Contour+ performance on Male Pelvis synthetic CTs

Results of MVision AI’s solutions scientific evaluation were presented at the previous ASTRO annual meeting. One of the posters included data from paired MRI and synthetic CT scans (MRCAT) of male pelvis and showed the accuracy and consistency of Contour+ models. The abstract was published in  Volume 117 Issue 2 Supplement of the international Journal of Radiation Oncology *Biology* Physics and can be found here.

Nowadays MRI scans are used for target delineation and the CTs are needed for planning. However, an MR-only workflow using synthetic CTs (MRCAT) is expected to simplify the process, improve delineation’s accuracy and therefore reduce the risk of systematic errors.

We are adding the full poster content below, with the authors’ approval.

(Text and images copyright retained by the authors)

Clinical Evaluation of Deep Learning-Based Auto-Segmentation toward MRI-Only Prostate Radiotherapy

J.A. Perez-Sanchez 1, R.J. Brisson 1, K.E. Hitchcock 1, R.A. Zlotecki 1, Y. Badyal 2, C. Liu 1, G. Yan 1

1 Department of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL

2 MVision AI, Helsinki, Finland


Due to its superior soft-tissue contrast, MRI is the imaging modality of choice for target and organs-at-risk delineation in prostate radiotherapy. A recent advanced MR scanner allows for an MRI-only workflow owing to its ability to generate synthetic CT (MRCAT) that enables accurate dose calculation and treatment planning. While auto-segmentation algorithms for CT images are increasingly being adopted in clinical practice, there are fewer such algorithms for MRI images. The purpose of this study was to evaluate a deep learning-based auto-segmentation tool (from MVision AI) for both MRI and MRCAT images. We hypothesize that such algorithms could produce accurate contours that require little manual editing, resulting in significant time saving and workflow efficiency in MRI-only treatment planning.


T2 MRI and MRCAT image pairs (acquired on a Philips Ingenia 1.5T MRI Scanner) for 10 prostate patients treated in an MRI-only workflow at our institution were retrospectively collected with IRB approval. These image volumes were then segmented using deep learning-based auto-segmentation models trained for MRI and CT images, respectively. The structures were manually edited by an experienced physician in a commercial imaging informatics system (Velocity). Finally, the auto-segmented and the post-edit structures were compared using Dice Similarity Coefficient (DSC) and 95% Hausdorff Distance (HD95).


Excellent agreement was found between the auto-segmented and post-edit structures. For MRCAT, the average DSC was 0.92±0.03, 0.99±0.00, 0.90±0.06, 0.89±0.07, 0.89±0.06, 0.95±0.04, and 0.96±0.03 for prostate, bladder, rectum, penile bulb, seminal vesicles, left femur, and right femur, respectively. The average HD95 for these structures were 0.65±0.51mm, 0.01±0.01mm, 2.07±1.68mm, 1.10±0.91mm, 1.67±1.81mm, 0.86±1.24mm, and 0.67±1.49mm, respectively. For MRI, the average DSC was 0.96±0.02, 0.99 ±0.00, 0.96±0.03, 0.87±0.03, 0.94±0.06, and 0.89±0.11 for prostate, bladder, rectum, penile bulb, seminal vesicles, and Barrigel spacer, respectively. The average HD95 was 0.25±0.21mm, 0.00±0.00mm, 1.60±1.87mm, 1.43±0.98mm, 0.50±0.86mm, and 1.56±2.26mm, respectively.

Figure 1. MVision segmentations (red) vs. physician-edited segmentations (blue) on MR images of Patient 1. (a) bladder, (b) penile bulb, (c) prostate, (d) rectum, (e) seminal vesicles. Left-to-right: coronal, axial, sagittal cross-sections.

Figure 2. MVision segmentations (red) vs. physician-edited segmentations (blue) on MRCAT images of patient 3. (a) bladder, (b) penile bulb, (c) prostate, (d) rectum, and (e) seminal vesicles. Left-to-right: coronal, axial, sagittal cross-sections.


We evaluated the clinical acceptability of deep learning-based auto-segmentation models for prostate radiotherapy, one for MRI and one for MRCAT delineation. Both models demonstrated superior accuracy, and a high degree of consistency, offering the potential for significant time saving in treatment planning and improved workflow efficiency.

These excellent results are in line with previous studies and external validation of Contour+ quality The novelty of this study is the confirmation of MVision AI male pelvis models performance on synthetic CTs and opens up new opportunities for using it in the MR-only workflow for treating prostate cancer. 

Our Newsletter

Subscribe to get information, latest news and other interesting offers about MVision AI

Related Posts


World Cancer Day – Joining efforts for a better cancer care

World cancer day, celebrated on the 4th of February, was established more than 20 years ago. It represents an initiative of the Union for International Cancer Control (UICC) , the oldest and largest global membership organization dedicated to taking action on cancer. UICC has over 1150 member organizations in 172…



How can AI-based auto-contouring improve the life of patients who need palliative radiation therapy?

Cancer is one of the leading causes of morbidity and death, worldwide. Approximately 18 million people are diagnosed with cancer each year and the number is expected to increase. Pain is experienced by two thirds of patients having advanced, metastatic or terminal disease (1). Other symptoms caused by compression on…



MVision AI Announces Partnership with Medron Medical Systems (Canada)

HELSINKI, Finland - January 22, 2024 MVision AI, a premier software service provider of GBS™ (Guideline-Based Segmentation) solution for same-day radiotherapy treatment planning, proudly announces a new distribution partnership with Medron Medical Systems, a leading health equipment provider in Canada. This collaboration follows the receipt of Medical Device License (MDL)…

Press Releases