Impact of deep learning-based multiorgan segmentation methods on patient-specific internal dosimetry in PET/CT imaging: A comparative study

Abstract

Purpose: Accurate and fast multiorgan segmentation is essential in image-based internal dosimetry in nuclear medicine. While conventional manual PET image segmentation is widely used, it suffers from both being time-consuming as well as subject to human error. This study exploited 2D and 3D deep learning (DL) models. Key organs in the trunk of the body were segmented and then used as a reference for networks.

Methods: The pre-trained p2p-U-Net-GAN and HighRes3D architectures were fine-tuned with PET-only images as inputs. Additionally, the HighRes3D model was alternatively trained with PET/CT images. Evaluation metrics such as sensitivity (SEN), specificity (SPC), intersection over union (IoU), and Dice scores were considered to assess the performance of the networks. The impact of DL-assisted PET image segmentation methods was further assessed using the Monte Carlo (MC)-derived S-values to be used for internal dosimetry.

Results: A fair comparison with manual low-dose CT-aided segmentation of the PET images was also conducted. Although both 2D and 3D models performed well, the HighRes3D offers superior performance with Dice scores higher than 0.90. Key evaluation metrics such as SEN, SPC, and IoU vary between 0.89-0.93, 0.98-0.99, and 0.87-0.89 intervals, respectively, indicating the encouraging performance of the models. The percentage differences between the manual and DL segmentation methods in the calculated S-values varied between 0.1% and 6% with a maximum attributed to the stomach.

Conclusion: The findings prove while the incorporation of anatomical information provided by the CT data offers superior performance in terms of Dice score, the performance of HighRes3D remains comparable without the extra CT channel. It is concluded that both proposed DL-based methods provide automated and fast segmentation of whole-body PET/CT images with promising evaluation metrics. Between them, the HighRes3D is more pronounced by providing better performance and can therefore be the method of choice for 18F-FDG-PET image segmentation.

Keywords: PET/CT; deep learning; internal dosimetry; segmentation.

FIGURE 1

General workflow of the present study to compare 2D U‐Net GAN and HighRes3D_PET for PET image segmentation.

FIGURE 2

The segmented organs using thresholding, region growing, and level tracing methods.

FIGURE 3

The paring of input (PET images) with output (labeled images) for both 2D and 3D networks. The middle row represents the corresponding low‐dose CT images.

FIGURE 4

Sketch of the p2p‐U‐Net‐GAN used in the present study to segment PET/CT images.

FIGURE 5

Sketch of the HighRes3D_PET architecture used in the present study to segment PET images.

FIGURE 6

Representative segmentation results of one subject from the test dataset.

FIGURE 7

The difference map of p2p‐U‐Net‐GAN (2D) and HighRes3D_PET (3D) models.

FIGURE 8

Representation of the inaccurate area of the bladder in generated slices.

FIGURE 9

The Bland–Altman plots of spleen volume for the ten patients for the HighRes3D_PET and p2p‐U‐Net‐GAN architectures. The regression line has been also plotted in each graph.

FIGURE 10

The comparison of the different organ's S‐values with manual and DL segmentation methods.