Enhanced Surface Registration Techniques for Patient Positioning Control in Breast Cancer Radiotherapy

Optimal marker placement in hadrontherapy: intelligent optimization strategies with augmented Lagrangian pattern search

PURPOSE

In high precision photon radiotherapy and in hadrontherapy, it is crucial to minimize the occurrence of geometrical deviations with respect to the treatment plan in each treatment session. To this end, point-based infrared (IR) optical tracking for patient set-up quality assessment is performed. Such tracking depends on external fiducial points placement. The main purpose of our work is to propose a new algorithm based on simulated annealing and augmented Lagrangian pattern search (SAPS), which is able to take into account prior knowledge, such as spatial constraints, during the optimization process.

MATERIAL AND METHODS

The SAPS algorithm was tested on data related to head and neck and pelvic cancer patients, and that were fitted with external surface markers for IR optical tracking applied for patient set-up preliminary correction. The integrated algorithm was tested considering optimality measures obtained with Computed Tomography (CT) images (i.e. the ratio between the so-called target registration error and fiducial registration error, TRE/FRE) and assessing the marker spatial distribution. Comparison has been performed with randomly selected marker configuration and with the GETS algorithm (Genetic Evolutionary Taboo Search), also taking into account the presence of organs at risk.

RESULTS

The results obtained with SAPS highlight improvements with respect to the other approaches: (i) TRE/FRE ratio decreases; (ii) marker distribution satisfies both marker visibility and spatial constraints. We have also investigated how the TRE/FRE ratio is influenced by the number of markers, obtaining significant TRE/FRE reduction with respect to the random configurations, when a high number of markers is used.

CONCLUSIONS

The SAPS algorithm is a valuable strategy for fiducial configuration optimization in IR optical tracking applied for patient set-up error detection and correction in radiation therapy, showing that taking into account prior knowledge is valuable in this optimization process. Further work will be focused on the computational optimization of the SAPS algorithm toward fast point-of-care applications.

A semi-automated 3-D annotation method for breast ultrasound imaging: system development and feasibility study on phantoms

Spatial annotation is an essential step in breast ultrasound imaging, because the follow-up diagnosis and treatment are based on this annotation. However, the current method for annotation is manual and highly dependent on the operator's experience. Moreover, important spatial information, such as the probe tilt angle, cannot be indicated in the clinical 2-D annotations. To solve these problems, we developed a semi-automated 3-D annotation method for breast ultrasound imaging. A spatial sensor was fixed on an ultrasound probe to obtain the image spatial data. Three-dimensional virtual models of breast and probe were used to annotate image locations. After the reference points were recorded, this system displayed the image annotations automatically. Compared with the conventional manual annotation method, this new annotation system has higher accuracy as indicated by the phantom test results. In addition, this new annotation method has good repeatability, with intra-class correlation coefficients of 0.907 (average variation: ≤3.45%) and 0.937 (average variation: ≤2.85%) for the intra-rater and inter-rater tests, respectively. Breast phantom experiments simulating clinical breast scanning further indicated the feasibility of this system for clinical applications. This new annotation method is expected to facilitate more accurate, intuitive and rapid breast ultrasound diagnosis.

Tumor tracking based on correlation models in scanned ion beam therapy: an experimental study

Accurate dose delivery to extra-cranial lesions requires tumor motion compensation. An effective compensation can be achieved by real-time tracking of the target position, either measured in fluoroscopy or estimated through correlation models as a function of external surrogate motion. In this work, we integrated two internal/external correlation models (a state space model and an artificial neural network-based model) into a custom infra-red optical tracking system (OTS). Dedicated experiments were designed and conducted at GSI (Helmholtzzentrum für Schwerionenforschung). A robotic breathing phantom was used to reproduce regular and irregular internal target motion as well as external thorax motion. The position of a set of markers placed on the phantom thorax was measured with the OTS and used by the correlation models to infer the internal target position in real-time. Finally, the estimated target position was provided as input for the dynamic steering of a carbon ion beam. Geometric results showed that the correlation models transversal (2D) targeting error was always lower than 1.3 mm (root mean square). A significant decrease of the dosimetric error with respect to the uncompensated irradiation was achieved in four out of six experiments, demonstrating that phase shifts are the most critical irregularity for external/internal correlation models.

Assessment of set-up variability during deep inspiration breath hold radiotherapy for breast cancer patients by 3D-surface imaging

PURPOSE

To quantify set-up uncertainties during voluntary deep inspiration breath hold (DIBH) radiotherapy using 3D-surface imaging in patients with left sided breast cancer.

MATERIAL AND METHODS

Nineteen patients were included. Cone-beam CT-scan (CBCT) was used for online set-up correction while patients were instructed to perform a voluntary DIBH. The reproducibility of the DIBH during treatment was monitored with 2D-fluoroscopy and portal imaging. Simultaneously, a surface imaging system was used to capture 3D-surfaces throughout CBCT acquisition and delivery of treatment beams. Retrospectively, all captured surfaces were registered to the planning-CT surface. Interfraction, intra-fraction and intra-beam set-up variability were quantified in left-right, cranio-caudal and anterior-posterior direction.

RESULTS

Inter-fraction systematic (Σ) and random (σ) translational errors (1SD) before and after set-up correction were between 0.20-0.50 cm and 0.09-0.22 cm, respectively, whereas rotational Σ and σ errors were between 0.08 and 1.56°. The intra-fraction Σ and σ errors were ≤ 0.14 cm and ≤ 0.47°. The intra-beam SD variability was ≤ 0.08 cm and ≤ 0.28° in all directions.

CONCLUSION

Quantification of 3D set-up variability in DIBH RT showed that patients are able to perform a very stable and reproducible DIBH within a treatment fraction. However, relatively large inter-fraction variability requires online image guided set-up corrections.

Accuracy in breast shape alignment with 3D surface fitting algorithms

Surface imaging is in use in radiotherapy clinical practice for patient setup optimization and monitoring. Breast alignment is accomplished by searching for a tentative spatial correspondence between the reference and daily surface shape models. In this study, the authors quantify whole breast shape alignment by relying on texture features digitized on 3D surface models. Texture feature localization was validated through repeated measurements in a silicone breast phantom, mounted on a high precision mechanical stage. Clinical investigations on breast shape alignment included 133 fractions in 18 patients treated with accelerated partial breast irradiation. The breast shape was detected with a 3D video based surface imaging system so that breathing was compensated. An in-house algorithm for breast alignment, based on surface fitting constrained by nipple matching (constrained surface fitting), was applied. Results were compared with a commercial software where no constraints are utilized (unconstrained surface fitting). Texture feature localization was validated within 2 mm in each anatomical direction. Clinical data show that unconstrained surface fitting achieves adequate accuracy in most cases, though nipple mismatch is considerably higher than residual surface distances (3.9 mm vs 0.6 mm on average). Outliers beyond 1 cm can be experienced as the result of a degenerate surface fit, where unconstrained surface fitting is not sufficient to establish spatial correspondence. In the constrained surface fitting algorithm, average surface mismatch within 1 mm was obtained when nipple position was forced to match in the [1.5; 5] mm range. In conclusion, optimal results can be obtained by trading off the desired overall surface congruence vs matching of selected landmarks (constraint). Constrained surface fitting is put forward to represent an improvement in setup accuracy for those applications where whole breast positional reproducibility is an issue.

Integration of Enhanced Optical Tracking Techniques and Imaging in IGRT

In external beam radiotherapy, modern technologies for dynamic dose delivery and beam conformation provide high selectivity in radiation dose administration to the pathological volume. A comparable accuracy level is needed in the 3-D localization of tumor and organs at risk (OARs), in order to accomplish the planned dose distribution in the reality of each irradiation session. In-room imaging techniques for patient setup verification and tumor targeting may benefit of the combined daily use of optical tracking technologies, supported by techniques for the detection and compensation of organ motion events. Multiple solutions to enhance the use of optical tracking for the on-line correction of target localization uncertainties are described, with specific emphasis on the compensation of setup errors, breathing movements and non-rigid deformations. The final goal is the implementation of customized protocols where appropriate external landmarks, to be tracked in real-time by means of non-invasive optical devices, are selected as a function of inner target localization. The presented methodology features high accuracy in patient setup optimization, also providing a valuable tool for on-line patient surveillance, taking into account both breathing and deformation effects. The methodic application of optical tracking is put forward to represent a reliable and low cost procedure for the reduction of safety margins, once the patient-specific correlation between external landmarks and inner structures has been established. Therefore, the integration of optical tracking with in-room imaging devices is proposed as a way to gain higher confidence in the framework of Image Guided Radiation Therapy (IGRT) treatments.

Breast movement during normal and deep breathing, respiratory training and set up errors: implications for external beam partial breast irradiation

This study was designed to evaluate interfraction and intrafraction breast movement and to study the effect of respiratory training on respiratory indices. Five patients were immobilized in supine position in a vacuum bag and three-dimensional set up errors, respiratory movement of the breast during normal and deep breathing, tidal volume and breath hold time were recorded. All patients underwent respiratory training and all the respiratory indices were re-evaluated at the end of training. Cumulative maximum movement error (CMME) was calculated by adding directional maximum set up error and maximum post training movement during normal breathing. The mean set up deviation was 1.3 mm (SD +/- 0.5 mm), 1.3 mm (SD +/- 0.3 mm) and 4.4 mm (SD +/- 2.6 mm) in the mediolateral, superoinferior and anteroposterior dimensions. Pre-training mean of the maximum marker movement during normal breathing was 1.07 mm, 1.94 mm and 1.86 mm in the mediolateral, superoinferior and anteroposterior dimensions. During deep breathing these values were 2 mm, 5.5 mm and 4.8 mm. While respiratory training had negligible effect on breast movement during normal breathing, it resulted in a modest reduction during deep breathing (p = 0.2). The mean CMME recorded for these patients was 3.4 mm, 4.5 mm and 7.1 mm in the mediolateral, superoinferior and anteroposterior dimension. Respiratory training also resulted in an increase in breath hold time from a mean of 31 s to 44 s (p = 0.04) and tidal volume from a mean of 560 cm(3) to 1160 cm(3) (p = 0.04). With patients immobilized in the vacuum bag the CMMEs are relatively less. Individualized directional margins may aid in reduction of planning target volume (PTV).

Patient set-up verification by infrared optical localization and body surface sensing in breast radiation therapy

BACKGROUND AND PURPOSE

The aim of the study was to investigate the clinical application of a technique for patient set-up verification in breast cancer radiotherapy, based on the 3D localization of a hybrid configuration of surface control points.

MATERIALS AND METHODS

An infrared optical tracker provided the 3D position of two passive markers and 10 laser spots placed around and within the irradiation field on nine patients. A fast iterative constrained minimization procedure was applied to detect and compensate patient set-up errors, through the control points registration with reference data coming from treatment plan (markers reference position, CT-based surface model).

RESULTS

The application of the corrective spatial transformation estimated by the registration procedure led to significant improvement of patient set-up. Median value of 3D errors affecting three additional verification markers within the irradiation field decreased from 5.7 to 3.5 mm. Errors variability (25-75%) decreased from 3.2 to 2.1 mm. Laser spots registration on the reference surface model was documented to contribute substantially to set-up errors compensation.

CONCLUSIONS

Patient set-up verification through a hybrid set of control points and constrained surface minimization algorithm was confirmed to be feasible in clinical practice and to provide valuable information for the improvement of the quality of patient set-up, with minimal requirement of operator-dependant procedures. The technique combines conveniently the advantages of passive markers based methods and surface registration techniques, by featuring immediate and robust estimation of the set-up accuracy from a redundant dataset.

Clinical experience with a 3D surface patient setup system for alignment of partial-breast irradiation patients

PURPOSE

To assess the utility of surface imaging on patient setup for accelerated partial-breast irradiation (APBI).

METHODS AND MATERIAL

A photogrammetry system was used in parallel to APBI setup by laser and portal imaging. Surface data were acquired after laser and port-film setup for 9 patients. Surfaces were analyzed in comparison to a reference surface from the first treatment session by use of rigid transformations. The surface model after laser setup was used in a simulated photogrammetry setup procedure. In addition, breathing data were acquired by surface acquisition at a frame rate of 7 Hz.

RESULTS

Mean 3D displacement was 7.3 mm (SD, 4.4 mm) and 7.6 mm (SD, 4.2 mm) for laser and port film, respectively. Simulated setup with the photogrammetry system yielded mean displacement of 1 mm (SD, 1.2 mm). Distance analysis resulted in mean distances of 3.7 mm (SD, 4.9 mm), 4.3 mm (SD, 5.6 mm), and 1.6 mm (SD, 2.4 mm) for laser, port film, and photogrammetry, respectively. Breathing motion at isocenter was smaller than 3.7 mm, with a mean of 1.9 mm (SD, 1.1 mm).

CONCLUSIONS

Surface imaging for PBI setup appears promising. Alignment of the 3D breast surface achieved by stereo-photogrammetry shows greater breast topology congruence than when patients are set up by laser or portal imaging. A correlation of breast surface and CTV must be quantitatively established.

A Neural Network Based Method for Optical Patient Set-up Registration in Breast Radiotherapy

Patient set-up optimization is required in breast-cancer radiotherapy to fill the accuracy gap between personalized treatment planning and uncertainties in the irradiation set-up. Opto-electronic systems allow implementing automatic procedures to minimize the positional mismatches of light-reflecting markers located on the patient surface with respect to a corresponding reference configuration. The same systems are used to detect the position of the irradiated body surface by means of laser spots; patient set-up is then corrected by matching the control points onto a CT based reference model through surface registration algorithms. In this paper, a non-deterministic approach based on Artificial Neural Networks is proposed for the automatic, real-time verification of geometrical set-up of breast irradiation. Unlike iterative surface registration methods, no passive fiducials are used and true real-time performance is obtained. Moreover, the non-deterministic modeling performed by the neural algorithm minimizes sensitivity to intra-fractional and inter-fractional non-rigid motion of the breast. The technique was validated through simulated activities by using reference CT data acquired on four subjects. Results show that the procedure is able to detect and reduce simulated set-up errors and revealed high reliability in patient position correction, even when the surface deformation is included in testing conditions.