Browsing by Author "Craciunescu, Oana"

Purpose

Materials and methods

Results/conclusions

Treatment errors can occur in high dose rate (HDR) brachytherapy, but, currently, there is a lack of real-time treatment verification systems that are clinically available and thus many errors are only determined retrospectively, or not at all. Due to the rapid dose fall-off in HDR brachytherapy, small uncertainties can result in large dose variations. These errors can have a large impact on the patient if not detected during the treatment. In-vivo dosimetry is one potential way to detect these errors in real-time. The purpose of this work is to explore further the use of a nano-fiber optic detector (NanoFOD) for real-time dosimetry during HDR brachytherapy treatments. The NanoFOD consists of an inorganic nano-crystalline scintillator fixed on an optical fiber. It is small enough to be placed into clinical catheters and needles, allowing for in-vivo measurements, and is able to measure point doses to sub-millimeter resolution. Previous studies proved the feasibility of using the NanoFOD for real-time measurements in both cylinder and T&O HDR treatments. The purpose of this work is to: (1) determine a way to calibrate the NanoFOD at small source to detector distances, with the end goal to make the NanoFOD usable in more types of HDR deliveries; (2) test the feasibility of using the NanoFOD to measure real-time, in-vivo dose measurements during an US-based HDR prostate phantom end-to-end-test; (3) design and test a novel platform to allow for real time tracking of HDR treatments.

In this thesis a new calibration technique was developed and tested for the calibration of the NanoFOD at short distances from an Ir-192 HDR source. This calibration provides a way to convert the real-time measured voltage to a dose-rate-to-water value for a range of source-to-detector distances, beginning at 0.5 cm, allowing for more accurate dose measurements in HDR brachytherapy applications in which there are typically small volumes, such as prostate HDR, or any other interstitial type implant. Verification of this calibration found accuracy within 7% of the expected cumulative dose values, with a potential uncertainty being the accuracy with which we can position the NanoFOD versus the known location of the source. A second potential uncertainty is the assumption that the entire NanoFOD is water equivalent; when in reality some components are not. Preliminary Monte Carlo simulations were able to determine a relative dose rate value when the NanoFOD material was present, with results indicating the NanoFOD impacts the calculated dose.

This NanoFOD calibration was then used to test the ability of the fiber to take real-time measurements during an HDR prostate phantom end-to-end. Dose differences between the planned and measured cumulative doses when using flexi needles were found to be approximately 18%. This differences between the measured and planned dose values include both uncertainties in the NanoFOD system, as well as US-based HDR prostate brachytherapy workflow uncertainties, and are on the same order of magnitude with other reported in-vivo systems (Mason, Mamo, Al-Qaisieh, Henry, & Bownes, 2016). Through a Type A and Type B uncertainty analysis, the overall uncertainty of the dose measurements achieved by the NanoFOD was determined to be 15.7% and the uncertainty in the measured TPS dose values for HDR prostate treatments was determined to be 15.4%.

For real time tracking of delivered dose during HDR treatments, a novel platform using the NanoFOD detector was designed. The designed prototype interface was successful in displaying in real-time the measured voltage, correctly accounting for background noise, and also automatically detecting the location of the first dwell to start the overlay with the expected signal. When large errors were introduced into a phantom delivery treatment, the interface detected large differences between the measured and expected signals. This platform provides real-time feedback to the user, allowing for in-vivo verification of the dose delivered to the patient, as compared to expected values imported from the TPS. In doing this, the interface has the ability to help identify and reduce potential treatment errors in real-time that could otherwise remain undetected.

In recent years, there has been a decline in the use of brachytherapy for the treatment of several cancers. One of the several reasons attributed to this is the unease in implementing the procedures at the initial stage of the practice due to limited practice. These procedures are performed by the radiation oncologist that acquire their training during their residency. Although the period of residency exposes them to the various steps of the procedure by observing treatments of patients with different pathologies, it is evident that being able to simulate these skills on a phantom will be an added advantage. Specifically, due to the complex nature of the HDR procedure for cervical brachytherapy, it requires to a great extent exposure to hands-on training. This requirement established the need to develop a modular phantom with which various intracavitary and/or interstitial HDR GYN brachytherapy procedures will be done became apparent. To address this need, a brachytherapy prototype kit was designed by our group1. This kit was specifically designed to focus on High-Dose Rate (HDR) cervical cancer therapy. The treatment involves the use of applicators such as tandem and ovoids and/or needles. Dimensional and pathological variability were investigated by performing a 50-patient study after which a kit was designed that comprised of a vaginal canal, uteri (small, average, and large), High-Risk Clinical Target Volumes (HRCTV) and organs of interest (rectum and bladder). Upon the completion of the prototype kit, some limitations became apparent. The width of the vaginal canal was not wide enough to accommodate the various sizes of the ovoid caps. In addition, the original design incorporated the training kit in a small box hence making its use a bit challenging l. These limitations led to the designing of an improved modular training phantom. The goal of this project is to construct a modular brachytherapy phantom by integrating an improved version of the prototype kit with an existing gynecologic phantom, Gaumard (S400.100). First, we investigated the feasibility of integrating the existing phantom with our in-house kit. This was done by acquiring Computed Tomography (CT) images of the existing GYN phantom and contouring out a model using the 3D Slicer 4.10.2 application. Following this, parts from the initial kit were evaluated with respect to potential location in the existing phantom as well as general fit. The next step was to design new parts of the modular training kit and integrate them with the existing gynecologic phantom. The main part was the pelvic insert which comprised a combination of the vaginal canal and the pelvic floor. This newly designed anatomical part was printed in multi-material Agilus with a shore value of 30A. In addition, a frame was built to secure the training kit in the phantom. The modular connectivity of the uterus to the cervix of the vaginal canal had to be re-designed. The peg locking mechanism used in the prototype had a limitation. The uterus had to be Boolean subtracted from the cervix of the vaginal canal and this hole was the connecting point of the vaginal canal to the uterus. The peg was then used to keep both parts together. However, the size of the large uterus made it difficult to perform the Boolean subtraction as the cervix area was too small. This led to the designing of a new locking mechanism using flanges. The modular phantom was evaluated by a radiation oncology resident to assess usability and ease of use. The feedback is a resource in improving the phantom to fit their training needs. In conclusion, the brachytherapy modular phantom was built by integrating our in-house brachytherapy prototype kit with the GYN phantom.

In radiation oncology, 18F-FDG Positron Emission Tomography (PET) is used for determining metabolic activity of cancers as well as delineating gross tumor volumes (GTV) for treatment planning. More recently, PET is being utilized for adaptive therapies for gynecological malignancies in which tumor response may be estimated and treatments adjusted during the course of radiation. In addition to treatment assessment, 18F-FDG PET has become a tool in the prediction of tumor response because of the derived Standard Uptake Value (SUV), a measure of the metabolic activity of a tumor. In this study, we seek to establish texture analysis as complimentary to SUV for predicting tumor response as well as understanding temporal changes during treatment in gynecological cancers. An additional experiment was performed studying the variability of texture features from baseline and intra-treatment PET scans due to reconstruction parameters in order to identify features that show statistically significant changes during treatment and that are independent of reconstruction parameters.

In this IRB approved clinical research study, 29 women with node positive gynecological malignancies visible on PET including cervical, endometrial, vulvar, and vaginal cancers are treated with radiation therapy. Prescribed dose varied between 45-50.4Gy, with a 55-70Gy boost to the PET positive nodes. A baseline, intra-treatment (between 30-36Gy), and post-treatment PET-CT were obtained with tumor response determined by a physician according to post-treatment RECIST. All volumes were re-contoured on the intra-treatment PET-CT. Primary GTVs were segmented both with the 40% SUVmax threshold method and a validated gradient-based contouring tool, PET Edge (MIM Software Inc., Cleveland, OH). A MATLAB Graphical User Interface (GUI) called Duke FIRE (Functional Imaging Research Environment) was developed for this study in order to calculate four mathematical algorithms representing the spatial distribution of pixels in an image: gray level co-occurrence matrix (GLCM), gray level run length matrix (GLRLM), gray level size zone matrix (GLSZM), and the neighborhood gray level difference matrix (NGLDM). Features representing characteristics of the image are derived from these texture matrices: 12 local features from the GLCM, 11 regional features from the GLRLM, 11 regional features from the GLSZM, and 5 local features from the NGLDM. Additionally, 6 global SUV histogram features including SUVmean, SUVmedian, SUVmax, skewness, kurtosis, and variance as well as metabolic volume (MV) and total lesion glycolysis (TLG) are extracted. The prognostic power of each baseline feature derived from both gradient-based and threshold segmentation methods was determined using the Wilcoxon rank-sum test. Receiver operating characteristic (ROC) curves were calculated to understand the sensitivity and specificity of baseline texture features compared to SUV metrics. Changes in features from baseline to intra-treatment PET-CT were determined using the Wilcoxon signed-rank test. A subset of 7 patient baseline and intra-treatment raw PET data was reconstructed 6 times using a TrueX+TOF algorithm on a Siemens Biograph mCT with varying iterations and Gaussian filter widths. Texture features were derived from the GTV as before. Texture features per patient were normalized to the respective clinical baseline value in order to limit variability to reconstruction parameters. Mean percent ranges of each feature at baseline and intra-treatment were determined and the change in features was compared using the Wilcoxon signed-rank test.

Of the 29 patients, there were 16 complete responders, 7 partial responders, and 6 non-responders. Comparing CR/PR vs. NR for the gradient-based GTVs, 7 texture values had a p < 0.05. The threshold GTVs yielded 4 texture features with p < 0.05. ROC and logistic regression was performed and texture features from both PET Edge and thresholding yielding a higher area under the curve (AUC) than SUV metrics. Features derived from PET Edge GTVs also showed higher AUCs than the threshold GTVs. From baseline to intra-treatment, 16 texture features changed with p < 0.05. Texture analysis of PET imaged gynecological tumors is considerably more powerful than SUV in early prognosis of tumor response, especially when using a gradient based method.

We then took the 16 texture features showing significant changes (p < 0.05) between baseline and intra-treatment PET scans in 29 patients and tested these against the subset of reconstructed features to determine if these changes were dependent upon the method in which the scans were reconstructed. A total of 13 features (including entropy, zone non-uniformity, and complexity) were found to be consistently different even when subjected to different means of reconstruction, however 3 of the 16 (inverse variance, run percentage, and zone percentage) were found to be dependent upon these reconstruction parameters. Texture features such as entropy, zone non-uniformity, and complexity are excellent candidates for future investigations of changes in texture analysis during radiation therapy of gynecological cancers. Caution should be taken with inverse variance, run percentage, and zone percentage due to their dependence upon reconstruction parameters.

This comprehensive work characterizes gynecological cancers using texture analysis in order to identify texture features that may be used for predicting tumor response as well as reflecting changes during treatment. It is the first study to our knowledge that utilizes all 4 texture matrices (GLCM, GLRLM, GLSZM, and NGLDM) and found 7 statistically significant features classifying responding and non-responding gynecological tumors: energy, entropy, max probability, zone gray level non uniformity, zone size non uniformity, contrast (NGLDM), and complexity. A novel method was implemented extending the NGLDM and its respective features to 3D space for this study. It is also the first study concluding that a semi-automatic gradient-based segmentation method results in more, stronger predictors than using a 40% SUVmax threshold method. Finally, this is the first study to examine variability of texture features with reconstruction parameters and to identify texture features as reliable and independent of reconstruction. In conclusion, texture analysis is a promising method of characterizing tumors visible on PET and should be considered for future studies.

Introduction:In interstitial gynecological (GYN) high-dose rate (HDR) brachytherapy, the lack of real-time imaging during procedures is a recognized limitation, often resulting in inaccuracies in needle placement. Traditionally, clinicians rely on a combination of pre-treatment MRI scans and clinical examinations to guide needle placement, but this approach can lead to suboptimal outcomes. To address these challenges and improve the clinical workflow, this research introduces the utilization of a 3D-printed pelvic phantom designed to evaluate the feasibility and efficacy of intraoperative 3D transrectal-ultrasound (TRUS) guidance for needle insertion. Materials and Methods: This research developed needle-tracking techniques utilizing real-time TRUS guidance specifically tailored for GYN malignancies within the context of hybrid HDR brachytherapy. Leveraging the Eigen Health system, originally designed for prostate biopsy, we adapted its real-time imaging capabilities for GYN applications. To facilitate this, anatomical phantoms that replicate female pelvic anatomy were created and optimized. These phantoms served as the cornerstone for a series of feasibility tests that examined the integration of TRUS guidance into brachytherapy workflows. Additionally, comprehensive quality assurance (QA) protocols were established to ensure the reliability of needle tracking and characterization in relation to our ground-truth CT images. The system's MR/TRUS deformable registration algorithm was evaluated through two different experiments. Results: The execution of feasibility tests yielded promising results, demonstrating the Eigen Health system's compatibility with existing brachytherapy workflows and its potential to significantly improve needle placement accuracy. The study also highlighted the system's adaptability to accommodate multiple needle insertions and implement rod-based planning. Through quality assurance evaluations of the reliability and precision of needle characterization, a scaling issue within the needle-tracking system was discovered. Furthermore, the MR-3DUS surface-based elastic registration process was thoroughly examined, with uncertainty variables acknowledged, the preliminary results of this experiment showed that the system can adapt to anatomical variation. Conclusion: The integration of real-time TRUS guidance into GYN HDR brachytherapy significantly improves needle placement accuracy, promising enhanced treatment efficacy and minimized complications. This research contributes valuable insights to medical physics and cancer treatment, paving the way for future clinical trials and potential integration into standard brachytherapy protocols.

Purpose

Material and methods

Results

Conclusions

Purpose Gynecological malignancies present challenges in determining an appropri- ate volumetric dose due to the highly variable physiologic activity of the surrounding tissue. Because of the high doses used in brachytherapy, surrounding structures have the potential to move around in the dose region and receive unknown amounts of radiation. Deformable image registration could overcome challenges in determining the true delivered dose through a dose accumulation process. This study uses two dose summation techniques to determine the efficacy of a deformation registration for the Syed applicators and cylinder applicators.

Methods and Materials Data for patients treated with a vaginal cylinder or Syed template were imported into the MIM software (Cleveland, OH). The bladder, rectum and applicator were contoured on each CT. The deformable registration was applied to structures only by masking other image data to a single intensity with the purpose of focusing the registration on the high dose area, as well as to minimize any uncertainty from the CT data. The deformable registration flow consisted of the following steps: 1) Using a different Hounsfield Unit, the CTs were masked so that each of the structures-of-interest (SOIs) had one unique intensity value; 2) Perform a rigid registration between two image sets with alignment based on the applicator position; 3) Perform a deformable registration; 4) Refine registration by using local manual alignment in area with large contour changes; 5) Repeat steps 1 to 3 to register the desired structure from all the subsequent fractions to the first fraction structure; 5) Transfer each deformed contoured to the first CT. The deformed structure accuracy was measured by a comparison to the first fraction using the dice similarity coefficient. Two dose summation techniques were investigated: a) deform the desired structure to one fraction and determine the fractional dose to these new structures, and b) deform the fractional dose to one CT and accumulate to find the total dose. Point doses, D2cc, were used as a comparison value for each method.

Results The cylinder set of patients had a DSC ranging from 0.82 to 0.96 for bladder and 0.82 to 0.94 for rectum. The contour deformation addition method has variations up to 23% from the initial clinical point dose for the cylinder applicators. The dose deformation accumulation method gave up to a 7.1% difference from the clinical point dose. The Syed applicator patient set has DSC ranging from 0.53 to 0.97 for the bladder and 0.75 and 0.95 for the rectum. These registrations' dose additions varied up to 34.78% and the dose deformation accumulation varied up to 34.97%.

Conclusions With the changing anatomy in brachytherapy, deforming the dose with the end point of dose summation leads to different volumetric doses then when dose is recalculated on deformed structures, raising concerns about the accuracy of the deformed dose. Dice Similarity Coefficients alone cannot be used to establish the accuracy of a deformation for brachy dose summation purpose.

This thesis consists of two studies. Study 1 is an assessment of dose-volume metrics of an 18F-FDG PET adaptive radiation therapy for vulvar and cervical cancer patients.Study 2 is an evaluation of cumulative dose distributions from external beam radiation therapy using CT-to-CBCT deformable image registration (DIR) for cervical cancer patients.

Study 1: Assessment of dose-volume metrics of an 18F-FDG PET adaptive radiation therapy for vulvar and cervical cancer patientsPurpose: Adaptive radiation therapy (ART) enables treatment to be modified with the goal of improving the dose distribution to the patient due to changes in anatomy. Fluorodeoxyglucose positron emission tomography and computed tomography (FDG-PET/CT) is used for staging, treatment planning, and assessing treatment response, but can also be used to adapt treatment. In an adaptive PET/CT study, an additional PET/CT scan is acquired for planning purposes after a certain prescribed dose has been delivered. The intratreatment PET/CT is used to re-contour the volumes and create a new treatment plan that is used to deliver the remaining dose for the treatment. The goal of adaptive radiation therapy (ART) is to reduce the dose to normal tissues while maintaining the prescribed dose to the adapted PTV. Materials and Methods: In this IRB-approved protocol, patients with vulvar and cervical cancer received a planning PET/CT and an intratreatment PET/CT. Radiation therapy consisted of either intensity modulated radiotherapy (IMRT) or volumetric modulated arc therapy (VMAT) with 1.8 Gy once daily to a total of 45 to 50.4 Gy and simultaneous integrated boosts (SIB) to involved pelvic or para-aortic (PA) lymph nodes. The primary tumor was treated to 64.4 to 66.4 Gy with sequential boosts for the vulvar cancer patients. Cervical cancer patients were boosted with brachytherapy. SIB dose ranged from a total of 64.4 Gy to 66.4 Gy in 25 fractions determined by the treating physician and organs-at risk (OAR) tolerance. An intratreatment PET/CT was obtained at 12-20 fractions when the delivered dose was between 30 to 36 Gy. All patients were re-planned with revised OAR, gross tumor volume (GTV) and planning target volume (PTV) contours. The same dose goals remained on the adapted plan. Dosimetric metrics for OARs were compared using the Wilcoxon signed rank test. The criteria for determining statistical significance was established as a p-value less than 0.05. Results: In the vulvar analysis, out of 20 eligible patients, ART resulted in significant reductions in OAR doses. For bladder, max dose (Dmax) median reduction (MR) was 1.1 Gy ((IQR 0.48 – 2.3 Gy), p < 0.001) and for D2cc MR was 1.5 Gy ((IQR 0.51 – 2.1 Gy), p < 0.001). For bowel, Dmax MR was 1.0 Gy ((IQR 0.11 – 2.9 Gy), p < 0.001), for D2cc MR was 0.39 Gy ((IQR 0.023 – 1.7 Gy), p < 0.001), and for D15cc MR was 0.19 Gy ((IQR 0.026 – 0.47 Gy), p = 0.002)). For rectum, mean dose (Dmean) MR was 0.66 Gy ((IQR 0.17 – 1.7 Gy) p = 0.006) and for D2cc MR was 0.46 Gy ((IQR 0.17 – 0.80 Gy), p = 0.006). Thirty-seven cervical patients were analyzed. ART resulted in significant reductions in OAR doses. For bladder, max dose (Dmax) median reduction (MR) was 0.89 Gy ((IQR 0.23 – 2.14 Gy), p = 0.001) and for D2cc MR was 0.38 Gy ((IQR 0.12 – 1.36 Gy), p<0.0001). For bowel, Dmax MR was 3.27 Gy ((IQR 0.50 – 5.41 Gy), p < 0.0001). For D2cc MR was 2.09 Gy ((IQR 0.30 – 4.97 Gy), p < 0.0001), and for D15cc MR was 0.57 Gy (IQR 0.22 – 2.07 Gy)). For rectum, Dmean MR was 0.13 Gy ((IQR 0.09 – 0.24 Gy) p = 0.0025), and for D2cc MR was 0.44 Gy ((IQR 0.14 – 1.02 Gy), p < 0.0001). Conclusions: Based on the analysis and response to ART of 20 eligible patients with vulvar cancer and 37 eligible patients with cervical cancer, it can be concluded that ART resulted in a significant reduction in OAR doses, including bladder, bowel, and rectum. Overall, these findings suggest that ART can effectively reduce the radiation dose to OARs and improve treatment outcomes for patients with gynecological cancers.

Study 2: Evaluation of cumulative dose distributions from external beam radiation therapy using CT-to-CBCT deformable image registration (DIR) for cervical cancer patients Purpose: Organ motion during radiation therapy in the pelvic region can potentially lead to uncertainties with the dose delivered to critical organs during fractionated treatment. The purpose of this study is to investigate, by means of using deformable image registration (DIR) and dose summation techniques, the differences between the planning dose and the delivered dose as calculated from the longitudinal cone-beam CT (CBCT) images for cervical cancer patients. Materials and Methods: Cervical cancer patients treated with external beam radiation therapy (EBRT) received a planning CT (pCT) and five CBCTs, once every five fractions of treatment. The “Merged CBCT” feature in MIM Maestro (MIM Software, Cleveland, OH) was performed between the pCT and each CBCT to generate an extended field-of-view (FOV) CBCT (mCBCT). A free-form multi-modality DIR was then performed between the pCT and the mCBCT to deform the pCT structures onto the mCBCT. DIR-generated bladder and rectum contours were further adjusted by a physician, and Dice Similarity Coefficients (DSC) were calculated between the two. After deformation, the investigated doses on the mCBCT were: 1) recalculated in Eclipse TPS (Varian Medical Systems, Palo Alto, CA) using original plan parameters (ecD), and 2) deformed from planning dose (pD) using the deformation matrix (mdD). Dose summation was performed to the first week’s mCBCT. Bladder D2cc, Dmax, Dmean, V45, and D50, rectum D2cc, Dmax, Dmean, and D50, and PTV45 D90 and D98 were compared between the three calculated doses. Dose distributions were compared in terms of dose volume histograms (DVHs) and gamma analysis. The Wilcoxon signed rank test was used to compare dosimetric metrics with statistical significance defined at p < 0.05. Results: For the ten patients analyzed, the average DSC were 0.72 ± 0.15 for bladder and 0.80 ± 0.11 for rectum. For most cases, only the superior and inferior slices were edited by physician. Regardless of the method of dose calculation (ecD or mdD), D2cc (bladder and rectum), and D90 and D98 (PTV45) were within 5% of pD for at least 9 out of 10 patients. For one patient each for bladder, rectum, and PTV45, the agreement was worse than 5%, with the largest difference of 15.3% for bladder D2cc in a patient with large bladder filling differences. For the Eclipse calculated dose on the merged CBCT (ecD) and t;he MIM deformed dose on merged CBCT (mdD), the bladder Dmax was within 5% for 8 out of 10 patients, and rectum Dmax was within 5% for 7 out of 10 patients. All 10 patients for ecD and mdD were >5% for bladder V45 due to the large variations in bladder volume throughout treatment. Statistically significant differences for bladder D2cc between the ecD and the mdD (p = 0.047). For bladder D50, significant differences between pD and ecD (p = 0.009) and ecD and mdD (p = 0.005). Statistically significant differences for rectum D2cc between the pD and ecD (p = 0.028) as well as ecD and mdD (p = 0.005). Statistically significant differences for D98 between the pD and ecD (p = 0.028) and pD and mdD (p = 0.007). The gamma analysis between the ecD and pD matched 90% of the voxels for 3 out of 10 patients and between the mdD and pD for 1 out of 10 patients. Conclusions: In this study, we evaluated cumulative doses based on weekly CBCTs using a commercially-available DIR software. Using DIR and the new Merged CBCT feature, we determined that reporting the initial planning dose would not introduce a more than 5% difference in 90% of cases studied. Our results indicate that the mdD produces similar dose values as the ecD for the OARs and PTV. The proposed workflow should be used on a case-by-case basis when the weekly CBCT shows marked difference in organs-at-risk from the planning CT.

The utilization of brachytherapy practice in clinics has been declining over the years. The decline has been linked to a variety of factors including a lack of training opportunities. To improve the quality of intracavitary and interstitial HDR brachytherapy education, a multi-material modular 3D printed pelvic phantom kit prototype simulating normal and cervix pathological conditions has been developed. This comprehensive training phantom is intended to serve as a novel aid in the “300 in 10 Strategy” put forward by the American Brachytherapy Society which calls to train 30 competent brachytherapists per year over the next 10 years.

Patient anatomy was derived from anonymized pelvic CT and MRI scans from different representative patients who had been diagnosed with cervical cancer. The dimensions of patients’ uterine canal sizes and uterine body sizes were measured and used to construct a variety of uteri based off of the averages and standard deviations of the subjects in our study.

The length of the uterine body was measured from top of the fundus to the top of the cervix. The width of the uterine body was measured at the top of the cervix and also measured across the midpoint between the cervix and the fundus. High risk clinical target volumes (HR-CTV) were also extracted from these patients. 3D Slicer (Slicer 4.10.1) was used to import and convert contoured DICOM data into a 3D model. Organs of interest for our prototype include the vaginal canal, uterus, and cervix. A standard rectum, and bladder were also printed.

Individual STL files were imported into Autodesk Meshmixer (3.5.4) and manipulated to include more representative features such as hollowed out cavities and canals. Modular components of the phantom were designed and integrated into patient anatomy using 3D modeling software Shapr3D (Stratasys, 3.35.0).

Flexible and rigid materials were assigned to each component of the phantom. Vero Clear (Stratasys) was assigned to rigid design components including modularity connections. Agilus30 (Stratasys) was assigned to the more flexible components including the vaginal canal, uterus, rectum, and bladder. Each flexible component was assigned a shore hardness value ranging from 30-80 to further individualize the level of flexibility. The finalized prototype was printed using a Stratasys J750 PolyJet printer.

The prototype kit consists of four uteri. The three anteverted uteri in the kit are based on the smallest, the average, and the largest dimensions from our patient set. The fourth uterus is retroverted and uses average dimensions. The four uteri incorporate two embedded HR-CTVs through color staining in the uterine body prints. Four clip-on HR-CTV sections that expand outside the cervix and uterine body are also part of the kit to mimic different pathology. All uterus bodies and the vaginal canal are printed using clear Agilus (shore 30a), and the HR-CTVs are printed externally and into the uterine bodies using a blend of colored Vero and Agilus (shore 40a) as a means of evaluation for tandem/ovoids and needle placement. The bladder surface and rectum are printed in Agilus, shore 35a and 70a, respectively. The outer box was printed using Vero Clear, shore 90. The full kit which consists of an outer box, vulvar entrance, vaginal canal, four uteruses, two embedded HR-CTVs, four clip-on HR-CTVs, a standard bladder, and standard rectum, costs $631 in printing material expenses. This low-cost comprehensive training kit may be used to improve resident comfort in performing gynecological brachytherapy procedures.

The thesis has two parts. The first project relates to real time dosimetry in Gynecological (GYN) High Dose Rate (HDR) Brachytherapy Treatments and the second part investigates a new workflow for Image Guided Brachytherapy (IGBT) for centers that do not have access to MRI with applicator in-situ.

Project 1: Real-time Dosimetry in HDR Brachytherapy using the Duke designed NanoFOD dosimeter: Limits of Error Detection in Clinical Applications

Purpose: Previously, an optical fiber radiation detector system was shown to be capable of identifying potential high dose rate (HDR) brachytherapy delivery errors in real time. The purpose of this work is to determine this detector’s limits of error detection in vaginal cylinder and tandem and ovoid-type HDR gynecological brachytherapy treatments.

Method and Materials: The system consists of a scintillating nanoparticle-terminated fiber-optic dosimeter (NanoFOD) and a LabView platform (Versions 2015 and 2017; National Instruments Corporation, Austin, TX) which displays the real-time voltages measured by the NanoFOD during HDR treatment delivery. The platform allows for the measured voltage to be overlaid on the expected detector signal. To test the limitation of error detectability in vaginal cylinder brachytherapy, the NanoFOD was taped 1.5-cm from the tip of a 3-cm diameter Varian manufactured stump cylinder. This setup was

imaged using CT to localize the NanoFOD, and a plan was generated based on one of the institutional 3-cm stump cylinder templates with 9 dwell positions. For each dwell position, the expected voltage and doses were calculated using the dose distribution exported from the treatment planning system (TPS), a previously obtained distance-based calibration curve for the NanoFOD and TG-229 along and away table. The voltage values were imported to the LabView for real time monitoring. With a known location of the NanoFOD tip, the expected doses from the NanoFOD at each dwell position were calculated for all applicators for: 1) the clinical plan; 2) wrong source guide tube (SGT); 3) wrong cylinder; 4) wrong treatment template; 5) wrong step size, as well as 6) incorrect positioning of the cylinder insert. Measurements of voltage from each delivered plan were converted to dose per dwell and compared to the expected dose values.

In addition to experiments, a simulation-based limit of error detection with cylinder was studied and the results obtained from simulation were compared to experimental results. Simulation was done by modifying the applicator parameters and the incorrect plans were generated for 1) wrong SGT lengths 5-mm to 70-mm longer than correct length with a 5-mm interval; 2) wrong cylinder; 3) wrong treatment template; 4) wrong step size, as well as 5) incorrect positioning of the cylinder insert. Doses per dwell were also calculated and compared to the doses from correct plan.

Furthermore, simulations-only for T&O plans were also performed. Two representative clinical T&O-based plans with extra needles were used to establish the simulation-based limits of error detectability of the NanoFOD system. With a known location of the NanoFOD tip, the expected doses from the NanoFOD at each dwell position were calculated for all applicators for: 1) the clinical plan; 2) wrong treatment lengths (TL); 3) wrong connection to afterloader; 4) wrong digitization direction; and 5) wrong step size.

From previous work, it has been determined that the overall uncertainty in dose of this system in HDR brachytherapy is 20%. The percent difference (PD) between expected and measured doses and, between the simulated doses from correct and incorrect plans was calculated and compared to 20% to determine if a certain potential error can be detected with the NanoFOD system.

Results: For the cylinder experiments, when the correct treatment was delivered, the median PD over 9 dwells between measured and expected dose from each dwell was 11%. When the wrong SGT, wrong cylinder size, or wrong clinical template was used, the system detected PD values of -91.6%, -71.4% and -29.6% at the first dwell. With wrong step, the system showed large discrepancies starting at the third dwell position (58.6%). When incorrect cylinder insert length was 0.5cm, 1.0cm and 1.5cm, the significant PD values captured were -29.4%, -28.1% and -22.5% at 3rd, 2nd and 1st dwell, respectively.

In TPS simulations for the cylinder applicator, PD from wrong SGT, wrong cylinder size and wrong template and were -93.6%, -90.8% and -57.8% at first dwell. For wrong step size, the simulation predicted the PD to be greater than 20% (47.2%) starting at first dwell. For all incorrect insert length, simulation predicted greater than 20% PD (-38.6%, -34.9%, -38.6%) at first dwell. For incorrect SGT lengths range from 5mm to 70mm, simulation-based PD was from -28.0% to -82.3% at the first dwell, indicating the NanoFOD caught this error when the SGT was only 5-mm longer.

In TPS simulations with two T&O-based plans, with incorrect treatment length (TL, 5mm-70mm), the PD ranges for the two patients were 26.4%-26.8%, 20.4%-30.9%, and 30.9%-57% in 10mm TL tandem, 5mm TL ovoid and 10mm TL needle at first dwell. PD values were -44.1% to -36.4%, -75.3% to 96.4%, and -39.6% to 298.1% respectively at the first dwell when connections between tandem and left ovoid, left needle and left ovoid, left needle and tandem were swapped. When the two needles were switched, PD values were 43.1% and 63.6% at the 1st and 3rd dwell for patient A and patient B. Switching connection between the two ovoids did not produce significant PD for patient A because the two ovoids were identically loaded. For patient B, the PD was 87.5% when ovoids were swapped. Incorrect direction of digitization made the signal too low to be detected, an indication that treatment should be stopped. With wrong step size in tandem, ovoid and needle, simulation-based PD range in dose for two patients were 25.8%-34.9%, 34.3%-37.4%, and 20.7%-81.5%, respectively at 3rd, 2nd and 1st dwell for both patients.

Conclusion:

Using the 20% error detection threshold, the NanoFOD system was able to detect errors in real time cylinder tests when the wrong cylinder size, wrong SGT and wrong template were used, all at the first dwell position. For subtle errors such as small incorrect catheter insertions, the real-time system can detect errors starting with the second or third dwell position. The Labview interface is a good tool to use for real time tracking of the delivery.

TPS simulation predicted comparable discrepancies to PD obtained from experimental measurements when wrong cylinder size and wrong source guide tube was used. The disagreement between simulation-based PD and experiment-based PD is due to uncertainty in positioning during the experiment, as the nanoFOD in current form does not have a radio opaque marker to help with easy identification. TPS simulation was thus used for more complex applicators, knowing that simulations and experiments match.

With the 20% PD as an indication of wrong treatment in T&O-based HDR cases, the NanoFOD can capture all simulated gross errors within the first few dwells into the treatment, indicating it is capable of real-time verification of T&O HDR brachytherapy. Although the two T&O plans were different, the NanoFOD was able to capture the errors at similar dwell positions regardless of the difference in planning. Overall, the NanoFOD can catch, in real time, potential gross errors in clinical HDR.

Project 2: Investigation of Contour-based Deformable Image Registration (cbDIR) of pre-brachytherapy MRI (pbMRI) for target definition in HDR cervical cancer treatments

Purpose: While MRI-guided brachytherapy has been considered the gold standard for IGBT treatment for cervical cancer, practical factors such as cost, MRI access and clinical flow efficiency have limited the use of MRI for brachytherapy treatments. However, a pre-brachytherapy MRI (pbMRI) is usually taken before the brachytherapy to assess the tumor regression after external beam treatments. The purpose of this project is to investigate the role of contour-based deformable image registration of the pbMRI in target definition for high dose rate (HDR) cervical cancer treatments.

Method and Materials: Thirty-five patients with locally advanced cervical cancer treated at Duke University with Tandem and Ovoids (T&O) applicators were studied. A pbMRI was acquired for all patients. Each patient also had MRI and CT taken with applicator in situ at the time of the first fraction. The uterus structure contoured on pbMRI was deformed using contour-based (cb) and hybrid contour-based (hcb) DIR to the same structure contoured on the CT of first HDR fraction (MIM Software Inc., v 6.7.3 and v 6.8.11). The deformation matrix was used to deform the dHRCTV, which was then compared with the ground truth HRCTV (gtHRCTV) obtained on the MRI of 1st HDR fraction. Dice Similarity Coefficients (DSC) and Hausdorff Distance (med_HD, max_HD) were used to evaluate the overlap and mismatch at boundaries between the deformed HRCTV and gtHRCTV using both DIR methods. The process of generating dHRCTV requires user’s manual input. Therefore, inter- and intra-user variability using cbDIR, as well as the difference in the two DIR methods were investigated.

Results: The pbMRI scan was acquired on average 5.9+/-3.8 days before first HDR fraction. Median DSC for uterus was 0.9 (IQR 0.88-0.93) and 0.93 (IQR 0.92-0.94), and for HRCTV was 0.64 (IQR 0.52-0.71) and 0.68 (IQR 0.59-0.72) for cbDIR and hcbDIR, respectively. The median med_HD for uterus was 0.09cm (IQR 0.08-0.1cm) and 0.08cm (IQR 0.06-0.09cm). The median max_HD for HRCTV was 1.37cm (IQR 1.1-2.3cm) and 1.39cm (IQR 1.1-1.6cm) with cbDIR and hcbDIR respectively. Limited inter-user, intra-user, and DSC variability between DIR versions were found (p=0.53, p=0.77, p=0.18, respectively).

Conclusion: Contour-based DIR based on uterus/cervix structure is proven to be relatively user independent, improves when hybrid contoured based DIR algorithms are used, and is expected to serve as a good starting point for HRCTV definition for HDR planning in the absence of MRI with applicators in-situ.

Purpose: The purpose of this study was to assess the feasibility of using dynamic contrast-enhanced MRI to monitor early treatment-induced changes in pharmacokinetic (PK) parameters in head and neck cancer patients. The intrinsic variability of three parameters (Ktrans, v_e, and iAUC60) without treatment intervention was measured and compared to the treatment-induced variability.

Materials and Methods: Twenty patients were imaged while undergoing chemoradiation therapy (CRT) for head and neck malignancies. The imaging protocol included two baseline scans one week apart (B1, B2), and a third scan after 1-2 weeks of chemoradiation (ETX - early treatment). The images were acquired on a 1.5T scanner in the coronal plane with a temporal resolution of 10 sec. A population-averaged arterial input function was calculated from plasma concentration curves in both the left and the right carotids of each patient at each time point (B1, B2, ETX). The statistical significance of using a left/right AIF or a time-point-specific AIF was evaluated using Bland-Altman plots. To further ensure the correct calculation of PK parameters, the accuracy of the flip angles produced by the MR scanner was measured in phantoms and a volunteer. PK analysis was performed in iCAD (Nashua, NH) based on the modified Tofts model. This study focuses on two PK parameters used in the Tofts model (Ktrans, v_e), and one semi-quantitative parameter that was also calculated in iCAD using an uptake integral approach (iAUC60). Ktrans, v_e, and iAUC60 were averaged over regions of interest (ROIs), some of which covered primary tumors, and others of which covered known nodal metastases. Bland-Altman plots were used to describe the intrinsic variability in each parameter between the two baseline scans. The coefficient of repeatability (CR) between the baseline values was determined from the Bland-Altman plots and compared to the magnitude of the observed treatment-induced changes.

Results: The plasma parameters for the population-averaged AIF were a₁ = 27.1135 kg/liter, a₂ = 17.6486 kg/liter, m₁ = 11.7525 min-1, and m₂ = 0.2054 min-1. The use of a left- or right-sided AIF was determined to be unnecessary, as it did not give statistically different PK parameters than the population-averaged AIF. The use of a time-point-specific AIF was not necessary in most cases, though it may give more accurate results when Ktrans values are > 1 min-1. The flip angle tests revealed high inaccuracies at a flip angle of 5¡ã, so flip angles ¡Ü 5¡ã were not used in PK analysis. The intrinsic variability of Ktrans, v_e, and iAUC60 was very high. For nodes, the CRs from the B1-B2 Bland-Altman plots were 0.725 min-1 for Ktrans, 0.315 for v_e, and 18.15 mM-sec for iAUC60. The fractions of node ROIs which showed treatment-induced changes greater than the CR were 3 out of 14 for Ktrans, 3 out of 17 for v_e, and 7 out of 17 for iAUC60. For primaries, the CRs were 1.385 min-1 for Ktrans, 0.305 for v_e, and 62.85 mM-sec for iAUC60. The fractions of primary ROIs which showed treatment-induced changes greater than the CR were 0 out of 9 for Ktrans, 1 out of 11 for v_e, and 2 out of 12 for iAUC60.

Conclusions: A population-averaged AIF for head and neck was generated that accounts for differences in right vs. left carotids, day-to-day AIF fluctuations, and treatment-induced AIF changes. It is not necessary to use a side-specific or a time-point-specific AIF. When Ktrans is greater than 1 min-1, PK parameter accuracy may be improved with the use of a time-point-specific AIF. Using the average AIF, large intrinsic fluctuations were observed in ROI-averaged values of Ktrans, v_e, and iAUC60, making these parameters poor evaluators of early treatment response in head and neck cancer. Nodes were slightly more likely than primaries to show significant treatment-induced changes. Overall, the use of averaged MR-based PK parameters to assess early treatment response is limited and challenging. An analysis of voxel-based variability might be better suited to this task.

Purpose: To evaluate dosimetric impact of heterogeneity from applicators and patient tissue between plans treated with tandem and ovoid applicators and interstitial needles calculated with the traditional Task Group 43 formulism and a model-based dose calculation algorithm (MBDCA) approach - Acuros®BV (Varian Medical Systems (VMS), Inc). Methods: For this analysis, patients with locally advanced cervical cancer treated with high dose rate (HDR) brachytherapy with a prescription of 5.5 Gy to D90% high-risk clinical target volume (CTV-HR) were selected. Fletcher-Suit Delclos style tandem and ovoid (T&O) applicators and interstitial needles were used and stabilized by vaginal balloons filled with a 95% saline and 5% IsoVue solution. Imaging techniques for applicator delineation and planning was done using Acuity CBCT or AIRO CT, depending on the modality used at the time of treatment. Standard planning for TG43 based dose calculation consisted of manually digitizing applicators on the CT or CBCT images. For plans calculated with Acuros®BV, solid applicator renderings were used, and the interstitial needles were left unchanged due to no solid model being available. Dose to medium in medium Dm,m was reported for Acuros plans and compared to dose to water in water D w,w from TG43 plans. Equivalent dose in 2Gy per fraction (EQD2) values were reported for target structures and organs at risk (OAR) including bladder, sigmoid, rectum, bowel, and vagina. As an additional comparison we investigated dosimetric changes based on the imaging modality used (CBCT vs. CT) and the presence of interstitial needles versus plans with T&O applicators alone. To investigate correlation and regression between DCAs both linear regression and Bland Altman Analysis (BAA) was performed. Conclusions: Based on data presented in this retrospective study, the average ± standard deviation of dosimetric reduction in D90% was 4.33 ± 0.09% (0.41 ± 0.13 Gy) for CTV-HR and 4.12 ± 0.09% (0.21 ± 0.07Gy) for CTV-IR . The reduction to OARs D2cm3 was: 4.99 ± 0.15% (0.34 ± 0.17 Gy) for bladder, 7.87 ± 0.16% (0.20 ± 0.09 Gy) for rectum, 5.79 ± 0.17% (0.21 ± 0.09 Gy) for sigmoid, 6.91 ± 0.14% (0.25 ± 0.13 Gy) for bowel, and 4.55 ± 0.14% (0.46 ± 0.27 Gy) for vagina. This indicates that plans calculated using Acuros®BV showed a dosimetric reduction to all structures when compared to TG43. This trend is attributed to heterogeneities within the treatment volume including applicators, tissue, and presence of air.

MR-based planning provides many benefits to planning of the HDR brachytherapy component of treatments for cervical cancer patients but requires the acquisition of different imaging sets that need to be registered for planning. However, registration between images introduces unknown uncertainties which impact the dose metrics that are essential to assessing brachytherapy plans. The goal of this project is to determine the overall effects of uncertainties in image registration between 1) T1- and T2-weighted MR images, and 2) planning CT and T1-weighted MR. The study looks at a total of 60 treatment fractions from thirteen patients for tandem and ovoid treatments. Workflows were created to compare clinical registrations with an automated intensity-based registration. In evaluating the overall uncertainty budget due to registration, the two different contour sets from the two different registrations were compared by calculation of the percent differences in important dosimetric values for the treatment target and also organs at risk. The data collected show that for both T1-MR to T2-MR and T1-MR to CT registration the average deviation for the target was very close to 0 with a reasonable tight standard deviation within 3%. However, the average deviation for organs at risk is greater which may mean patients are receiving more dose to those organs at risk than indicated due to uncertainties of registration caused by shifts and deformation between CT and T1- and T2-weighted MR.