Deep Learning-based Auto-segmentation on CT and MRI for Abdominal Structures

Abstract

When asked 'How would you rate your transportation experience today?' 82% responded Above Average. To the question 'Would you have been able to attend your appointment today if this program did not exist?' 92% answered No. Conclusion: This study shows that the cost of rideshare transportation can be significantly less than the cost of no-shows. This suggests that a proactive virtual transportation hub can help address transportation barriers, drive patient satisfaction and reduce the waste of no-shows. Radiation therapy represents an ambulatory medicine crucible for patients with limited transportation and social support. Scaling up rideshare innovations from radiation oncology has the potential to drive broader ambulatory strategy.

Automatic Contour Refinement for Deep Learning Auto-segmentation of Complex Organs in MRI-guided Adaptive Radiation Therapy

“…Bobo et al 7 described their CNN for the multiorgan segmentation on abdominal MRI, with DSCs of 0.556 and 0.691 in the stomach and pancreas. Chen et al 8 and Amjad et al 9 reported similar results from their DL auto-segmentation solutions. These previous studies indicate that the DL auto-segmentation on MRI for complex organs (eg, bowels) is generally unacceptable for clinical use.…”

“…Although DL-based segmentation methods have enabled the organ auto-contouring and achieved great success in many clinical applications, the current auto-segmented contours of challenging organs, such as the bowels, can still be clinically unacceptable. 6 , 7 , 8 , 9 Inevitably, manual editing needs to be performed subsequently to make the contours acceptable. The manual editing is generally time-consuming and labor-intensive, especially for inaccurate contours with irregular shapes (eg, complex abdominal organs).…”

“…In recent years, deep learning (DL) techniques, particularly the convolutional neural networks (CNNs), have been successfully applied to automatically segment organs from medical images including MRI. 6 , 7 , 8 , 9 , 10 , 11 Although it has been well documented that such auto-segmentations can be more efficient compared with the time-consuming and labor-intensive manual delineations, available MRI-based auto-segmentation methods still have limited success, particularly for complex structures such as those in the abdomen. Fu et al 6 reported their DL auto-segmentation on MRI in the abdomen achieved average Dice similarity coefficients (DSCs) of 0.953, 0.931, 0.850, 0.866, and 0.655 for the liver, kidneys, stomach, bowels, and duodenum, respectively.…”

A deep learning model based on the attention mechanism for automatic segmentation of abdominal muscle and fat for body composition assessment

“…The proposed model can overcome such shortcomings by increasing the variability and heterogeneity of the training images, as well as by including images from a wider range of institutions. We used CT imaging to develop a deep learning model instead of using magnetic resonance imaging (MRI) or a combination of both, which might be more effective (46). The segmentation of small targets also remains a class of challenges to be solved.…”

Multi‐parametric magnetic resonance imaging for radiation treatment planning

“…Additionally, DL models are being developed for fast segmentation of complex organs such as abdominal organs. The first DL model employing a voxel-wise label prediction CNN in conjunction with a correction network to segment liver, kidneys, stomach, bowel, and duodenum on TrueFISP 3D MRI was presented by Fu et al 62 For innovative techniques such as adaptive RT, multilayer learning including self -adaptive, active learning classification algorithm (SAALC) on T1w trueFISP, 63 DL architecture with short-and long-range residue and U-Net skip connection on T1w Dixon and T2w HASTE, 64 and 2D U-net and dense net on T1 VIBE MRIs, 65 have been proposed to deliver fast segmentation with high accuracy (dice > 0.8) for most of the abdominal organs.…”