Real-time markerless tumour tracking with patient-specific deep learning using a personalised data generation strategy: proof of concept by phantom study

Takahashi, Wataru; Oshikawa, Shota; Mori, Shinichiro

doi:10.1259/bjr.20190420

Cited by 19 publications

(39 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared with the previous works, 11,16 this work shows advantages in the following aspects. First, our work predicted the contour of the tumor and determined the tumor position based on the centroid of the contour.…”

Section: Discussionmentioning

confidence: 95%

“…In their work, two‐dimensional (2D) localization was not sufficient for tumor tracking, and the appropriateness of localization with the top‐left corner of the bounding box to evaluate performance is worth discussing. In contrast, several investigators developed a three‐dimensional (3D) localization model 14–16 . Their model is based on a fully convolutional neural network 17 and predicts the target position based on the center of the target probability map.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Development of a deep learning‐based patient‐specific target contour prediction model for markerless tumor positioning

et al. 2022

View full text Add to dashboard Cite

Purpose For pancreatic cancer patients, image guided radiation therapy and real‐time tumor tracking (RTTT) techniques can deliver radiation to the target accurately. Currently, for the radiation therapy machine with kV X‐ray imaging systems, internal markers must be implemented to facilitate tumor tracking. The purpose of this study was to develop a markerless deep learning‐based pancreatic tumor positioning procedure for real‐time tumor tracking with a kV X‐ray imaging system. Methods and materials Fourteen pancreatic cancer patients treated with intensity‐modulated radiation therapy from six fixed gantry angles with a gimbal‐head radiotherapy system were included in this study. For a gimbal‐head radiotherapy system, the three‐dimensional (3D) intrafraction target position can be determined using an orthogonal kV X‐ray imaging system. All patients underwent four‐dimensional computed tomography (4DCT) simulations for treatment planning, which were divided into 10 respiratory phases. After a patient's 4DCT was acquired, for each X‐ray tube angle, 10 digitally reconstructed radiograph (DRR) images were obtained. Then, a data augmentation procedure was conducted. The data augmentation procedure first rotated the CT volume around the superior–inferior and anterior–posterior directions from −3° to 3° in 1.5° intervals. Then, the Super‐SloMo model was adapted to interpolate 10 frames between respiratory phases. In total, the data augmentation procedure expanded the data scale 250‐fold. In this study, for each patient, 12 datasets containing the DRR images from each specific X‐ray tube angle based on the radiation therapy plan were obtained. The augmented dataset was randomly divided into training and testing datasets. The training dataset contained 2000 DRR images with clinical target volume (CTV) contours labeled for fine‐tuning the pre‐trained target contour prediction model. After the fine‐tuning, the patient and X‐ray tube angle‐specific CTV contour prediction model was acquired. The testing dataset contained the remaining 500 images to evaluate the performance of the CTV contour prediction model. The dice similarity coefficient (DSC) between the area enclosed by the CTV contour and predicted contour was calculated to evaluate the model's contour prediction performance. The 3D position of the CTV was calculated based on the centroid of the contour in the orthogonal DRR images, and the 3D error of the prediction position was calculated to evaluate the CTV positioning performance. For each patient, the DSC results from 12 X‐ray tube angles and 3D error from 6 gantry angles were calculated, representing the novelty of this study. Results The mean and standard deviation (SD) of all patients’ DSCs were 0.98 and 0.015, respectively. The mean and SD of the 3D error were 0.29 mm and 0.14 mm, respectively. The global maximum 3D error was 1.66 mm, and the global minimum DSC was 0.81. The mean calculation time for CTV contour prediction was 55 ms per image. This fulfills the requirement of RTTT. Conclusions Regarding...

show abstract

Section: Discussionmentioning

confidence: 95%

Section: Introductionmentioning

confidence: 99%

Development of a deep learning‐based patient‐specific target contour prediction model for markerless tumor positioning

et al. 2022

View full text Add to dashboard Cite

show abstract

“…When tested on DRRs, the tracking error was approximately 1 mm with a processing time of 25 ms/frame for contouring and tracking. A DL approach for segmenting lung tumours was also developed by Takahashi et al 56 Their approach uses a FCN trained on patient-specific DRRs. The system was tested in a phantom study with tumours of 1, 2 and 3 cm in size.…”

Section: Markerless-based Approachesmentioning

confidence: 99%

A review of artificial intelligence applications for motion tracking in radiotherapy

2021

View full text Add to dashboard Cite

During radiotherapy, the organs and tumour move as a result of the dynamic nature of the body; this is known as intrafraction motion. Intrafraction motion can result in tumour underdose and healthy tissue overdose, thereby reducing the effectiveness of the treatment while increasing toxicity to the patients. There is a growing appreciation of intrafraction target motion management by the radiation oncology community. Real‐time image‐guided radiation therapy (IGRT) can track the target and account for the motion, improving the radiation dose to the tumour and reducing the dose to healthy tissue. Recently, artificial intelligence (AI)‐based approaches have been applied to motion management and have shown great potential. In this review, four main categories of motion management using AI are summarised: marker‐based tracking, markerless tracking, full anatomy monitoring and motion prediction. Marker‐based and markerless tracking approaches focus on tracking the individual target throughout the treatment. Full anatomy algorithms monitor for intrafraction changes in the full anatomy within the field of view. Motion prediction algorithms can be used to account for the latencies due to the time for the system to localise, process and act.

show abstract

“…One disadvantage of some vision-based approaches is their initialization step. Many such algorithms require manual initialization [33,63] or semiautomatic initialization [125]. Even when they initialize automatically, they usually have to start from a well-known point [90].…”

Section: Vision-based Tracking and Registration Challengesmentioning

confidence: 99%

A Survey of Marker-Less Tracking and Registration Techniques for Health & Environmental Applications to Augmented Reality and Ubiquitous Geospatial Information Systems

Sadeghi‐Niaraki

Choi

2020

Sensors

View full text Add to dashboard Cite

Most existing augmented reality (AR) applications are suitable for cases in which only a small number of real world entities are involved, such as superimposing a character on a single surface. In this case, we only need to calculate pose of the camera relative to that surface. However, when an AR health or environmental application involves a one-to-one relationship between an entity in the real-world and the corresponding object in the computer model (geo-referenced object), we need to estimate the pose of the camera in reference to a common coordinate system for better geo-referenced object registration in the real-world. New innovations in developing cheap sensors, computer vision techniques, machine learning, and computing power have helped to develop applications with more precise matching between a real world and a virtual content. AR Tracking techniques can be divided into two subcategories: marker-based and marker-less approaches. This paper provides a comprehensive overview of marker-less registration and tracking techniques and reviews their most important categories in the context of ubiquitous Geospatial Information Systems (GIS) and AR focusing to health and environmental applications. Basic ideas, advantages, and disadvantages, as well as challenges, are discussed for each subcategory of tracking and registration techniques. We need precise enough virtual models of the environment for both calibrations of tracking and visualization. Ubiquitous GISs can play an important role in developing AR in terms of providing seamless and precise spatial data for outdoor (e.g., environmental applications) and indoor (e.g., health applications) environments.

show abstract

Real-time markerless tumour tracking with patient-specific deep learning using a personalised data generation strategy: proof of concept by phantom study

Cited by 19 publications

References 22 publications

Development of a deep learning‐based patient‐specific target contour prediction model for markerless tumor positioning

Development of a deep learning‐based patient‐specific target contour prediction model for markerless tumor positioning

A review of artificial intelligence applications for motion tracking in radiotherapy

A Survey of Marker-Less Tracking and Registration Techniques for Health & Environmental Applications to Augmented Reality and Ubiquitous Geospatial Information Systems

Contact Info

Product

Resources

About