Anthony F. Waller

A novel method for volumetric MRI response assessment of enhancing brain tumors.

Current radiographic response criteria for brain tumors have difficulty describing changes surrounding postoperative resection cavities. Volumetric techniques may offer improved assessment, however usually are time-consuming, subjective and require expert opinion and specialized magnetic resonance imaging (MRI) sequences. We describe the application of a novel volumetric software algorithm that is nearly fully automated and uses standard T1 pre- and post-contrast MRI sequences. T1-weighted pre- and post-contrast images are automatically fused and normalized. The tumor region of interest is grossly outlined by the user. An atlas of the nasal mucosa is automatically detected and used to normalize levels of enhancement. The volume of enhancing tumor is then automatically calculated. We tested the ability of our method to calculate enhancing tumor volume with resection cavity collapse and when the enhancing tumor is obscured by subacute blood in a resection cavity. To determine variability in results, we compared narrowly-defined tumor regions with tumor regions that include adjacent meningeal enhancement and also compared different contrast enhancement threshold levels used for the automatic calculation of enhancing tumor volume. Our method quantified enhancing tumor volume despite resection cavity collapse. It detected tumor volume increase in the midst of blood products that incorrectly caused decreased measurements by other techniques. Similar trends in volume changes across scans were seen with inclusion or exclusion of meningeal enhancement and despite different automated thresholds for tissue enhancement. Our approach appears to overcome many of the challenges with response assessment of enhancing brain tumors and warrants further examination and validation.

A measure to evaluate deformable registration fields in clinical settings

Deformable registration has migrated from a research topic to a widely used clinical tool that can improve radiotherapeutic treatment accuracy by tracking anatomical changes. Although various mathematical formulations have been reported in the literature and implemented in commercial software, we lack a straightforward method to verify a given solution in routine clinical use. We propose a metric using concepts derived from vector analysis that complements the standard evaluation tools to identify unrealistic wrappings in a displacement field. At the heart of the proposed procedure is identification of vortexes in the displacement field that do not correspond to underlying anatomical changes. Vortexes are detected and their intensity quantified using the CURL operator and presented as a vortex map overlaid on the original anatomy for rapid identification of problematic regions. We show application of the proposed metric on clinical scenarios of adaptive radiotherapy and treatment response assessment, where the CURL operator quantitatively detected errors in the displacement field and identified problematic regions that were invisible to classical voxel‐based evaluation methods. Unrealistic warping not visible to standard voxel‐based solution assessment can produce erroneous results when the deformable solution is applied on a secondary dataset, such as dose matrix in adaptive therapy or PET data for treatment response assessment. The proposed metric for evaluating deformable registration provides increased usability and accuracy of detecting unrealistic deformable registration solutions when compared to standard intensity‐based approaches. It is computationally efficient and provides a valuable platform for the clinical acceptance of image‐guided radiotherapy. PACS numbers: 87.57.nj; 87.55.Qr; 87.57.cp

Voxel clustering for quantifying PET-based treatment response assessment.

PURPOSE Imaging biomarkers are crucial in managing treatment options for cancer patients. They are extremely powerful tools since they allow personalized treatment assessment early during therapy by using repeated imaging to detect and quantify tumor response. Currently, treatment response assessment from consecutive imaging is measured by simple global measures that do not capture a tumors heterogeneous response. The authors present an automated, multivoxel metric that groups voxels into clusters of changes for a local definition of radiation treatment efficiency from multiple PET imaging studies acquired at different time periods for assessing therapeutic response. METHODS The algorithm employs level-set mathematics to extract changing features to classify voxels into response patterns. First, pretreatment and post-treatment PET images were aligned using a deformable registration to correct for posture and soft tissue changes. The detailed mapping was modeled by free form deformations B-spline optimized using the limited memory L-BFGS algorithm. The posture-corrected datasets are then subtracted to produce an image of molecular changes embedded with noise. Once images were aligned and subtracted, a segmentation algorithm combining the concepts of voxel and distance-based techniques classified voxels into patterns of signal reduction or enhancement. Although signal reduction is evidence of successful treatment, signal-enhancing regions are an indication of treatment failure. For an in depth analysis of potential treatment errors, patterns of signal enhancement were correlated with the radiation treatment dose and anatomical structures from the treatment plan using image registration methods. RESULTS The algorithm was retrospectively applied to PET∕CT and radiotherapy (RT) oncology data from an NCI-sponsored clinical trial (81 clinical cases from RTOG 0522 Trial) for combined drug and radiation therapy in head and neck carcinomas. This clinical trial dataset presented a realistic environment for implementing and validating our algorithm to correlate local response as observed in serial PET with delivered dose. The technique was instrumental in detecting geographical and segmentation misses on the actual clinical cases by providing accurate voxel-by-voxel analysis of metabolic changes. Results of the level-set based clustering algorithm are saved as a detailed report of enhancing∕nonenhancing regions and their location, and can be further displayed as a colorwash laid over the original anatomy for in depth analysis. CONCLUSIONS The automated technique was instrumental in analyzing treatment response in the clinical cases and provided an useful tool for accurate, outcome-based response assessment of the radiation treatment process. The developed method is general and should be extendable to other high-resolution diagnostic imaging with minor modifications.

Multileaf collimator-based linear accelerator radiosurgery: five-year efficiency analysis.

BACKGROUND In 1989, Emory University initiated a linear accelerator (linac) radiosurgery program using circular collimators. In 2001, the program converted to a multileaf collimator. Since then, the treatment parameters of each patient have been stored in the record-and-verify system. Three major changes have occurred in the radiosurgery program in the past 6 years: in 2002, treatment was changed from static conformal beams to dynamic conformal arc (DCA) therapy, and all patients were imaged before treatment. Beginning in 2005, a linac was used, with the opportunity to treat at higher dose rates (600-1,000 monitor units/min). The aim of this study was to analyze the time required to deliver radiosurgery and the factors affecting treatment delivery. Benchmark data are provided for centers contemplating initiating linac radiosurgery programs. MATERIALS AND METHODS Custom software was developed to mine the record-and-verify system database and automatically perform a chart review on patients who underwent stereotactic radiosurgery from March 2001 to October 2006. The software extracted 510 patients who underwent stereotactic radiosurgery, and the following information was recorded for each patient: treatment technique, treatment time (from initiation of imaging, if done, to completion of therapy), number of isocenters, number of fields, total monitor units, and dose rate. RESULTS Of the 510 patients, 395 were treated with DCA therapy and 115 with static conformal beams. The average number of isocenters treated was 1.06 (range, 1-4). The average times to deliver treatment were 24.1 minutes for patients who underwent DCA therapy and 19.3 minutes for those treated with static conformal beams, reflecting the lack of imaging in the latter patients. Eighty percent of patients were treated in <30 minutes. For the patients who underwent DCA therapy, the times required to treat 1, 2, 3, and 4 isocenters were 23.9, 24.8, 33.1, and 37.8 minutes, respectively. Average beam-on time for these patients was 11.4 minutes. There has been no significant reduction in treatment delivery with the use of 1,000 monitor units/min, reflecting the fact that beam-on time is not the major determinant of overall treatment time. CONCLUSIONS Multileaf collimator-based linac radiosurgery can be delivered efficiently in <30 minutes in the vast majority of patients. Given the limited treatment room utilization required for stereotactic radiosurgery treatments, this study calls into question the need for a dedicated radiosurgery unit for even busy treatment centers.

PET Lesion Segmentation Using Automated Iso-intensity Contouring in Head and Neck Cancer

To improve the objectivity of the integration of positron emission tomography (PET), we used the conformality index (CI) to measure the goodness of fit of a given PET iso-SUV (standardized uptake value) level with the GTV defined on PET (GTVPET) and CT (GTVCT). Twenty-two datasets involving 20 head and neck cancer patients were identified. GTVPET and GTVCT were delineated manually. An iso-intensity method was developed to automatically segment GTVPET-ISO using (a) SUV and (b) maximum intensity thresholding (%Max), over a range of intensities. For each intensity, GTVPET-ISO was compared to GTVPET using the conformality index CIPET (and, similarly, to GTVCT using CICT). Comparing GTVPET to GTVPET-ISO vs comparing GTVCT to GTVPET-ISO, the average peak CI was 0.68 ± 0.09 vs 0.49 ± 0.12 (p<0.001), the optimum iso-SUV was 2.7 ± 0.7 vs 2.9 ± 1.0 (p=0. 253), and the %Max SUV was 21.8% ± 7.6% vs 23.8% ± 8.6% (p=0. 310), respectively. The radiation oncologists volumes corresponded to a lower iso-SUV (3.02 ± 0.58 vs 4.36 ± 0.77, p < 0.001) and lower %Max SUV (24.1 ± 9.1% vs 34.3 ± 11.2%, p<0.001) than those drawn by the nuclear medicine physician. Though manual editing may still be necessary, PET iso-contouring is one method to improve the objectivity of GTV definition in head and neck cancer patients. Iso-SUVs can also be used to study the differences between PETs role as a nuclear medicine diagnostic test versus a radiation oncology treatment planning tool.

WE‐D‐BRB‐09: Quality Assurance (QA) for Deformable Registration: Measures for Evaluating Displacement Fields in Clinical Settings

Purpose: Deformable registration is an essential tool in adaptive radiotherapy, as it accounts for anatomical changes during treatment. In recent years, research has focused on proposing different deformable registration algorithms and inter‐comparing their results in academic settings. We contend that finding an efficient method for quality assurance of deformable registration in clinical settings is crucial for a global acceptance of adaptive radiotherapy. This study proposes measures derived from computational fluid dynamics as a simple and efficient tool to quantify a displacement field. Method: Our aim was to develop quantitative metrics of registration quality designed for routine use that are algorithm‐independent, labor‐efficient, and accurately identify errors in a given displacement field. The quality assurance (QA) framework identifies unrealistic anatomical motion through vortexes in the displacement field as detected using the CURL operator and presented as a vortex map overlaid on the original anatomy for a quick identification of problematic regions. Regions of compression/expansion are identified through the determinant of the Jacobian matrix. The warp energy measure is proposed as a global measure of displacement field smoothness. Results: The new evaluation approach was tested on numerous inter and intra patient cases using both single and multi‐modality registration algorithms. The CURL operator quantitatively detected errors in the displacement field and identified problematic regions that were invisible to classical voxel‐based evaluation methods. Warping emerges above 1 indicated unrealistic displacement fields. Conclusions: The proposed QA framework for deformable image registration provides increased usability and accuracy in detecting unrealistic warping over classical registration assessment methods. It is computationally efficient and provides a valuable platform for the clinical acceptance of adaptive therapy in the future.