Automatic detection of patient identification and positioning errors in radiation therapy treatment using 3-dimensional setup images

Jani, S; Low, Daniel A.; Lamb, James

doi:10.1016/j.prro.2015.06.004

Cited by 7 publications

(11 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As compared to a similar work that uses non–DL techniques 14 to find vertebral misalignment errors in thoracic CBCT‐guided radiotherapy treatments, EDM 2 resulted in higher sensitivity (0.95 vs. 0.90) for a fixed specificity of 99%. Additionally, our model was validated on a larger test set composed of unseen data from two different institutions as compared to the non‐DL techniques that were validated using a 10‐fold cross‐validation method for a training‐testing dataset composed of 57 patients from a single institution.…”

Section: Discussionmentioning

confidence: 80%

Development and interinstitutional validation of an automatic vertebral‐body misalignment error detector for cone‐beam CT‐guided radiotherapy

et al. 2022

Self Cite

View full text Add to dashboard Cite

Background: In cone-beam computed tomography (CBCT)-guided radiotherapy, off -by-one vertebral-body misalignments are rare but serious errors that lead to wrong-site treatments. Purpose: An automatic error detection algorithm was developed that uses a three-branch convolutional neural network error detection model (EDM) to detect off -by-one vertebral-body misalignments using planning computed tomography (CT) images and setup CBCT images. Methods: Algorithm training and test data consisted of planning CTs and CBCTs from 480 patients undergoing radiotherapy treatment in the thoracic and abdominal regions at two radiotherapy clinics. The clinically applied registration was used to derive true-negative (no error) data. The setup and planning images were then misaligned by one vertebral-body in both the superior and inferior directions, simulating the most likely misalignment scenarios. For each of the aligned and misaligned 3D image pairs, 2D slice pairs were automatically extracted in each anatomical plane about a point within the vertebral column. The three slice pairs obtained were then inputted to the EDM that returned a probability of vertebral misalignment. One model (EDM 1 ) was trained solely on data from institution 1. EDM 1 was further trained using a lower learning rate on a dataset from institution 2 to produce a fine-tuned model, EDM 2 . Another model, EDM 3 , was trained from scratch using a training dataset composed of data from both institutions. These three models were validated on a randomly selected and unseen dataset composed of images from both institutions, for a total of 303 image pairs. The model performances were quantified using a receiver operating characteristic analysis. Due to the rarity of vertebral-body misalignments in the clinic, a minimum threshold value yielding a specificity of at least 99% was selected. Using this threshold, the sensitivity was calculated for each model, on each institution's test set separately. Results: When applied to the combined test set, EDM 1 , EDM 2 , and EDM 3 resulted in an area under curve of 99.5%, 99.4%, and 99.5%, respectively. EDM 1 achieved a sensitivity of 96% and 88% on Institution 1 and Institution 2 test set, respectively. EDM 2 obtained a sensitivity of 95% on each institution's test set. EDM 3 achieved a sensitivity of 95% and 88% on Institution 1 and Institution 2 test set, respectively. Conclusion:The proposed algorithm demonstrated accuracy in identifying off -by-one vertebral-body misalignments in CBCT-guided radiotherapy that 6410

show abstract

Section: Discussionmentioning

confidence: 80%

Development and interinstitutional validation of an automatic vertebral‐body misalignment error detector for cone‐beam CT‐guided radiotherapy

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…This technique was able to correctly identify patients using image pairs focused on the prostate, cranial, and thoracic/lumbar spine regions. This work was extended to evaluate 3D‐3D radiographic image registrations using an automated process by Jani et al (16) In this work they showed that registration similarity metrics could accurately identify patients and setup misalignments for head and neck, pelvis, and spine cases.…”

Section: Discussionmentioning

confidence: 99%

“…In that work they found 0%‐10% of patients wrongly identified depending on threshold selection, and misclassification probabilities of 0.0045 and 0.014 for the prostate and spine, respectively. Jani et al (16) used 3D‐3D radiographic image registrations to classify head and neck, pelvis, and spine patients. They found

< 5 %

of patients wrongly identified for the evaluated sites depending on the technique, and misclassification probabilities of 0.0066, 0.0167, and 0 for the head and neck, pelvis, and spine, respectively.…”

Section: Discussionmentioning

confidence: 99%

A novel method for radiotherapy patient identification using surface imaging

Wiant

Verchick

Gates

et al. 2016

J Applied Clin Med Phys

View full text Add to dashboard Cite

Performing a procedure on the wrong patient or site is one of the greatest errors that can occur in medicine. The addition of automation has been shown to reduce errors in many processes. In this work we explore the use of an automated patient identification process using optical surface imaging for radiotherapy treatments. Surface imaging uses visible light to align the patient to a reference surface in the treatment room. It is possible to evaluate the similarity between a daily set‐up surface image and the reference image using distance to agreement between the points on the two surfaces. The higher the percentage overlapping points within a defined distance, the more similar the surfaces. This similarity metric was used to intercompare 16 left‐sided breast patients. The reference surface for each patient was compared to 10 daily treatment surfaces for the same patient, and 10 surfaces from each of the other 15 patients (for a total of 160 comparisons per patient), looking at the percent of points overlapping. For each patient, the minimum same‐patient similarity score was higher than the maximum different‐patient score. For the group as a whole a threshold was able to classify correct and incorrect patients with high levels of accuracy. A 10‐fold cross‐validation using linear discriminant analysis gave cross‐validation loss of 0.0074. An automated process using surface imaging is a feasible option to provide nonharmful daily patient identification verification using currently available technology.PACS number(s): 87.53.Jw, 87.55.N‐, 87.55.Qr, 87.57.N‐, 87.63.L‐

show abstract

“…In order to study the behavior of DRRs with arbitrary 6D CT transformations, it may be necessary to perform offline DRR rendering. For example, at our institution, we are developing algorithms to detect patient positioning errors in IGRT images 6,7 . Algorithm training required the use of simulated errors produced by matching clinically acquired X‐ray projections with DRRs of adjacent (“wrong”) vertebral bodies.…”

Section: Introductionmentioning

confidence: 99%

“…For example, at our institution, we are developing algorithms to detect patient positioning errors in IGRT images. 6 , 7 Algorithm training required the use of simulated errors produced by matching clinically acquired X‐ray projections with DRRs of adjacent (“wrong”) vertebral bodies. The ExacTrac offline preparation and review station allows simulation of alignment to the wrong vertebral body but does not allow for export of the corresponding optimized DRRs.…”

Section: Introductionmentioning

confidence: 99%

Offline generator for digitally reconstructed radiographs of a commercial stereoscopic radiotherapy image‐guidance system

Charters

Bertram

Lamb

2022

J Applied Clin Med Phys

Self Cite

View full text Add to dashboard Cite

Purpose Image‐guided radiotherapy (IGRT) research sometimes involves simulated changes to patient positioning using retrospectively collected clinical data. For example, researchers may simulate patient misalignments to develop error detection algorithms or positioning optimization algorithms. The Brainlab ExacTrac system can be used to retrospectively “replay” simulated alignment scenarios but does not allow export of digitally reconstructed radiographs (DRRs) with simulated positioning variations for further analysis. Here we describe methods to overcome this limitation and replicate ExacTrac system DRRs by using projective geometry parameters contained in the ExacTrac configuration files saved for every imaged subject. Methods Two ExacTrac DRR generators were implemented, one with custom MATLAB software based on first principles, and the other using libraries from the Insight Segmentation and Registration Toolkit (ITK). A description of perspective projections for DRR rendering applications is included, with emphasis on linear operators in real projective space double-struckP3${\mathbb{P}^3}$. We provide a general methodology for the extraction of relevant geometric values needed to replicate ExacTrac DRRs. Our generators were tested on phantom and patient images, both acquired in a known treatment position. We demonstrate the validity of our methods by comparing our generated DRRs to reference DRRs produced by the ExacTrac system during a treatment workflow using a manual landmark analysis as well as rigid registration with the elastix software package. Results Manual landmarks selected between the corresponding DRR generators across patient and phantom images have an average displacement of 1.15 mm. For elastix image registrations, we found that absolute value vertical and horizontal translations were 0.18 and 0.35 mm on average, respectively. Rigid rotations were within 0.002 degrees. Conclusion Custom and ITK‐based algorithms successfully reproduce ExacTrac DRRs and have the distinctive advantage of incorporating any desired 6D couch position. An open‐source repository is provided separately for users to implement in IGRT patient positioning research.

show abstract

Automatic detection of patient identification and positioning errors in radiation therapy treatment using 3-dimensional setup images

Cited by 7 publications

References 21 publications

Development and interinstitutional validation of an automatic vertebral‐body misalignment error detector for cone‐beam CT‐guided radiotherapy

Development and interinstitutional validation of an automatic vertebral‐body misalignment error detector for cone‐beam CT‐guided radiotherapy

A novel method for radiotherapy patient identification using surface imaging

Offline generator for digitally reconstructed radiographs of a commercial stereoscopic radiotherapy image‐guidance system

Contact Info

Product

Resources

About