advanced pattern recognition and computational capabilities of diffusion model, the project intends to develop a system that can automatically and accurately translate between different photon energy levels, enhancing the utility and clarity of PCCT images.

The ultimate goal is to provide a robust and efficient framework that not only improves the diagnostic quality of PCCT images but also expands the practical applications of this technology in clinical settings. This involves intricate work in both the development of diffusion model and the understanding of the physics underlying PCCT.

Requirements:

Completion of Deep Learning is mandatory.
Proficiency in PyTorch is essential.
Strong analytical and problem-solving skills.

~~Prospective candidates are warmly invited to send their CV and transcript to yipeng.sun@fau.de.~~