3D Reconstruction and alignment by consumer RGB-D sensors and fiducial planar markers for patient positioning in radiation therapy

Facial landmark-guided surface matching for image-to-patient registration with an RGB-D camera

BACKGROUND

Fiducial marker-based image-to-patient registration is the most common way in image-guided neurosurgery, which is labour-intensive, time consuming, invasive and error prone.

METHODS

We proposed a method of facial landmark-guided surface matching for image-to-patient registration using an RGB-D camera. Five facial landmarks are localised from preoperative magnetic resonance (MR) images using deep learning and RGB image using Adaboost with multi-scale block local binary patterns, respectively. The registration of two facial surface point clouds derived from MR images and RGB-D data is initialised by aligning these five landmarks and further refined by weighted iterative closest point algorithm.

RESULTS

Phantom experiment results show the target registration error is less than 3 mm when the distance from the camera to the phantom is less than 1000 mm. The registration takes less than 10 s.

CONCLUSIONS

The proposed method is comparable to the state-of-the-arts in terms of the accuracy yet more time-saving and non-invasive.

Patient posture correction and alignment using mixed reality visualization and the HoloLens 2

PURPOSE

The purpose of this study was to develop and preliminarily test a radiotherapy system for patient posture correction and alignment using mixed reality (MixR) visualization. The write-up of this work also provides an opportunity to introduce the concepts and technology of MixR for a medical physics audience who may be unfamiliar with the topic.

METHODS

A MixR application was developed for on optical-see-through head-mounted display (HoloLens 2) allowing a user to simultaneously and directly view a patient and a reference hologram derived from their simulation CT scan. The hologram provides a visual reference for the exact posture needed during treatment and is initialized in relation to the origin of a radiotherapy device using marker-based tracking. The system further provides marker-less tracking that allows the user tofreely navigate the room as they view and align the patient from various angles. The system was preliminarily tested using both a rigid (pelvis) and nonrigid (female mannequin) anthropomorphic phantom. Each phantom was aligned via hologram and accuracy quantified using CBCT and CT.

RESULTS

A fully realized system was developed. Rigid registration accuracy was on the order of 3.0 ± 1.5 mm based on the performance of three users repeating alignment five times each. The lateral direction showed the most variability among users and was associated with the largest off-sets (approximately 2.0 mm). For nonrigid alignment, the MixR setup outperformed a setup based on three-point alignment and setup photos, the latter of which showed a difference in arm position of 2 cm and a torso roll of 6-7°.

CONCLUSIONS

MixR visualization is a rapidly emerging domain that has the potential to significantly impact the field of medicine. The current application is an illustration of this and highlights the advantages of MixR for patient setup in radiation oncology. The key feature of the system is the way in which it transforms nonrigid registration into rigid registration by providing an efficient, portable, and cost-effective mechanism for reproducing patient posture without the use of ionizing radiation. Preliminary estimates of registration accuracy indicate clinical viability and form the foundation for further development and clinical testing.

Joint scene and object tracking for cost-Effective augmented reality guided patient positioning in radiation therapy

BACKGROUND AND OBJECTIVE

The research is done in the field of Augmented Reality (AR) for patient positioning in radiation therapy is scarce. We propose an efficient and cost-effective algorithm for tracking the scene and the patient to interactively assist the patient's positioning process by providing visual feedback to the operator. Up to our knowledge, this is the first framework that can be employed for mobile interactive AR to guide patient positioning.

METHODS

We propose a pointcloud processing method that, combined with a fiducial marker-mapper algorithm and the generalized ICP algorithm, tracks the patient and the camera precisely and efficiently only using the CPU unit. The 3D reference model and body marker map alignment is calculated employing an efficient body reconstruction algorithm.

RESULTS

Our quantitative evaluation shows that the proposed method achieves a translational and rotational error of 4.17 mm/0.82^∘ at 9 fps. Furthermore, the qualitative results demonstrate the usefulness of our algorithm in patient positioning on different human subjects.

CONCLUSION

Since our algorithm achieves a relatively high frame rate and accuracy employing a regular laptop (without a dedicated GPU), it is a very cost-effective AR-based patient positioning method. It also opens the way for other researchers by introducing a framework that could be improved upon for better mobile interactive AR patient positioning solutions in the future.