Related Experiment Video
Updated: Dec 13, 2025
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Convolutional neural networks for head and neck tumor segmentation on 7-channel multiparametric MRI: a leave-one-out
Lars Bielak1,2, Nicole Wiedenmann3,4, Arnie Berlin5
1Department of Radiology, Medical Physics, Medical Center University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany. lars.bielak@uniklinik-freiburg.de.
Background:
Automatic tumor segmentation based on Convolutional Neural Networks (CNNs) has shown to be a valuable tool in treatment planning and clinical decision making. We investigate the influence of 7 MRI input channels of a CNN with respect to the segmentation performance of head&neck cancer.
Methods:
Head&neck cancer patients underwent multi-parametric MRI including T2w, pre- and post-contrast T1w, T2*, perfusion (ktrans, ve) and diffusion (ADC) measurements at 3 time points before and during radiochemotherapy. The 7 different MRI contrasts (input channels) and manually defined gross tumor volumes (primary tumor and lymph node metastases) were used to train CNNs for lesion segmentation. A reference CNN with all input channels was compared to individually trained CNNs where one of the input channels was left out to identify which MRI contrast contributes the most to the tumor segmentation task. A statistical analysis was employed to account for random fluctuations in the segmentation performance.
Results:
The CNN segmentation performance scored up to a Dice similarity coefficient (DSC) of 0.65. The network trained without T2* data generally yielded the worst results, with ΔDSCGTV-T = 5.7% for primary tumor and ΔDSCGTV-Ln = 5.8% for lymph node metastases compared to the network containing all input channels. Overall, the ADC input channel showed the least impact on segmentation performance, with ΔDSCGTV-T = 2.4% for primary tumor and ΔDSCGTV-Ln = 2.2% respectively.
Conclusions:
We developed a method to reduce overall scan times in MRI protocols by prioritizing those sequences that add most unique information for the task of automatic tumor segmentation. The optimized CNNs could be used to aid in the definition of the GTVs in radiotherapy planning, and the faster imaging protocols will reduce patient scan times which can increase patient compliance.
Trial Registration:
The trial was registered retrospectively at the German Register for Clinical Studies (DRKS) under register number DRKS00003830 on August 20th, 2015.
Insights
This study shows that T2* MRI sequences are crucial for accurate head and neck cancer segmentation using Convolutional Neural Networks (CNNs). Omitting T2* data significantly reduces performance, suggesting its importance in optimizing CNN-based tumor segmentation protocols.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Oncology
- Radiotherapy Planning
Background:
- Convolutional Neural Networks (CNNs) are valuable for automatic tumor segmentation in cancer treatment planning.
- Investigating the impact of different Magnetic Resonance Imaging (MRI) input channels on CNN segmentation performance for head and neck cancer is essential.
- Multi-parametric MRI data, including T2w, T1w (pre- and post-contrast), T2*, perfusion (ktrans, ve), and diffusion (ADC), were acquired at multiple time points.
Purpose of the Study:
- To evaluate the influence of seven different MRI input channels on the segmentation performance of CNNs for head and neck cancer.
- To identify which specific MRI contrast contributes most significantly to accurate tumor segmentation.
- To optimize MRI protocols for faster patient scans without compromising segmentation accuracy.
Main Methods:
- Head and neck cancer patients underwent multi-parametric MRI scans before and during radiochemotherapy.
- Manually defined gross tumor volumes (primary tumor and lymph node metastases) were used to train CNNs.
- A reference CNN with all seven input channels was compared against CNNs trained with one channel excluded to assess individual channel contributions.
Main Results:
- The CNN segmentation achieved a Dice Similarity Coefficient (DSC) of up to 0.65.
- Excluding T2* data resulted in the poorest segmentation performance, with a decrease in DSC of 5.7% for primary tumors and 5.8% for lymph nodes.
- The Apparent Diffusion Coefficient (ADC) channel had the least impact, with minimal DSC reduction (2.4% for primary tumors, 2.2% for lymph nodes).
Conclusions:
- A method was developed to prioritize MRI sequences that provide the most unique information for automatic tumor segmentation.
- Optimized CNNs can assist in defining Gross Tumor Volumes (GTVs) for radiotherapy planning.
- Reduced patient scan times through optimized MRI protocols can improve patient compliance.
