Engineering Papers⌕ Search

Engineering topics

Zhang, M

Publications and source records attributed to Zhang, M.

At least 19 records

SU-D-207B-07: Development of a CT-Radiomics Based Early Response Prediction Model During Delivery of Chemoradiation Therapy for Pancreatic Cancer

Purpose: Pilot study developing a CT-texture based model for early assessment of treatment response during the delivery of chemoradiation therapy (CRT) for pancreatic cancer. Methods: Daily CT data acquired for 24 pancreatic head cancer patients using CT-on-rails, during the routine CT-guided CRT delivery with a radiation dose of 50.4 Gy in 28 fractions, were analyzed. The pancreas head was contoured on each daily CT. Texture analysis was performed within the pancreas head contour using a research tool (IBEX). Over 1300 texture metrics including: grey level co-occurrence, run-length, histogram, neighborhood intensity difference, and geometrical shape features were calculated for each daily CT. Metric-trend information was established by finding the best fit of either a linear, quadratic, or exponential function for each metric value verses accumulated dose. Thus all the daily CT texture information was consolidated into a best-fit trend type for a given patient and texture metric. Linear correlation was performed between the patient histological response vector (good, medium, poor) and all combinations of 23 patient subgroups (statistical jackknife) determining which metrics were most correlated to response and repeatedly reliable across most patients. Control correlations against CT scanner, reconstruction kernel, and gated/nongated CT images were also calculated. Euclidean distance measure was used to group/sort patient vectors based on the data of these trend-response metrics. Results: We found four specific trend-metrics (Gray Level Coocurence Matrix311-1InverseDiffMomentNorm, Gray Level Coocurence Matrix311-1InverseDiffNorm, Gray Level Coocurence Matrix311-1 Homogeneity2, and Intensity Direct Local StdMean) that were highly correlated with patient response and repeatedly reliable. Our four trend-metric model successfully ordered our pilot response dataset (p=0.00070). We found no significant correlation to our control parameters: gating (p=0.7717), scanner (p=0.9741), and kernel (p=0.8586). Conclusion: We have successfully created a CT-texture based early treatment response prediction model using the CTs acquired during the delivery of chemoradiation therapy for pancreatic cancer. Future testing is required to validate the model with more patient data.

60 APPLIED LIFE SCIENCES↗

SU-F-T-134: Can We Use the Same Dose Constrains Learnt From Photon World to Plan Proton for Lung Cancer?

Purpose: To evaluate if the same DVH constrains used in photon plans can be safely used to plan proton therapy for lung cancer. Since protons and photons have different dose deposition patterns, the hypothesis is following DVH constrains derived from photon world is not safe for proton. Methods: We retrospectively evaluated plans for 11 lung cancer patients. Each patient was planned with photon and proton following the same dose constrains. Dose statistics on PTV, normal lung, heart and esophagus were extracted for comparison. gEUD for normal lung was calculated and compared between proton and photon plans. We calculated series of gEUDs for each plan by varying the parameter “a” in gEUD formula from 0.1 to 3, covering the whole confidence interval. Results: For all patients, proton plans yield similar PTV coverage and lower dose to heart and esophagus than photon plans. Normal lung V5 was 32.3 % on average in proton plans than 55.4 % in photon. Normal lung gEUD monotonically increased with increasing “a” for all proton and photon plans. For a given patient, the gEUD-proton(a) had a steeper slope than gEUD-photon(a). The two curves crossed for 8 out of 11 patients when “a” = [0.1, 3]. a-crossing ranged from 0.8 to 2.44 with an average of 1.15. For a<a-crossing, gEUD-proton was less than gEUD-photon and vice versa. Conclusion: The current clinical guideline is the lower normal lung V5 would associated with less complications. However, proton plans with a lower normal lung V5 could yield a higher gEUD than photon if the real “a” is larger than a-crossing. Since a-crossing was within the possible range of real “a”, simply following the normal lung V5 guideline for proton plan would not be a good practice. More comprehensive methods should be developed to evaluate the proton plan.

60 APPLIED LIFE SCIENCES↗

SU-F-T-179: Fast and Accurate Profile Acquisition for Proton Beam Using Multi-Ion Chamber Arrays

Purpose: Proton beam profile measurement is more time-consuming than photon beam. Due to the energy modulation during proton delivery, chambers have to move step-by-step instead of continuously. Multi-ion chamber arrays are appealing to this task since multiple measurements can be performed at once. However, their utilization suffers from sparse spatial resolution and potential intrinsic volume-averaging effect of the disk-shaped ion chambers. We proposed an approach to measure proton beam profiles accurately and efficiently. Methods: Mevion S250 proton system and IBA Matrixx ion chamber arrays were used in this study. Matrixx has interchamber distance of 7.62 mm, and chamber diameter of 4.5 mm. We measured the same beam profile by moving the Matrixx seven times with 1 mm each time along y axis. All 7 measurements were superimposed to get a “finer” profile with 1 mm spatial resolution. Coarser resolution profiles of 2 mm and 3 mm were also generated by using subsets of measurements. Those profiles were compared to the TPS calculated beam profile. Gamma analysis was performed for 2D dose maps to evaluate the difference to TPS dose plane. Results: Preliminary results showed a large discrepancy between the TPS calculated profile and the single measurement profile with 7.6 mm resolution. A good match could be achieved when the resolution reduced to 3 mm by adding one extra measurement. Gamma analysis for 2D dose map of a 10×10 field showed a passing rate (γ ≤ 1) of 90.6% using a 3% and 3mm criterion for single measurement, which increased to 92.3% for 2-measurement superimposition, and slightly further increased to 92.9% for 7-measurement superimposition. Conclusion: The results indicated that 2 measurements shifted by 3mm using Matrixx generated a smooth proton beam profile with good matching to Eclipse beam profile. We suggest using this 2-measurement approach in clinic for double scattering proton beam profile measurement.

60 APPLIED LIFE SCIENCES↗

SU-F-T-216: Evaluating Dosimetry Accuracy of a Treatment Planning System On Small Proton Fields

Purpose: This study is aiming to identify the smallest field size for which a treatment planning system (TPS) can accurately calculate the relative dose distribution. The finding would be used as a guideline to choose the smallest proton field for clinical treatment. Methods: Mevion S250™ double scattering proton delivery system and Eclipse™ TPS (Varian) with pencil beam convolution (PBC) dose algorithm were used in this study. Square sized fields of 1 cm, 2 cm, 3 cm, 4 cm, 5 cm, and 10 cm were planned on a cubical water phantom with iso-center placed at 10 cm depth. All beams used the same proton beam option: range 15 cm and modulation 10 cm. Dose in water was calculated without any compensator. Gafchromic™ EBT3 film and diode detectors were used to measure the central axis dose distribution and lateral dose profiles at 5 cm, 10 cm, and 14 cm depth. Results: The preliminary film measurement shows good agreement between Eclipse calculated lateral dose profiles for all tested field sizes. The differences on full width half maximum were ≤ 1 mm while the differences on the penumbras were between 1 mm and 2 mm between Eclipse and film. For the depth dose, Eclipse results matched well with film measurements for field sizes down to 2 cm{sup 2}. With smaller field size of 1 cm{sup 2}, Eclipse was able to predict the decreasing of SOBP due to the lack of lateral charged particle equilibrium in depth. However, it did not match the film measurement. Diode measurement results will be available at the time of presentation. Conclusion: The PBC dose algorithm in Eclipse can accurately calculate relative dose distribution in double scattered proton system for field size down to 2 cm{sup 2}.

60 APPLIED LIFE SCIENCES↗

SU-F-T-499: Anatomic Features for Selection of Electronic Tissue Compensation Radiotherapy in Early-Stage Breast Cancer Patients After Breast-Conserving Surgery

Purpose: To explore the correlations between anatomic features and dose-volumetric parameters in 3DCRT and eComp whole breast irradiation and identify the feasibility of anatomic parameters to predict the planning method selection. Methods: We compared the effectiveness between conventional three-dimensional conformal radiotherapy (3DCRT) and electronic tissue compensation (eComp) for whole breast irradiation. 3DCRT and eComp planning techniques were used to generate treatment plans for 60 whole breast patients, respectively. The planning goal was to cover 95% of the planning target volume (PTV) with 95% of the prescription dose while minimizing dose to lung, heart, and skin. Statistical analyses were performed between critical organ doses and patient anatomic features, i.e., central lung distance (CLD), maximal heart distance (MHD), maximal heart length (MHL) and breast separation (BS). Results: Comparing to 3DCRT plans, on the average, eComp treatment planning process was about 7 minutes longer, but resulted in lower lung V20Gy, lower mean skin dose, with similar heart dose. The benefits were more pronounced for larger breast patients. To keep the lung V20Gy lower than 20% and mean skin dose lower than 85% of the prescription dose, eComp was the preferred method for patients with more than 2.3 cm CLD or larger than 22.5 cm BS. Conclusion: The study results may be useful in providing a handy criterion in clinical practice allowing us to easily choose between different planning techniques to satisfy the planning goal with minimal increase in complexity and cost. This study was supported by National Natural Science Foundation of China (NO. 31420103915) and Chongqing Health and Family Planning Commission Project (2015MSXM012).

60 APPLIED LIFE SCIENCES↗

SU-G-201-16: Thermal Imaging in Source Visualization and Radioactivity Measurement for High Dose Rate Brachytherapy

Purpose: High Dose Rate (HDR) brachytherapy poses a special challenge to radiation safety and quality assurance (QA) due to its high radioactivity, and it is thus critical to verify the HDR source location and its radioactive strength. This study demonstrates a new method for measuring HDR source location and radioactivity utilizing thermal imaging. A potential application would relate to HDR QA and safety improvement. Methods: Heating effects by an HDR source were studied using Finite Element Analysis (FEA). Thermal cameras were used to visualize an HDR source inside a plastic applicator made of polyvinylidene difluoride (PVDF). Using different source dwell times, correlations between the HDR source strength and heating effects were studied, thus establishing potential daily QA criteria using thermal imaging Results: For an Ir1?2 source with a radioactivity of 10 Ci, the decay-induced heating power inside the source is ∼13.3 mW. After the HDR source was extended into the PVDF applicator and reached thermal equilibrium, thermal imaging visualized the temperature gradient of 10 K/cm along the PVDF applicator surface, which agreed with FEA modeling. For Ir{sup 192} source activities ranging from 4.20–10.20 Ci, thermal imaging could verify source activity with an accuracy of 6.3% with a dwell time of 10 sec, and an accuracy of 2.5 % with 100 sec. Conclusion: Thermal imaging is a feasible tool to visualize HDR source dwell positions and verify source integrity. Patient safety and treatment quality will be improved by integrating thermal measurements into HDR QA procedures.

60 APPLIED LIFE SCIENCES↗

SU-G-BRB-04: Automated Output Factor Measurements Using Continuous Data Logging for Linac Commissioning

Purpose: Linac commissioning is a time consuming and labor intensive process, the streamline of which is highly desirable. In particular, manual measurement of output factors for a variety of field sizes and energy greatly hinders the commissioning efficiency. In this study, automated measurement of output factors was demonstrated as ‘one-click’ using data logging of an electrometer. Methods: Beams to be measured were created in the recording and verifying (R&V) system and configured for continuous delivery. An electrometer with an automatic data logging feature enabled continuous data collection for all fields without human intervention. The electrometer saved data into a spreadsheet every 0.5 seconds. A Matlab program was developed to analyze the excel data to monitor and check the data quality. Results: For each photon energy, output factors were measured for five configurations, including open field and four wedges. Each configuration includes 72 fields sizes, ranging from 4×4 to 20×30 cm{sup 2}. Using automation, it took 50 minutes to complete the measurement of 72 field sizes, in contrast to 80 minutes when using the manual approach. The automation avoided the necessity of redundant Linac status checks between fields as in the manual approach. In fact, the only limiting factor in such automation is Linac overheating. The data collection beams in the R&V system are reusable, and the simplified process is less error-prone. In addition, our Matlab program extracted the output factors faithfully from data logging, and the discrepancy between the automatic and manual measurement is within ±0.3%. For two separate automated measurements 30 days apart, consistency check shows a discrepancy within ±1% for 6MV photon with a 60 degree wedge. Conclusion: Automated output factor measurements can save time by 40% when compared with conventional manual approach. This work laid ground for further improvement for the automation of Linac commissioning.

43 PARTICLE ACCELERATORS↗

SU-G-BRA-03: PCA Based Imaging Angle Optimization for 2D Cine MRI Based Radiotherapy Guidance

Purpose: To develop an imaging angle optimization methodology for orthogonal 2D cine MRI based radiotherapy guidance using Principal Component Analysis (PCA) of target motion retrieved from 4DCT. Methods: We retrospectively analyzed 4DCT of 6 patients with lung tumor. A radiation oncologist manually contoured the target volume at the maximal inhalation phase of the respiratory cycle. An object constrained deformable image registration (DIR) method has been developed to track the target motion along the respiration at ten phases. The motion of the center of the target mass has been analyzed using the PCA to find out the principal motion components that were uncorrelated with each other. Two orthogonal image planes for cineMRI have been determined using this method to minimize the through plane motion during MRI based radiotherapy guidance. Results: 3D target respiratory motion for all 6 patients has been efficiently retrieved from 4DCT. In this process, the object constrained DIR demonstrated satisfactory accuracy and efficiency to enable the automatic motion tracking for clinical application. The average motion amplitude in the AP, lateral, and longitudinal directions were 3.6mm (min: 1.6mm, max: 5.6mm), 1.7mm (min: 0.6mm, max: 2.7mm), and 5.6mm (min: 1.8mm, max: 16.1mm), respectively. Based on PCA, the optimal orthogonal imaging planes were determined for cineMRI. The average angular difference between the PCA determined imaging planes and the traditional AP and lateral imaging planes were 47 and 31 degrees, respectively. After optimization, the average amplitude of through plane motion reduced from 3.6mm in AP images to 2.5mm (min:1.3mm, max:3.9mm); and from 1.7mm in lateral images to 0.6mm (min: 0.2mm, max:1.5mm), while the principal in plane motion amplitude increased from 5.6mm to 6.5mm (min: 2.8mm, max: 17mm). Conclusion: DIR and PCA can be used to optimize the orthogonal image planes of cineMRI to minimize the through plane motion during radiotherapy guidance.

60 APPLIED LIFE SCIENCES↗

MO-FG-BRA-02: A Feasibility Study of Integrating Breathing Audio Signal with Surface Surrogates for Respiratory Motion Management

Purpose: Tracking the surrogate placed on patient skin surface sometimes leads to problematic signals for certain patients, such as shallow breathers. This in turn impairs the 4D CT image quality and dosimetric accuracy. In this pilot study, we explored the feasibility of monitoring human breathing motion by integrating breathing sound signal with surface surrogates. Methods: The breathing sound signals were acquired though a microphone attached adjacently to volunteer’s nostrils, and breathing curve were analyzed using a low pass filter. Simultaneously, the Real-time Position Management™ (RPM) system from Varian were employed on a volunteer to monitor respiratory motion including both shallow and deep breath modes. The similar experiment was performed by using Calypso system, and three beacons taped on volunteer abdominal region to capture breath motion. The period of each breathing curves were calculated with autocorrelation functions. The coherence and consistency between breathing signals using different acquisition methods were examined. Results: Clear breathing patterns were revealed by the sound signal which was coherent with the signal obtained from both the RPM system and Calypso system. For shallow breathing, the periods of breathing cycle were 3.00±0.19 sec (sound) and 3.00±0.21 sec (RPM); For deep breathing, the periods were 3.49± 0.11 sec (sound) and 3.49±0.12 sec (RPM). Compared with 4.54±0.66 sec period recorded by the calypso system, the sound measured 4.64±0.54 sec. The additional signal from sound could be supplement to the surface monitoring, and provide new parameters to model the hysteresis lung motion. Conclusion: Our preliminary study shows that the breathing sound signal can provide a comparable way as the RPM system to evaluate the respiratory motion. It’s instantaneous and robust characteristics facilitate it possibly to be a either independently or as auxiliary methods to manage respiratory motion in radiotherapy.

60 APPLIED LIFE SCIENCES↗

TU-AB-202-06: Quantitative Evaluation of Deformable Image Registration in MRI-Guided Adaptive Radiation Therapy

Purpose: To assess the performance of the deformable image registration algorithm used for MRI-guided adaptive radiation therapy using image feature analysis. Methods: MR images were collected from five patients treated on the MRIdian (ViewRay, Inc., Oakwood Village, OH), a three head Cobalt-60 therapy machine with an 0.35 T MR system. The images were acquired immediately prior to treatment with a uniform 1.5 mm resolution. Treatment sites were as follows: head/neck, lung, breast, stomach, and bladder. Deformable image registration was performed using the ViewRay software between the first fraction MRI and the final fraction MRI, and the DICE similarity coefficient (DSC) for the skin contours was reported. The SIFT and Harris feature detection and matching algorithms identified point features in each image separately, then found matching features in the other image. The target registration error (TRE) was defined as the vector distance between matched features on the two image sets. Each deformation was evaluated based on comparison of average TRE and DSC. Results: Image feature analysis produced between 2000–9500 points for evaluation on the patient images. The average (± standard deviation) TRE for all patients was 3.3 mm (±3.1 mm), and the passing rate of TRE<3 mm was 60% on the images. The head/neck patient had the best average TRE (1.9 mm±2.3 mm) and the best passing rate (80%). The lung patient had the worst average TRE (4.8 mm±3.3 mm) and the worst passing rate (37.2%). DSC was not significantly correlated with either TRE (p=0.63) or passing rate (p=0.55). Conclusions: Feature matching provides a quantitative assessment of deformable image registration, with a large number of data points for analysis. The TRE of matched features can be used to evaluate the registration of many objects throughout the volume, whereas DSC mainly provides a measure of gross overlap. We have a research agreement with ViewRay Inc.

60 APPLIED LIFE SCIENCES↗

TU-FG-BRB-08: Challenges, Limitations and Future Outlook Towards Clinical Translation of Proton Acoustic Range Verification

Purpose: To report our investigations of proton acoustic imaging, including computer simulations and preliminary experimental studies at clinical facilities. The ultimate achievable accuracy, sensitivity and clinical translation challenges are discussed. Methods: The acoustic pulse due to pressure rise was estimated using finite element model. Since the ionoacoustic pulse is highly dependent on the proton pulse width and energy, multiple pulse widths were studied. Based on the received signal spectrum at piezoelectric ultrasound transducer with consideration of random thermal noise, maximum spatial resolution of the proton-acoustic imaging modality was calculated. The simulation studies defined the design specifications of the system to detect proton acoustic signal from Hitachi and Mevion clinical machines. A 500 KHz hydrophone with 100 dB amplification was set up in a water tank placed in front of the proton nozzle A 40 MHz data acquisition was synchronized by a trigger signal provided by the machine. Results: Given 30–800 mGy dose per pulse at the Bragg peak, the minimum number of protons detectable by the proton acoustic technique was on the order of 10×10^6 per pulse. The broader pulse widths produce signal with lower acoustic frequencies, with 10 µs pulses producing signals with frequency less than 100 kHz. As the proton beam pulse width increases, a higher dose rate is required to measure the acoustic signal. Conclusion: We have established the minimal detection limit for protonacoustic range validation for a variety of pulse parameters. Our study indicated practical proton-acoustic range verification can be feasible with a pulse shorter than 10 µs, 5×10^6 protons/pulse, 50 nA beam current and a highly sensitive ultrasonic transducer. The translational challenges into current clinical machines include proper magnetic shielding of the measurement equipment, providing a clean trigger signal from the proton machine, providing a shorter proton beam pulse and higher dose per pulse.

60 APPLIED LIFE SCIENCES↗

WE-AB-202-04: Statistical Evaluation of Lung Function Using 4DCT Ventilation Imaging: Proton Therapy VS IMRT

Purpose: Variation in function of different lung regions has been ignored so far for conventional lung cancer treatment planning, which may lead to higher risk of radiation induced lung disease. 4DCT based lung ventilation imaging provides a novel yet convenient approach for lung functional imaging as 4DCT is taken as routine for lung cancer treatment. Our work aims to evaluate the impact of accounting for spatial heterogeneity in lung function using 4DCT based lung ventilation imaging for proton and IMRT plans. Methods: Six patients with advanced stage lung cancer of various tumor locations were retrospectively evaluated for the study. Proton and IMRT plans were designed following identical planning objective and constrains for each patient. Ventilation images were calculated from patients’ 4DCT using deformable image registration implemented by Velocity AI software based on Jacobian-metrics. Lung was delineated into two function level regions based on ventilation (low and high functional area). High functional region was defined as lung ventilation greater than 30%. Dose distribution and statistics in different lung function area was calculated for patients. Results: Variation in dosimetric statistics of different function lung region was observed between proton and IMRT plans. In all proton plans, high function lung regions receive lower maximum dose (100.2%–108.9%), compared with IMRT plans (106.4%–119.7%). Interestingly, three out of six proton plans gave higher mean dose by up to 2.2% than IMRT to high function lung region. Lower mean dose (lower by up to 14.1%) and maximum dose (lower by up to 9%) were observed in low function lung for proton plans. Conclusion: A systematic approach was developed to generate function lung ventilation imaging and use it to evaluate plans. This method hold great promise in function analysis of lung during planning. We are currently studying more subjects to evaluate this tool.

60 APPLIED LIFE SCIENCES↗

SU-E-J-22: A Feasibility Study On KV-Based Whole Breast Radiation Patient Setup

Purpose: In room kilovoltage x-ray (kV) imaging provides higher contrast than Megavoltage (MV) imaging with faster acquisition time compared with on-board cone-beam computed tomography (CBCT), thus improving patient setup accuracy and efficiency. In this study we evaluated the clinical feasibility of utilizing kV imaging for whole breast radiation patient setup. Methods: For six breast cancer patients with whole breast treatment plans using two opposed tangential fields, MV-based patient setup was conducted by aligning patient markers with in room lasers and MV portal images. Beam-eye viewed kV images were acquired using Varian OBI system after the set up process. In house software was developed to transfer MLC blocks information overlaying onto kV images to demonstrate the field shape for verification. KV-based patient digital shift was derived by performing rigid registration between kV image and the digitally reconstructed radiography (DRR) to align the bony structure. This digital shift between kV-based and MV-based setup was defined as setup deviation. Results: Six sets of kV images were acquired for breast patients. The mean setup deviation was 2.3mm, 2.2mm and 1.8mm for anterior-posterior, superior-inferior and left-right direction respectively. The average setup deviation magnitude was 4.3±1.7mm for six patients. Patient with large breast had a larger setup deviation (4.4–6.2mm). There was no strong correlation between MV-based shift and setup deviation. Conclusion: A preliminary clinical workflow for kV-based whole breast radiation setup was established and tested. We observed setup deviation of the magnitude below than 5mm. With the benefit of providing higher contrast and MLC block overlaid on the images for treatment field verification, it is feasible to use kV imaging for breast patient setup.

60 APPLIED LIFE SCIENCES↗

SU-E-J-84: Quantitative Dosimetry Assessment of the Impact of Image Artifacts of Metal Implants in Spinal SABR Treatment

Purpose: Metal rods are frequently used to stabilize the spine in patients with metastatic disease. The high Z material causes imaging artifacts in the surrounding tissue in CT scans, which introduces dosimetric uncertainty when inhomogeneity correction is enabled for radiation treatment planning. The purpose of this study is to quantify the dosimetric deviations caused by the imaging artifacts and to evaluate the effectiveness of using Hounsfield units (HU) overwriting to reduce dosimetric uncertainties. Methods: We retrospectively reviewed treatment plans for 4 patients with metal implants who received stereotactic ablative radiation therapy (SABR) for metastatic disease to the spine on Tomotherapy HiArt. For all four patients, the region of imaging artifact surrounding the metal implants was contoured and the pixel HU’s were overwritten to be water equivalent. We then generated adaptive treatment plans for these patients using the MVCT pretreatment set up images and batched beamlets in the original treatment plans. The dosimetry deviation between the adaptive and original plans were compared and quantitatively analyzed. Results: For three out of four patient, the major OAR (spinal cord) dose (0.35cc or 10% according to protocols and fractionation) increased (2.7%, 5.5%, 0%, 3.9%, mean=3.0±2.3%, p=0.04), and the PTV dose (D90 or D95 as per prescription) increased for all four patients ( 2%, 5%, 0.7%, 3.6%, mean=2.8±1.9%, p=0.03) in the adaptive plan with HU overwriting. The average point dose deviation of the Tomotherapy DQA for the same patients was −1.0±1.0%. For plans without HU overwriting, the dose deviation from the treatment plan will increase. Conclusion: The metal implant and the imaging artifacts may cause a significant dosimetric impact on radiation treatment plans for spinal disease. The dose to the PTV and the spinal cord was under-calculated in treatment plans without considering the imaging artifacts. HU overwriting can reduce the dosimetry un-certainty.

60 APPLIED LIFE SCIENCES↗