Computed Tomography Planning Research Articles

Ten-year outcomes of cementless anatomical femoral implants after 3D computed tomography planning. Follow-up note

BackgroundPre-operative 3D planning based on computed tomography (CT) imaging is used to optimise the restitution of normal hip anatomy during primary total hip arthroplasty (THA). Although CT planning has been proven effective and reproducible, its influence on long-term THA outcomes is unknown. In this 10-year follow-up study of patients managed with the same technique of CT-planned primary anatomical THA, the objectives were to assess femoral implant survival, long-term functional outcomes, 10-year outcomes of titanium modular femoral necks, and associations with the dislocation rate.Hypothesis Pre-operative CT planning of primary THA ensures achievement of the NICE criterion of a lower than 5% femoral revision rate within 10 years. Material and methodsThe study included 61 patients (61 hips) managed between 2004 and 2007 by CT-planned primary THA via the posterior approach, with an uncemented anatomical femoral component (SPS®, Symbios); when deemed necessary by the surgeon to restore normal anatomy, a titanium modular femoral neck was used (35/61 patients). After 10 years, 17 patients had died and 3 were lost to follow-up, leaving 41 patients with a mean age of 76 years (range, 60–91 years) for re-evaluation. Clinical outcomes were assessed by determining the Harris Hip Score (HHS) and the Postel-Merle d’Aubigné (PMA) score, which were compared to baseline values. Radiographs were evaluated using the AGORA Roentgenographic Assessment system (ARA). ResultsThe 10-year femoral component survival rate was 96% (95%CI, 88;99%). Revision was required in 4 patients, to treat delayed peri-prosthetic fractures (n=2) or to correct initial cup malposition (n=2). No changes occurred from 2010 to last follow-up in the mean HHS (90 [95%CI, 84;95] and 91 [95%CI, 77;96], respectively) or mean PMA score (16 [95%CI, 14;17] and 15.5 [95%CI, 14;16.5], respectively). The mean ARA score was 5.2 (range, 3–6) at last follow-up. No complications related to the use of modular femoral necks were recorded. Dislocation occurred in 2 patients, but in neither was the pre-operative plan followed during surgery. DiscussionThe SPS® stem produced good 10-year clinical and radiographic outcomes. No patients experienced complications related to use of a titanium modular femoral neck. The restoration of anatomical hip geometry made possible by pre-operative CT planning provided sustained clinical improvements with a low complication rate. Level of evidenceIV, retrospective observational cohort study.

Impact of daily soft-tissue image guidance to prostate on pelvic lymph node (PLN) irradiation for prostate patients receiving SBRT.

PurposeTo determine the impact of using fiducial match for daily image‐guidance on pelvic lymph node (PLN) coverage for prostate cancer patients receiving stereotactic body radiation therapy (SBRT).MethodsThirty patients underwent SBRT treatment to the prostate and PLN from 2014 to 2016. Each patient received either 800cGy × 5 or 500cGy × 5 to the prostate and 500cGy × 5 to the PLN. A 5 mm clinical target volume (CTV)‐to‐planning target volume (PTV) margin around the PLN was used for planning. Two registrations with planning computed tomography (PCT) for each of the daily cone beam CTs (CBCTs) were performed: a rigid registration to fiducials and to the bony anatomy. The average translational difference between fiducial and bony match as well as percentage of fractions with differences > 5mm were calculated. Changes in bladder and rectal volume as well as center‐of‐mass (COM) position from simulation parameters, and their correlation with translational difference were also evaluated. The dosimetric impact of the translational differences was calculated by shifting the plan isocenter.ResultsThe average translational difference between fiducial and bony match was 0.06 ± 0.82, 2.1 ± 4.1, −2.8 ± 4.3, and 5.5 ± 4.2 mm for lateral, vertical, longitudinal, and vector directions. The average change in bladder and rectal volume from simulation was −67.2 ± 163.04 cc (−12 ± 52%) and −1.6 ± 18.75 (−2 ± 30%) cc. The average change in COM of bladder from the simulation position was 0.34 ± 2.49, 4.4 ± 8.1, and −3.9 ± 7.5 mm along the LR, AP, and SI directions. The corresponding COM change for the rectum was 0.17 ± 1.9, 1.34 ± 3.5, and −0.6 ± 5.2 mm.ConclusionsThe 5 mm margin covered ~75% of fractions receiving PLN irradiation with SBRT, daily CBCT and fiducial‐guided setup. The dosimetric impact on PLN coverage was significant in 19% of fractions or 25% of patients. A larger translational shift was due to variation in rectal volume and changes in COM position of the bladder and rectum. A consistent bladder positioning and/or rectum filling compared with presimulation volume were essential for adequate coverage of PLN in a hypofractionated treatment regime.

Evaluation of the Accuracy of Virtual Planning in Orthognathic Surgery: A Morphometric Study.

The purpose of this study was to prospectively test the accuracy of computer-aided orthognathic surgery comparing the virtual surgical planning with the three-dimensional (3D) outcome. Patients that underwent computer-assisted orthognathic surgery were retrospectively evaluated. The postoperative results were compared with the surgical plan, superimposing the postoperative computed tomography (CT) scan onto the virtual plan. Surface-based superimpositioning of the postoperative CT scan onto the 3D preoperative plan was carried out to visualize the discrepancy between preoperative virtual plan and postoperative 3D CT result. A total of 17 consecutive patients that underwent two-jaw computer-assisted orthognathic surgery were enrolled in the study.The average linear differences for selected points were <1 mm in 12 patients out of 17. In 5 patients out of 17, the average differences for selected points were <2 mm. An overall high degree of accuracy between the virtual plan and the postoperative result was found.

Evaluation of variation of interfraction doses to organs at risk during brachytherapy of cervical cancer

BackgroundTwo-dimensional treatment planning using radiographs or simulator films was the standard in planning brachytherapy for patients with cervical cancer. Three-dimensional (3D) treatment planning has improved treatment efficacy. This retrospective study compares conventional and 3D treatment planning of brachytherapy in patients with cervical cancer and interfraction dose variation to bladder and rectum (D2cc). MethodsThe mean doses to bladder and rectum (D2cc) were computed by computed tomography (CT)–based planning during 100 sessions of intracavitary brachytherapy for carcinoma cervix with the same source configuration as generated for conventional planning, and these estimates were compared with the doses at International Commission on Radiation Units and measurements (ICRU) rectal, bladder points and point A. Interfraction variation of doses to bladder and rectum during various sessions was also analysed. ResultThe mean ICRU bladder dose and D2cc of the bladder for all patients was 3.7 Gy and 7.4 Gy, respectively (p < 0.001). The mean ICRU rectal dose from conventional plan was 4.3Gy and with CT planning, 4.45 Gy (p = 0.04). Interfraction dose variations for D2cc of the bladder were min −5.3 Gy and max 4.8 Gy and those of the rectum were min −1.8 Gy and max 1.72Gy. ConclusionDosimetric evaluation of conventional and 3D CT-based treatment planning for the same brachytherapy sessions demonstrated underestimation of ICRU bladder dose points (p < 0.001) and the rectal ICRU point dose and D2cc (p=0.04). The doses to organs at risk did not show a statistically significant variation between the fractions. However, large variation was noted between the interfractional maximum and minimum doses to bladder and rectum.

Evaluation of the Geometric and Dosimetric Accuracy of Synthetic Computed Tomography Images for Magnetic Resonance Imaging-only Stereotactic Radiosurgery.

IntroductionStereotactic radiosurgery (SRS) plans created using synthetic computed tomography (CT) images derived from magnetic resonance imaging (MRI) data may offer the advantage of inhomogeneity correction by convolution algorithms, as is done for CT-based plans. We sought to determine and validate the clinical significance and accuracy of synthetic CT images for inhomogeneity correction in MRI-only stereotactic radiosurgery plans for treatment of brain tumors.MethodsIn this retrospective study, data from two patients with brain metastases and one with meningioma who underwent imaging with multiple modalities and received frameless SRS treatment were analyzed. The SRS plans were generated using a convolution algorithm to account for brain inhomogeneity using CT and synthetic CT images and compared with the original clinical TMR10 plans created using MRI images.ResultsSynthetic CT-derived SRS plans are comparable with CT-based plans using convolution algorithm, and for some targets, based on location, they provided better coverage and a lower maximum dose.ConclusionsThe results suggest similar dose delivery results for CT and synthetic CT-based treatment plans. Synthetic CT plans offered a noticeable improvement in target dose coverage and a more gradual dose fall-off relative to TMR10 MRI-based plans. The major disadvantage is a slightly increased dose (by 0.37%) to nearby healthy tissue (brainstem) for synthetic CT-based plans relative to those created using clinical MRI images, which may be a problem for patients undergoing high-dose treatment.

Feasibility of multi-atlas cardiac segmentation from thoracic planning CT in a probabilistic framework

Toxicity to cardiac and coronary structures is an important late morbidity for patients undergoing left-sided breast radiotherapy. Many current studies have relied on estimates of cardiac doses assuming standardised anatomy, with a calculated increase in relative risk of 7.4% per Gy (mean heart dose). To provide individualised estimates for dose, delineation of various cardiac structures on patient images is required. Automatic multi-atlas based segmentation can provide a consistent, robust solution, however there are challenges to this method. We are aiming to develop and validate a cardiac atlas and segmentation framework, with a focus on the limitations and uncertainties in the process. We present a probabilistic approach to segmentation, which provides a simple method to incorporate inter-observer variation, as well as a useful tool for evaluating the accuracy and sources of error in segmentation.A dataset consisting of 20 planning computed tomography (CT) images of Australian breast cancer patients with delineations of 17 structures (including whole heart, four chambers, coronary arteries and valves) was manually contoured by three independent observers, following a protocol based on a published reference atlas, with verification by a cardiologist. To develop and validate the segmentation framework a leave-one-out cross-validation strategy was implemented. Performance of the automatic segmentations was evaluated relative to inter-observer variability in manually-derived contours; measures of volume and surface accuracy (Dice similarity coefficient (DSC) and mean absolute surface distance (MASD), respectively) were used to compare automatic segmentation to the consensus segmentation from manual contours.For the whole heart, the resulting segmentation achieved a DSC of , with a MASD of mm. Quantitative results, together with the analysis of probabilistic labelling, indicate the feasibility of accurate and consistent segmentation of larger structures, whereas this is not the case for many smaller structures, where a major limitation in segmentation accuracy is the inter-observer variability in manual contouring.

A reliable method of determining glenohumeral offset in anatomic total shoulder arthroplasty

Glenohumeral offset (GHO) may change from the preoperative state after anatomic total shoulder arthroplasty (TSA), and has been identified as a factor that may affect shoulder mechanics, strength, and function. The primary objective was (1) to establish a reliable method of measuring GHO with standardized computed tomography (CT) imaging planes and (2) to determine whether an association exists between GHO and functional outcomes in TSA. Thirty-seven patients underwent TSA for glenohumeral osteoarthritis. Preoperative and postoperative CT scans were reformatted along standardized measurement planes for the glenoid and humerus separately. Inter-rater and intrarater reliability was determined for 3 methods to measure humeral offsetand 2 methods to measure glenoid offset. Univariate regression analysis was used to determine the association between GHO and functional outcomes including the Constant score and strength. Of all methods tested, the highest preoperative and postoperative inter-rater reliability was r = 0.84 and r = 0.8, and r = 0.7 and r = 0.8 for humeral and glenoid offset, respectively. Intrarater reliability was >0.94. There was a mean increase of 4.3 mm (standard deviation, 4.6; range, -10.6 to 10.8) in combined GHO from preoperative to postoperative time points. No associations were observed between change in offset and functional or strength scores. A reliable approach to measure prearthroplasty and postarthroplasty GHO with CT plane standardization has been described. A net increase in GHO was observed after TSA. No associations were found between change in offset after TSA and functional scores or strength up to 2 years postoperatively.

Preoperative 3D CT Planning for Cortical Bone Trajectory Screws: A Retrospective Radiological Cohort Study

Pedicle screw instrumentation is a standard procedure in lumbar spinal fusion. The cortical bone trajectory (CBT) screw is an alternative technique, less invasive but harder to perform. The identification of the entry point and the appropriate direction can be tricky especially to the surgeons just at the beginning. Therefore, the aim of this study is to evaluate the reliability of preoperative computed tomography (CT) planning in the CBT screw placement. We retrospectively evaluated 82 patients who underwent a single-level posterior lumbar arthrodesis with CBT screws. The correct entry point and the best trajectory were identified on a 3-dimensional (3D) multiplanar reconstruction CT view. The concordance between achieved and planned trajectories was assessed on a postoperative CT. A total of 328 screws was positioned. The average entry point distance from the target was 1.1 mm (standard deviation, ±0.15 mm). In 301 screws (91.8%), the obtained trajectory differed no more than 2° from the planned one. No screws misplaced or new neurological deficit was recorded. The CBT screw placement is a great alternative to the pedicle screw. Accurate preoperative 3D planning is useful to predict the entry point and the direction with accuracy similar to navigation systems, avoiding its costs and technical difficulties. The 3D CT planning is helpful in the customization of spine surgery, and the results underline the radiological reliability of this technique.

Fiducial markers in prostate cancer image-guided radiotherapy.

Background: Image-guided radiotherapy (IGRT) is recommended to reduce the risk of geometrical miss when modern radiotherapy technologies with high grades of conformity are used. The purpose of this study was to evaluate the efficacy of fiducial markers (FMs) for electronic portal imaging in prostate cancer radiotherapy in term of evaluating the complications associated with FMs implantation, quantifying inter-fraction prostate motion, and determination of optimal planning target volume (PTV) margins. Methods: In this single institution, prospective, consecutive study, 27 patients underwent implantation of three-gold seed FMs into the prostate gland before prostate radiotherapy. Prior to computed tomography planning, all patients were asked to report any complication associated with FMs implantation that have experienced to date. Daily pre-treatment electronic portal images were captured, and prostate position errors were corrected if they were greater than 2 mm along three translational directions. Optimal PTV expansions were computed using van Herk formula [PTV-margin= 2.5Σ + 0.7σ]. Results: FMs implantation was successful with an acceptable toxicity profile in all patients. Without IGRT, margins of 5.4 mm, 5.8 mm and 5.5 mm, in vertical, longitudinal and lateral directions, respectively, are needed for a 95% confidence level of complete clinical target volume (CTV) coverage in each treatment session. The PTV margins of 3.0 mm, 3.3 mm and 4.0 mm in corresponding directions were calculated when FMs based electronic portal imaging was applied. Conclusion: FMs based electronic portal imaging is an effective tool for prostate cancer IGRT.

Observed high incidence of prostatic calculi with the potential to act as natural fiducials for prostate image guided radiotherapy

This study aims to quantify the incidence and distribution of prostatic calculi in a population of prostate radiotherapy patients and assess their potential role in prostate image guided radiotherapy (IGRT). A retrospective analysis of trans-rectal ultrasound (TRUS), computed tomography (CT) planning and treatment verification cone beam CT (CBCT) scans from radical prostate radiotherapy patients (external beam and brachytherapy) between 2012 and 2014 was undertaken by a single experienced observer. An internationally validated schema from the Prostate Imaging Reporting and Data system (PIRADS) was used to map the location of calculi. The association of calculi with patient and disease characteristics was explored. Data was analysed using SPSS (IBM version 22.0) using descriptive statistical methods and logistic binary regression analysis. 389 scan sets from 254 patients were included in the analysis. The overall incidence of calculi was 85% (n = 218) of which 79% (n = 201) were intra-prostatic calculi. The mean number of intra-prostatic calculi was 2 (range 1-10) and the mean size of calculi was 3.7 mm (range 0.5-15 mm). Calculi were most frequently observed in the posterior of the mid-gland (PI-RADs 3p, 9p) and posterior of the apex (PI-RADs 5p, 11p). 99% (n = 135) of CT planning scans with a corresponding CBCT had calculi in the same PIRADs location and all calculi were visible at the last fraction. There was no statistically significant association of calculi and N stage, M stage or Gleason score. A significant proportion of prostate radiotherapy patients have prostatic calculi detectable on pre radiotherapy imaging. Calculi observed on CT were also detectable on CBCT in 99% of cases and remain visible at the end of treatment. These findings add to the growing evidence base supporting the potential of calculi as an alternative to fiducial markers to aid prostate IGRT.

Dosimetric effects of the kV based image-guided radiation therapy of prone breast external beam radiation: Towards the optimized imaging frequency.

PurposeFor prone breast treatment, daily image‐guided radiation therapy (IGRT) allows couch shifting to correct breast position relative to the treatment field. This work investigates the dosimetric effect of reducing kV imaging frequencies and the feasibility of optimizing the frequency using patient anatomy or their first 3‐day shifts.MethodThirty‐seven prone breast patients who had been treated with skin marker alignment followed by daily kV were retrospectively analyzed. Three IGRT schemes (daily‐kV, weekly‐kV, no‐kV) were simulated, assuming that fractions with kV imaging deliver a dose distribution equivalent to that in computed tomography (CT) planning, whereas other fractions yield a dose distribution as recreated by shifting the CT plan isocenter back to its position before the couch shift was applied. Treatment dose to targets (breast and lumpectomy cavity [LPC]) and organs at risks (OAR)s (heart, ipsilateral lung) in different schemes were calculated. Patient anatomy information on CT plans and first 3‐day couch shift data were analyzed to investigate whether these factors could guide imaging scheme optimization.ResultsWhen kV imaging frequency was reduced, the percentage dose changes (δD) for breast and LPC objectives (average <1%) were smaller than those for heart and lung (average 28%–31% for Dmean). In general, the δD of no‐kV imaging was approximately that of weekly kV imaging × a factor of 1.2–1.4. Although most dose objectives were not affected, the potential higher heart dose may be of concern. No strong correlation was found between δD for different kV frequencies and patient anatomy size/distance or the first 3‐day couch shift data.ConclusionsDespite resulting in lower imaging dose, time, cost, and similar target coverage, a reduction in kV imaging frequency may introduce higher heart complication risk. Daily kVs are needed more in left‐sided breast patients. A less frequent imaging schedule, if considered, cannot be individually optimized using CT anatomic features or early shift data.

Determination of the Effect of Patient-Specific Bladder Filling Protocol for Operated Prostate Cancer Patients on Bladder Volume and Doses Using Cone Beam Computed Tomography Dose Calculations

One of the problems in prostate cancer radiotherapy is the effect of bladder volume changes on irradiated target volume and bladder itself. The aim of this study is to develop a patient-specific bladder filling protocol (PSP) by the help of a bladder scanner (BS) in operated patients and to compare the volumetric and dosimetric parameters with that of the standard filling protocol (SP). For the PSP the patient was asked to empty his bladder and drink 500 mL of water. The changes in bladder volume was measured with BS every 5 minutes. The time to reach a 150-200 cc-volume was recorded. That specific time was used for CT simulation and treatment. For the SP after emptying the bladder and drinking 500 mL of water the patient was asked for urgency. Ten patients in each protocol received 72 Gy postprostatectomy salvage RT with VMAT technique. Catphan phantom was loaded into the Monaco 3.30 planning system after creating a CT-to-ED calibration graph with the aid of cone-beam computed tomography (CBCT) images. CBCT images were taken at five different fractions. OAR,CTV and PTV’s were contoured on CBCT images. Treatment plans done with the planning computed tomography (PCT) images were transferred to CBCTs without any mechanical or dosimetric changes. With the help of CBCT doses average treatment DVHs of each patient were obtained. PCT and BS correlation, changes in bladder volume and doses, standard deviations (SD), and their maximum and minimum values for PCT and CBCT were calculated for both protocols. Pearson correlation coefficient for PCT-BS was 0.74.For PSP average bladder volume and SDs, average bladder doses and SDs for PCT and CBCT were 136.92±25.4 cc and 130.38±25.1 cc, and 159.69±6.5 cGy and 160.79±10.1 cGy respectively. For SP the same values were 205.76±87.5 cc and 170.58±55.2 cc, 151.36±25.1 cGy and 163.54±18.3 cGy respectively. Average absolute difference of bladder volumes between PCT and CBCT were 48.47% for SP and 9.91% for PSP (p=0.007). Average absolute difference of bladder doses between PCT and CBCT for SP and PSP were 12.97% and 4.27% respectively (p=0.130). The dose difference between PCT and CBCT was not statistically significant for PSP (p=0.799). There was a trend for significance for SP (p=0.074). No statistically significant difference was found for bladder volumes between PCT and CBCT in both protocols (p=0.139 and p=0.139). With the PSP optimal bladder filling was obtained and maintained throughout the whole treatment course and it was reproducible in every fraction. Average CBCT doses were close to PCT doses, absolute difference of the bladder volumes and doses between PCT and CBCT, interfractional volume and dose changes were very small compared to SP. The findings of this study were reflected to our clinical practice and the determination of the changes in bladder volume and the time for obtaining optimal bladder volume by BS before and throughout the treatment in prostate cancer patients has become our routine clinical practice.

A Phase II Study of Two-Step Intensity Modulated Radiation Therapy (IMRT) with Chemotherapy for Loco-Regionally Advanced Nasopharyngeal Cancer (NPC) (JCOG1015)

A phase II study of adaptive two-step IMRT with chemotherapy (Cx) for NPC (JCOG1015, UMIN-CTR: UMIN000005448) was conducted to evaluate the efficacy and safety. Acute toxicities and treatment compliance had been reported, and this is the final report on efficacy. Patients (pts) aged 20-75 years with stages II-IVB NPC with performance status of 0-1 were enrolled. As adaptive two-step IMRT, computed tomography planning was done twice before the start of IMRT for initial plan of 46 Gy/23 fractions (fr) and at the third or fourth week of treatment for boost plan of 24 Gy/12 fr with a total radiation dose of 70 Gy/ 35 fr. Individual case review was performed. Cx (cisplatin 80 mg/ m2 /3-weeks × 3 cycles) was given concurrently with IMRT, followed by adjuvant Cx (cisplatin 70 mg/ m2 with 5-FU 700 mg/m2 for 5 days/4-weeks × 3 cycles). The primary endpoint was 3-year overall survival (OS). The sample size was set as 75 pts by precision-base so that the 95% confidence interval (CI) for the estimated 3-year OS would be ±10% around the expected value of 74% based on the predicted proportion of stage (UICC 7th) II:III:IV = 20%:40%:40%. If upper limit of the 95% CI of estimated 3-year OS based on the actual proportion of stage was not lower than the expected value of target population by the conventional treatment, we conclude that adaptive two-step IMRT method was to be considered as a standard treatment. As a secondary endpoint, xerostomia score was evaluated. The expected incidences of Grade 2 or more xerostomia at 1, 2, and 3 years after IMRT were 30%, 25%, and 20%. From April 2011 to October 2014, 75 pts were enrolled from 12 institutions. The pts consisted of 60 men and 15 women with the median age of 55 years (range, 28-75). At least 2 cycles of concurrent Cx with a total dose of 70 Gy was completed in 74 pts (99%). However, the full protocol treatment including 3 cycles of adjuvant Cx was completed for 39 pts (52%). The 3-year OS for the 75 pts was 88% [95% CI; 78%-94%]. The upper and lower limit of the 95% CI of the 3-year OS (94%) was higher than the target population’s expected 3-year OS of 75% adjusted by the stage with actual proportion of stage II:III:IV = 21%:44%:35%; thus the primary endpoint was met. The 3-year progression-free survival (PFS) and loco-regional PFS were 71% [59%-80%] and 77% [66%-85%]. Loco-regional recurrences were noted in 13 pts (17%). In terms of late toxicities, 15 pts (20%) showed grade 3 toxicities including hearing impaired, aspiration, and dysphagia. No grade 4-5 toxicities were noted. Grade 2 or more xerostomia was noted in 26% (18/70), 12% (8/69), and 9% (6/65) at 1, 2, and 3 years after start of IMRT. Adaptive two-step IMRT for NPC showed an excellent 3-year OS with acceptable toxicities, and the incidence of xerostomia was satisfactorily low. This method would be a new standard treatment for locally advanced NPC.

Cone beam CT for QA of synthetic CT in MRI only for prostate patients.

PurposeMagnetic resonance imaging (MRI)‐only radiotherapy is performed without computed tomography (CT). A synthetic CT (sCT) is used for treatment planning. The aim of this study was to develop a clinically feasible quality assurance (QA) procedure for sCT using the kV‐cone beam CT (CBCT), in an MRI‐only workflow for prostate cancer patients.Material and methodThree criteria were addressed; stability in Hounsfield Units (HUs), deviations in HUs between the CT and CBCT, and validation of the QA procedure. For the two first criteria, weekly phantom measurements were performed. For the third criteria, sCT, CT, and CBCT for ten patients were used. Treatment plans were created based on the sCT (MriPlannerTM). CT and CBCT images were registered to the sCT. The treatment plan was copied to the CT and CBCT and recalculated. Dose–volume histogram (DVH) metrics were used to evaluate dosimetric differences between the sCT plan and the recalculated CT and CBCT plans. HU distributions in sCT, CT, and CBCT were compared. Well‐defined errors were introduced in the sCT for one patient to evaluate efficacy of the QA procedure.ResultsThe kV‐CBCT system was stable in HU over time (standard deviation <40 HU). Variation in HUs between CT and CBCT was <60 HU. The differences between sCT–CT and sCT–CBCT dose distributions were below or equal to 1.0%. The highest mean dose difference for the CT and CBCT dose distribution was 0.6%. No statistically significant difference was found between total mean dose deviations from recalculated CT and CBCT plans, except for femoral head. Comparing HU distributions, the CBCT appeared to be similar to the CT. All introduced errors were identified by the proposed QA procedure, except all tissue compartments assigned as water.ConclusionThe results in this study shows that CBCT can be used as a clinically feasible QA procedure for MRI‐only radiotherapy of prostate cancer patients.

Prediction of Response to Stereotactic Radiosurgery for Brain Metastases Using Convolutional Neural Networks.

It is unclear whether radiomic phenotypes of brain metastases (BM) are related to radiation therapy prognosis. This study assessed whether a convolutional neural network (CNN)-based radiomics model which learned computer tomography (CT) image features with minimal preprocessing, could predict early response of BM to radiosurgery. Tumor images of 110 BM post stereotactic-radiosurgery (SRS) (within 3 months) were assessed (Response Evaluation Criteria in Solid Tumor, version 1.1) as responders (complete or partial response) or non-responders (stable or progressive disease). Datasets were axial planning CT images containing the tumor center, and the tumor response. Datasets were randomly assigned to training, validation, or evaluation groups repeatedly, to create 50 dataset combinations that were classified into five groups of 10 different dataset combinations with the same evaluation datasets. The CNN learned using training-group images and labels. Validation datasets were used to choose the model that best classified evaluation images as responders or non-responders. Of 110 tumors, 57 were classified as responders, and 53 as non-responders. The area under the receiver operating characteristic curve (AUC) of each CNN model for 50 dataset combinations ranged from 0.602 [95% confidence interval (CI)=36.5-83.9%] to 0.826 [95% CI, 64.3-100%]. The AUC of ensemble models, which averaged prediction results of 10 individual models within the same group, ranged from 0.761 (95% CI=55.2-97.1%) to 0.856 (95% CI=68.2-100%). A CNN-based ensemble radiomics model accurately predicted SRS responses of unlearned BM images. Thus, CNN models are able to predict SRS prognoses from small datasets.

CT imaging during treatment improves radiomic models for patients with locally advanced head and neck cancer

Background and purposeThe development of radiomic risk models to predict clinical outcome is usually based on pre-treatment imaging, such as computed tomography (CT) scans used for radiation treatment planning. Imaging data acquired during the course of treatment may improve their prognostic performance. We compared the performance of radiomic risk models based on the pre-treatment CT and CT scans acquired in the second week of therapy. Material and methodsTreatment planning and second week CT scans of 78 head and neck squamous cell carcinoma patients treated with primary radiochemotherapy were collected. 1538 image features were extracted from each image. Prognostic models for loco-regional tumour control (LRC) and overall survival (OS) were built using 6 feature selection methods and 6 machine learning algorithms. Prognostic performance was assessed using the concordance index (C-Index). Furthermore, patients were stratified into risk groups and differences in LRC and OS were evaluated by log-rank tests. ResultsThe performance of radiomic risk model in predicting LRC was improved using the second week CT scans (C-Index: 0.79), in comparison to the pre-treatment CT scans (C-Index: 0.65). This was confirmed by Kaplan–Meier analyses, in which risk stratification based on the second week CT could be improved for LRC (p = 0.002) compared to pre-treatment CT (p = 0.063). ConclusionIncorporation of imaging during treatment may be a promising way to improve radiomic risk models for clinical treatment adaption, i.e., to select patients that may benefit from dose modification.

P198] Cone beam CT as a clinical QA tool for prostate MRI-only workflows

Purpose Magnetic resonance imaging (MRI) only radiotherapy has in recent years been introduced as a clinical workflow for prostate cancer patients. MRI only radiotherapy is performed without computed tomography (CT) and instead, a synthetic CT (sCT) is used for treatment planning. With the introduction of new procedures for treatment planning, development of quality assurance (QA) tools must be considered. Today, no such tool has yet been presented. The aim of this study was to develop a clinically feasible QA procedure for sCT using the kV-cone beam CT (CBCT), in an MRI-only workflow for prostate cancer patients. Methods Three criteria were addressed; 1. Stability in Hounsfield Units (HUs), 2. Deviations in HUs between the CT and CBCT and 3. Validation of the QA procedure. For the two first criteria, phantom measurements were performed. For the third criteria, sCT, CT and CBCT for ten patients were used. Treatment plans were created based on the sCT (MriPlannerTM). CT and CBCT images were registered to the sCT. The treatment plan was recalculated on the CT and CBCT. Dose-volume histogram (DVH) metrics were used to evaluate dosimetric differences between the sCT plan and the recalculated CT and CBCT plans. Well defined errors were artificially introduced in the sCT for one patient to evaluate efficacy of the QA procedure. Results The kV-CBCT system was stable in HU over time (SD Conclusions The results in this study indicates that CBCT can be used as a clinically feasible QA tool for MRI-only radiotherapy of prostate cancer patients. The proposed method could easily be integrated in any modern image guided radiotherapy workflow without extensive work.

P276] Impact on proton range estimates using a novel mr-based artifact reduction algorithm

Purpose Metal implants lead to streak and cupping artifacts in computed tomography (CT) images, causing erroneous CT number estimates for radiation therapy (RT) planning. We recently introduced kerMAR, a novel Magnetic Resonance (MR)-based CT metal artifact reduction algorithm that combines kernel regression on uncorrupted CT value/MR patch pairs with a forward model of the CT metal artifacts. Here, we compare the impact of kerMAR and the clinical oMAR algorithm (Philips) on the proton range in calculated spot scanning (SS) dose plans for proton RT. Methods We acquired T1w MR and CT image sets of a veal shank phantom with and without metal markers inserted, and applied kerMAR and oMAR to the former. We then created SS plans in Eclipse v. 13.7 (Varian) delivering 150 MeV protons in a 0.55 cm single spot that passed 1) near a metal marker at shallow depth and 2) near two different markers located at shallow (near bone) and deep depths, respectively. Similar images were acquired for 4 head-and-neck (HN) RT patients with dental implants, and plans using the same phantom beam, angled through the artifact corrupted oral cavity, were devised, leading to a configuration most resembling phantom beam 2. We finally recalculated the dose plans for the same beamline settings and monitor units on all CT sets, and compared the proton ranges defined as the depth with 80% of the maximum dose. Results For the phantom plans, the following range deviations relative to the metal-free reference plan (uncorrected CT, oMAR, kerMAR) were observed: Plan 1: (0.6, 0.6, 0.6)%; plan 2: (−4, −1.8, −2.1)%. The patient plans showed the following deviations from the uncorrected CT plan: kerMAR: (10, 8, −7, 11)%; oMAR: (3, 0.1, −10, 0.2)%. Conclusions Proton range calculations on a phantom showed similar improvements compared to the uncorrupted reference with kerMAR and oMAR. Similar calculations on HN patients, however, displayed systematic proton range increases when applying kerMAR as compared to oMAR.

Automatic quantification of calcifications in the coronary arteries and thoracic aorta on radiotherapy planning CT scans of Western and Asian breast cancer patients

PurposeThis study automatically quantified calcifications in coronary arteries (CAC) and thoracic aorta (TAC) on breast planning computed tomography (CT) scans and assessed its reproducibility compared to manual scoring. Material and MethodsDutch (n = 1199) and Singaporean (n = 1090) breast cancer patients with radiotherapy planning CT scan were included. CAC and TAC were automatically scored using deep learning algorithm. CVD risk categories were based on Agatson CAC: 0, 1–10, 11–100, 101–400 and >400. Reliability between automatic and manual scoring was assessed in 120 randomly selected CT scans from each population, with linearly weighted kappa for CAC categories and intraclass correlation coefficient for TAC. ResultsMedian age was higher in Dutch patients than Singaporean patients: 57 versus 52 years. CAC and TAC increased with age and were more present in Dutch patients than Singaporean patients: 24.2% versus 17.3% and 73.0% versus 62.2%, respectively. Reliability of CAC categories and TAC was excellent in the Netherlands (0.85 (95% confidence interval (CI) = 0.77–0.93) and 0.98 (95% CI = 0.96–0.98) respectively) and Singapore (0.90 (95% CI = 0.84–0.96) and 0.99 (95% CI = 0.98–0.99) respectively). ConclusionsCAC and TAC prevalence was considerable and increased with age. Deep learning software is a reliable method to automatically measure CAC and TAC on radiotherapy breast CT scans.

Dosimetric Evaluation of Incorporating Patient Geometric Variations Into Adaptive Plan Optimization Through Probabilistic Treatment Planning in Head and Neck Cancers

Four-dimensional (4D) adaptive radiation therapy (ART) treatment planning is an alternative to the conventional margin-based treatment planning approach. In 4D ART, interfraction patient geometric variations, gathered from computed tomography (CT) or cone beam CT (CBCT) images acquired during the patient treatment course, are directly incorporated into the adaptive plan optimization using a probabilistic treatment planning method. The goal of the present planning study was to evaluate the dosimetric differences between 4D ART and conventional margin-based adaptive planning strategies for head and neck cancers. In addition, we examined whether the dose differences achieved with 4D ART would translate into clinically relevant toxicity reductions using the existing normal tissue complication probability (NTCP) models. For 18 head and neck cancer patients, the treatment plans were retrospectively generated for 4 different treatment strategies, including a solely image guided radiation therapy (IGRT) strategy (IGRT-only), 2 conventional adaptive treatment planning strategies using 3- and 0-mm planning target volume (PTV) margins, and the 4D ART strategy. In the IGRT-only strategy, a conventional 3-mm PTV margin treatment plan was applied for the entire treatment course. In the 2 conventional adaptive strategies, 2 new treatment plans were generated during the treatment course using diagnostic planning CT scans acquired after the 10th and 22nd fractions. The 4D ART followed the same adaptive schedule, except that the 4D adaptive plan was generated using 5 CBCT images acquired during the 5 most recent treatment fractions. For each strategy, the actual delivered dose for the entire treatment course was constructed by calculating the daily doses on 35 CBCT scans, deforming back to the pretreatment planning CT scan, and accumulating over all 35 fractions. The target coverage was evaluated using the percentage of target volume receiving ≥100% of the prescription dose (V100%) and the minimum dose to 99% of the target volume (D99). It was considered adequate if the V100% was ≥95% and the dose deficit in D99 was ≤2Gy (with respect to the prescription dose). For each strategy, the dose received by the organs at risk (OARs) was also evaluated, and the corresponding NTCP values were subsequently calculated using 3 NTCP models. Adequate target coverage was achieved for the primary clinical target volume (CTV1) and elective nodal CTV (CTV2) with a 3-mm PTV margin, regardless of adaptation. The 3-mm ART plan reduced the OAR mean dose by 1 to 2Gy compared with the IGRT-only plan. The 0-mm ART plan further reduced the OAR dose by another 2 to 3Gy at the expense of target coverage: 3 and 1 patient had V100% <95%, and 6 and 5 patients had a >2 Gy dose deficit in D99 for the CTV1 and CTV2, respectively. Use of 4D ART improved target coverage and attained OAR sparing similar to that with 0-mm ART. The number of patients with V100% <95% and >2 Gy D99 deficit decreased to 0 and 0 for CTV1 and 0 and 2 for CTV2, respectively. The NTCP calculations suggested that 4D ART could benefit a substantial portion of patients compared with IGRT-only because 17 and 12 patients had ≥5% and ≥10% NTCP reductions for parotid toxicity and 18 and 3 patients had ≥5% and ≥10% NTCP reductions for swallowing toxicity, respectively. Compared with margin-based adaptive planning strategies, 4D ART provides a better balance between target coverage and OAR sparing. NTCP estimation predicted for theoretical clinical benefits that warrant further clinical validation.