Board-certified Radiation Oncologists Research Articles

Evaluating Virtual Contrast-Enhanced Magnetic Resonance Imaging in Nasopharyngeal Carcinoma Radiation Therapy: A Retrospective Analysis for Primary Gross Tumor Delineation

PurposeTo investigate the potential of virtual contrast-enhanced MRI (VCE-MRI) for gross-tumor-volume (GTV) delineation of nasopharyngeal carcinoma (NPC) using multi-institutional data. Methods and MaterialsThis study retrospectively retrieved T1-weighted (T1w), T2-weighted (T2w) MRI, gadolinium-based contrast-enhanced MRI (CE-MRI) and planning CT of 348 biopsy-proven NPC patients from three oncology centers. A multimodality-guided synergistic neural network (MMgSN-Net) was trained using 288 patients to leverage complementary features in T1w and T2w MRI for VCE-MRI synthesis, which was independently evaluated using 60 patients. Three board-certified radiation oncologists and two medical physicists participated in clinical evaluations in three aspects: image quality assessment of the synthetic VCE-MRI, VCE-MRI in assisting target volume delineation, and effectiveness of VCE-MRI-based contours in treatment planning. The image quality assessment includes distinguishability between VCE-MRI and CE-MRI, clarity of tumor-to-normal tissue interface and veracity of contrast enhancement in tumor invasion risk areas. Primary tumor delineation and treatment planning were manually performed by radiation oncologists and medical physicists, respectively. ResultsThe mean accuracy to distinguish VCE-MRI from CE-MRI was 31.67%; no significant difference was observed in the clarity of tumor-to-normal tissue interface between VCE-MRI and CE-MRI; for the veracity of contrast enhancement in tumor invasion risk areas, an accuracy of 85.8% was obtained. The image quality assessment results suggest that the image quality of VCE-MRI is highly similar to real CE-MRI. The mean dosimetric difference of planning target volumes were less than 1Gy. ConclusionsThe VCE-MRI is highly promising to replace the use of gadolinium-based CE-MRI in tumor delineation of NPC patients.

Testing and Validation of a Custom Retrained Large Language Model for the Supportive Care of HN Patients with External Knowledge Base.

This study aimed to develop a retrained large language model (LLM) tailored to the needs of HN cancer patients treated with radiotherapy, with emphasis on symptom management and survivorship care. A comprehensive external database was curated for training ChatGPT-4, integrating expert-identified consensus guidelines on supportive care for HN patients and correspondences from physicians and nurses within our institution's electronic medical records for 90 HN patients. The performance of our model was evaluated using 20 patient post-treatment inquiries that were then assessed by three Board certified radiation oncologists (RadOncs). The rating of the model was assessed on a scale of 1 (strongly disagree) to 5 (strongly agree) based on accuracy, clarity of response, completeness s, and relevance. The average scores for the 20 tested questions were 4.25 for accuracy, 4.35 for clarity, 4.22 for completeness, and 4.32 for relevance, on a 5-point scale. Overall, 91.67% (220 out of 240) of assessments received scores of 3 or higher, and 83.33% (200 out of 240) received scores of 4 or higher. The custom-trained model demonstrates high accuracy in providing support to HN patients offering evidence-based information and guidance on their symptom management and survivorship care.

Knowledge and competences in hematological malignancies amongst radiation oncology residents in Germany-results from anational survey.

Radiation oncology is apivotal modality in the treatment of hematologic malignancies. To enable state-of-the-art patient care, structured education during residency is essential. However, given the lack of detailed data, the scope of educational opportunities available to trainees remains elusive. This prompted our group to perform anational survey amongst radiation oncology residents in Germany assessing the status quo of competences in the treatment of lymphoma and leukemia patients. Furthermore, areas of potential improvement were identified to further the goal of competence-based education for residents. Asurvey-based analysis was conducted to assess the knowledge and competence of radiation oncology residents in Germany regarding hematological malignancies. Adecisive questionnaire covering demographics, self-assessment of competences, and areas for improvement was developed in adaption of asurvey by the Association of Residents in Radiation Oncology and distributed amongst 1439 members of the German Society of Radiation Oncology. Responses were collected anonymously via an online survey tool and analyzed using descriptive statistics and chi-square tests. Atotal of 59complete and 22partial responses were collected, yielding a5.6% response rate. Participants' competence varied, with notable experience gaps in pediatric cases, proton therapy, and large-field techniques like total-skin irradiation or pediatric total body irradiation. While participants felt confident in treatment planning and patient counseling, they showed deficiencies in the definition of the planning target volume for modern involved site radiotherapy. Resources for education included national and international guidelines, scientific reviews, and textbooks. Board-certified radiation oncologists and physicians from specialized lymphoma centers demonstrated higher overall competence levels. This survey highlights the diversity of resident education regarding hematological malignancies in German radiation oncology programs. Knowledge gaps exist in key areas, including pediatric cases and specialized techniques. Competence-based education, interactive teaching formats, and rotations to specialized centers are potential strategies to address these gaps. The study contributes to the understanding of the federal educational landscape, underscoring the need for standardized and comprehensive training to ensure optimal patient care in hematological malignancies within the context of radiation oncology. Further research and collaborations are warranted to enhance training and expertise in this critical domain.

AAPM WGVRTO report 390: A survey of veterinary radiation oncology equipment and infrastructure in 2022.

Since 2010, there has been little published data on the state of equipment and infrastructure in veterinary radiation oncology clinical practice. These data are important not only to identify the status and use of technology within the veterinary radiation oncology community but also to help identify the extent of medical physics support. The purpose of our study is to report findings from a survey of veterinary radiation oncologists in the USA, Canada, and select centers outside of North America in 2022. A 40-question survey covering topics such as type of radiotherapy equipment, techniques offered, treatment planning systems and dose calculation algorithms, special techniques, board-certified radiation oncologists and residents, and extent of medical physics support was distributed through an online survey tool. Results from 40 veterinary radiation oncology institutions, with equipment explicitly used for veterinary care, suggest that the current state of practice is not dissimilar to what currently exists in human radiation oncology facilities; techniques and technologies commonly employed include flattening filter-free mode megavoltage beams, volumetric arc therapy, daily cone-beam computed tomography, image-guided radiation therapy, and sophisticated dose calculation algorithms. These findings suggest the need for modern radiation oncology acceptance testing, commissioning, and quality assurance programs within the veterinary community. The increase in veterinary radiation oncology residency positions and increasing sophistication of equipment suggests that increased levels of standardized medical physics support would benefit the veterinary radiation oncology community.

Evaluation of a deep image-to-image network (DI2IN) auto-segmentation algorithm across a network of cancer centers.

Due to manual OAR contouring challenges, various automatic contouring solutions have been introduced. Historically, common clinical auto-segmentation algorithms used were atlas-based, which required maintaining a library of self-made contours. Searching the collection was computationally intensive and could take several minutes to complete. Deep learning approaches have shown significant benefits compared to atlas-based methods in improving segmentation accuracy and efficiency in auto-segmentation algorithms. This work represents the first multi-institutional study to describe and evaluate an AI algorithm for the auto-segmentation of organs at risk (OARs) based on a deep image-to-image network (DI2IN). The AI-Rad Companion Organs RT (AIRC) algorithm (Siemens Healthineers, Erlangen, Germany) uses a two-step approach for segmentation. In the first step, the target organ region in the optimal input image is extracted using a trained deep reinforcement learning network (DRL), which is then used as input to create the contours in the second step based on DI2IN. The study was initially designed as a prospective single-center evaluation. The automated contours generated by AIRC were evaluated by three experienced board-certified radiation oncologists using a four-point scale where 4 is clinically usable and 1 requires re-contouring. After seeing favorable results in a single-center pilot study, we decided to expand the study to six additional institutions, encompassing eight additional evaluators for a total of 11 physician evaluators across seven institutions. One hundred and fifty-six patients and 1366 contours were prospectively evaluated. The five most commonly contoured organs were the lung (136 contours, average rating = 4.0), spinal cord (106 contours, average rating = 3.1), eye globe (80 contours, average rating = 3.9), lens (77 contours, average rating = 3.9), and optic nerve (75 contours, average rating = 4.0). The average rating per evaluator per contour was 3.6. On average, 124 contours were evaluated by each evaluator. 65% of the contours were rated as 4, and 31% were rated as 3. Only 4% of contours were rated as 1 or 2. Thirty-three organs were evaluated in the study, with 19 structures having a 3.5 or above average rating (ribs, abdominopelvic cavity, skeleton, larynx, lung, aorta, brachial plexus, lens, eye globe, glottis, heart, parotid glands, bladder, kidneys, supraglottic larynx, submandibular glands, esophagus, optic nerve, oral cavity) and the remaining organs having a rating of 3.0 or greater (female breast, proximal femur, seminal vesicles, rectum, sternum, brainstem, prostate, brain, lips, mandible, liver, optic chiasm, spinal cord, spleen). No organ had an average rating below 3. AIRC performed well with greater than 95% of contours accepted by treating physicians with no or minor edits. It supported a fully automated workflow with the potential for time savings and increased standardization with the use of AI-powered algorithms for high-quality OAR contouring.

VOARiability: Interobserver and intermodality variability analysis in OAR contouring from head and neck CT and MR images.

Accurate and consistent contouring of organs-at-risk (OARs) from medical images is a key step of radiotherapy (RT) cancer treatment planning. Most contouring approaches rely on computed tomography (CT) images, but the integration of complementary magnetic resonance (MR) modality is highly recommended, especially from the perspective of OAR contouring, synthetic CT and MR image generation for MR-only RT, and MR-guided RT. Although MR has been recognized as valuable for contouring OARs in the head and neck (HaN) region, the accuracy and consistency of the resulting contours have not been yet objectively evaluated. To analyze the interobserver and intermodality variability in contouring OARs in the HaN region, performed by observers with different level of experience from CT and MR images of the samepatients. In the final cohort of 27 CT and MR images of the same patients, contours of up to 31 OARs were obtained by a radiation oncology resident (junior observer, JO) and a board-certified radiation oncologist (senior observer, SO). The resulting contours were then evaluated in terms of interobserver variability, characterized as the agreement among different observers (JO and SO) when contouring OARs in a selected modality (CT or MR), and intermodality variability, characterized as the agreement among different modalities (CT and MR) when OARs were contoured by a selected observer (JO or SO), both by the Dice coefficient (DC) and 95-percentile Hausdorff distance (HD ). The mean (±standard deviation) interobserver variability was 69.0 ± 20.2% and 5.1 ± 4.1mm, while the mean intermodality variability was 61.6 ± 19.0% and 6.1 ± 4.3mm in terms of DC and HD , respectively, across all OARs. Statistically significant differences were only found for specific OARs. The performed MR to CT image registration resulted in a mean target registration error of 1.7 ± 0.5mm, which was considered as valid for the analysis of intermodalityvariability. The contouring variability was, in general, similar for both image modalities, and experience did not considerably affect the contouring performance. However, the results indicate that an OAR is difficult to contour regardless of whether it is contoured in the CT or MR image, and that observer experience may be an important factor for OARs that are deemed difficult to contour. Several of the differences in the resulting variability can be also attributed to adherence to guidelines, especially for OARs with poor visibility or without distinctive boundaries in either CT or MR images. Although considerable contouring differences were observed for specific OARs, it can be concluded that almost all OARs can be contoured with a similar degree of variability in either the CT or MR modality, which works in favor of MR images from the perspective of MR-only and MR-guidedRT.

Treatment Volume Determination in Small Cell Lung Cancer Patients

Objective: In this study, we evaluated treatment volume determination with PET-CT fusion for small cell lung cancer. Materials and methods: Primary goal of this study has been to evaluate treatment volume determination for fused PET-CT in patients with small cell lung cancer. We have carried out a comparative analysis of treatment volume determination by CT simulation images only or by integration of PET. While we primarily focused on evaluation of incorporated multimodality imaging for treatment volume determination, we also assessed critical organ contouring along with interobserver and intraobserver variations. Ground truth target volume has been utilized for comparative analysis, and it was determined by board certified radiation oncologists after detailed evaluation of all imaging and relevant data with thorough colleague peer review and consensus.

Pelvic Nodal Auto-Segmentation Using a Deep Image to Image Network (DI2IN) Auto-Segmentation Algorithm: Comparing Male vs. Female Pelvis

Deep Learning approaches have shown significant benefits compared to atlas-based methods in improving segmentation accuracy and efficiency in auto-segmentation algorithms. The AI-Rad Companion Organs RT pelvic nodal auto-segmentation feature was trained and developed for use in the male pelvis. There is no real-world data on its usability in male pelvis and whether it can be used for female pelvic nodal anatomy. This work represents the first multi-institutional study to describe and evaluate an AI algorithm for auto-segmentation of the pelvic nodal region in female patients based on a deep image-to-image network (DI2IN). The AIRC algorithm uses a two-step approach for segmentation. In the first step, the target organ region in the optimal input image is extracted using a trained Deep Reinforcement Learning network (DRL), which is then used as input to create the contours in the second step based on DI2IN. We retrospectively evaluated AIRC pelvic nodal auto-segmentation in both male and female patients treated at our network of institutions. The automated pelvic nodal contours generated by AIRC were evaluated by one board-certified radiation oncologist, specializing in prostate and gynecologic malignancies. A 4-point scale was used, where 4 is clinically usable and 1 requires re-contouring. Pelvic nodal regions included the right and left side of the common iliac, external iliac, internal iliac, obturator and midline presacral nodes. A chi-squared test was then used to compare the scores of male and female pelvic nodal cases. Fifty-two female and 51 male patients were included in the study, representing a total of 468 and 447 pelvic nodal regions, respectively. 96% (450 pelvic nodal contours) and 99% (443 pelvic nodal contours) required no or minor edits for female and male patients, respectively (p = 0.004). The right internal iliac was the only nodal group with a statistically significant difference between female (92% requiring no or minor edits) and male (100% requiring no or minimal edits) patients, p = 0.04. The percentage of patients requiring no, or minor edits was 87% (45 patients) and 92% (47 patients) for female and male patients, respectively (p = 0.36). AIRC pelvic nodal auto-segmentation performed very well in both male and female pelvic nodal regions, with the male pelvic nodal regions performing better especially in the right internal iliac nodal group. It is usable in female pelvic nodal regions, with 96% of contours requiring no or minor edits. As auto-segmentation becomes more widespread, it may be important to have equal representation from all genders in training and validation of auto-segmentation algorithms.

Patterns and Longitudinal Changes in The Practice of Breast Cancer Radiotherapy in Korea: Korean Radiation Oncology Group 22-01

To analyze contemporary practice patterns in breast cancer radiotherapy (RT) and to assess longitudinal changes therein over 5 years in Korea. A nationwide survey was conducted among board-certified radiation oncologists in Korea by the Division for Breast Cancer of the Korean Radiation Oncology Group in March 2022. The survey consisted of 44 questions related to six domains: hypofractionated (HypoFx) whole breast RT, accelerated partial breast RT (APBI), regional nodal irradiation (RNI), RT for ductal carcinoma in situ (DCIS), postmastectomy RT (PMRT), and tumor bed boost. In total, 70 radiation oncologists from 61 of 101 (60%) institutions participated in the survey. HypoFx RT was used by 62 respondents (89%), which has significantly increased from 36% in 2017. The HypoFx RT was commonly administered at 40-42.5 Gy in 15-16 fractions. APBI was used by 12 respondents (17%), which has increased from 5% in 2017. The use of RNI did not change significantly: ≥ pN2 (6%), ≥ pN1 (33%), and ≥ pN1 with pathological risk factors (61%). However, the indications for use of internal mammary lymph node (IMN) irradiation have expanded. In particular, the rates of routine treatment of IMN (11% from 6% in 2017) and treatment in cases of ≥ pN2 (27% from 14% in 2017) have doubled; however, the rate of treatment for only IMN involvement, identified on imaging, has decreased to 31% from 47% in 2017. With regard to DCIS, the use of hypoFx RT increased to 75% from 25%, and the rate of omission on of RT after breast-conserving surgery decreased to 38% from 48% in 2017. The use of hypoFx RT for PMRT also increased to 36% from 8% in 2017. The adoption of HypoFx RT after breast-conserving surgery in invasive breast cancer and DCIS has increased significantly, whereas that for PMRT has increased moderately, compared to 2017. Further studies are required to determine the optimal use of RNI.

Prevalence and Risk Factors for Retropharyngeal and Retro-Styloid Lymph Node Metastasis in Hypopharyngeal Carcinoma

We evaluated the prevalence and identified the risk factors for retropharyngeal and retro-styloid lymph node metastasis (LNM) in patients with hypopharyngeal carcinoma (HPC). This was achieved using a combination of MR and FDG-PET/CT images. Two board-certified radiation oncologists, who experienced 16 and 18 years in the diagnosis and treatment of head and neck cancers retrospectively reviewed pretreatment FDG-PET/CT images and contrast-enhanced thin-slice CT and MR images of 155 patients with HPC who underwent radiotherapy. Observers assessed these images without prior knowledge of the patient's clinical information and disagreements were resolved by consensus. The radiological diagnostic criteria for lateral retropharyngeal LNM (RPLNM) and retro-styloid LNM (RSLNM) were a short-axis diameter of ≥5 mm and/or necrosis and/or abnormal FDG uptake. Any visible medial RPLN was defined as LNM. RPLNM was confirmed in 20 (13%) patients. Of these, 10 (6%) had ipsilateral RPLNM, 5 (3%) had contralateral RPLNM, and 5 (3%) had bilateral RPLNM. No patients were diagnosed with medial RPLNM. Fisher's exact test revealed tumor site (p = 0.006), T category (p = 0.049), and cervical LNM (p < 0.001) to be significantly associated with RPLNM. Logistic regression analysis revealed posterior wall (PW) tumors (odds ratio [OR]: 4.128, 95% confidence interval [CI]: 1.339-12.727; p = 0.014) and bilateral or contralateral cervical LNM (OR: 11.577, 95% CI: 2.135-62.789; p = 0.005) to be significantly correlated with RPLNM. The RPLNM was found in 9 (32%) of the 28 patients with PW tumors. Of these 9 patients, 2 (7%) had ipsilateral RPLNM, 3 (11%) had contralateral RPLNM, and 4 (14%) had bilateral RPLNM. Fisher's exact test revealed the PW tumors to be significantly associated with contralateral RPLNM (p < 0.001). Retro-styloid LNM (RSLNM) was confirmed in two (1%) patients, both of whom had ipsilateral RSLNM with lymph nodes (LNs) of ≥15 mm in the upper limit of ipsilateral level II. Fisher's exact test revealed the LN of ≥15 mm in the upper limit of ipsilateral level II to be significantly associated with ipsilateral RSLNM (p = 0.001). The RPLNM was identified in 13% of patients with HPC. The PW tumors and bilateral or contralateral cervical LNM were risk factors for RPLNM; particularly, PW tumors were a specific risk factor for contralateral RPLNM. Although the RSLNM was rare, LNs of ≥15 mm in the upper limit of ipsilateral level II were a risk factor for ipsilateral RSLNM.

Performance Metric Analysis of a Radiation Oncology Educational Podcast

Asynchronous podcast education is a popular supplementary tool with up to 88% of medical residents reporting its use, and is perceived by faculty to have high educational value with convenience and connection to broader medical communities.1,2 Radiation oncology (RO) podcasts remain scarce compared to other specialties, and ones focused exclusively on education are largely absent. We analyze the early performance, listenership, and engagement of the first RO medical education podcast. Episode data and listener demographics were gathered from Spotify and iTunes. Episode engagement was defined as a percentage of plays on unique devices playing >40% of an episode within a single session. Listenership was defined by the number of plays per day (ppd) on unique devices, averaged over 60 days from publication date. Episodes were case based, categorized by disease subsite, and reviewed by a board-certified radiation oncologist. Quantitative endpoints included episode engagement and listenership. Qualitative comments were not solicited but received through email and Twitter. Pearson's correlation coefficient calculations were used for analysis. Eighteen total episodes had 8,517 total plays since July 2022 over 176 days. Median episode length was 13.8 minutes (range 9.2-20.1). Popular listening platforms included iTunes (53.5%) and Spotify (34.0%). Listener demographics included 59.4% male, 39.6% female, and 1.0% other, ranging from age 23-27 (14%), 28-34 (65%), 35-44 (14%), 45-59 (4%), and 60+ (1%). ATB was played in 48 countries, with the most listeners in North America (74.6%) followed by Asia (7.8%), Europe (7.6%), Australia (7.0%), Africa (2.0%), and South America (0.4%). There was a 464% increase in listenership since publication with median growth of 63.3% per month. Median listenership and engagement were 9.2 ppd (IQR, 7.7-9.9) and 77.8% (IQR, 68.1-81.2) for all episodes, respectively. Among 8 topics, head and neck (HN) episodes had the highest mean listenership with 17.8 ppd, followed by genitourinary (GU, 10.8) and lung (10.5). GU episodes had the highest mean engagement at 84.6%, followed by lung (82.3) and sarcoma (81.2). Dosimetry had the lowest listenership and engagement at 5.9 ppd and 63.1%, respectively. A significant negative relationship between episode length and engagement was observed, (r(18) = -0.469, p = 0.05). There was no statistically significant relationship between ppd and episode length, (r(18) = -0.303, p = 0.22). Evidenced by its significant rise in listenership, high listener engagement, and large international audience, there were previously unmet needs for RO medical education that may be supplemented by podcasts. HN episodes were most popular with GU exhibiting highest engagement. Longer episode length correlated with a significant decrease in engagement but no effect on popularity.

Crowdsourcing Deep Learning Algorithms to Automatically Contour GI Luminal Organs on Serial MRIs

A major barrier to a sustainable real-time adaptive MR-guided radiotherapy workflow is the time-consuming process of contouring the target and organs-at-risk (OARs) before the delivery of each fraction. While auto-contouring algorithms perform relatively well for many solid organs, the performance on luminal organs in the abdomen remain difficult due to the variability between patients and variability in daily shape and position. The purpose of this study is to evaluate the performance of crowdsourced deep learning algorithms to automatically contour GI luminal organs on serial MRIs. The stomach, small intestines, and large intestines were manually contoured on MRIs from patients who had undergone radiotherapy on an MR-Linac by a team of radiation therapists and medical physicists and were verified by a board-certified radiation oncologist. The MRIs and the contours were de-identified and uploaded to Kaggle, an online machine learning competition platform with portion of the data open to the public as training data and the remaining data hidden as a test set. Prize money was offered to teams submitting the best auto-contouring algorithms based on the Dice coefficient and Hausdorff distance evaluation metrics. The average performance of the winning algorithm and of manual contours were compared using unpaired t-test. Four hundred sixty-seven MRIs were collected from 107 patients who underwent 1-5 serial MRI sessions between 2015 and 2019. The most common anatomic site of treatment was the pancreas with 41 patients, followed by the liver with 38 patients. The manual contours of the stomach, small intestines, and large intestines on 4 representative MRIs had mean and standard deviation Dice coefficient of 0.90 +/- 0.02, 0.76 +/- 0.04, and 0.85 +/- 0.03 respectively and Hausdorff distance of 18.0 +/- 6.9, 35.5 +/- 12.6, and 32.3 +/- 12.3 mm respectively. The Kaggle competition was held from April to July 2022 and 1548 teams submitted algorithms for evaluation. The auto-contouring performance of the winning solution on the stomach, small intestines, and large intestines, when evaluated on a hold-out test set with 188 MRIs, had mean and standard deviation Dice coefficient of 0.92 +/- 0.04, 0.80 +/- 0.09, 0.85 +/- 0.08 respectively and Hausdorff distance of 15.2 +/- 11.2, 33.9 +/- 15.2, 34.8 +/- 20.3 mm respectively. Unpaired t-test was performed to compare the average performance across three organs of the human (N = 120) and of the algorithm (N = 564). The results suggest that average algorithm performance was statistically superior to manual contours for Dice (p = 0.01), yet not for Hausdorff (p = 0.64). Crowdsourced deep learning algorithms to auto-contour GI luminal organs on serial MRIs perform superiorly compared to manual contours when using a Dice coefficient metric but not when using a Haudorff distance metric. These auto-contouring algorithms may be used to efficiently adapt radiotherapy plans according to the anatomy of the day for patients with abdominal tumors on MR-Linacs.

Evaluation of a Deep Image to Image Network (DI2IN) Auto-Segmentation Algorithm across a Network of Cancer Centers

Due to manual OAR contouring challenges, various automatic contouring solutions have been introduced. Historically, common clinical auto-segmentation algorithms used were atlas based, which required maintaining a library of self-made contours. Searching the collection was computationally intensive and could take several minutes to complete. Deep learning approaches have shown significant benefits compared to atlas-based methods in improving segmentation accuracy and efficiency in auto-segmentation algorithms. This work represents the first multi-institutional study to describe and evaluate an AI algorithm for auto-segmentation of organs at risk (OARs) based on a deep image-to-image network (DI2IN). The AI-Rad Companion Organs RT (AIRC) algorithm (Siemens Healthineers, Erlangen, Germany) uses a two-step approach for segmentation. In the first step, the target organ region in the optimal input image is extracted using a trained Deep Reinforcement Learning network (DRL), which is then used as input to create the contours in the second step based on DI2IN. The study was initially designed as a prospective single-center evaluation. The automated contours generated by AIRC were evaluated by three experienced board-certified radiation oncologists using a 4-point scale where 4 is clinically usable and 1 requires re-contouring. After seeing favorable results in a single-center pilot study, we decided to expand the study to 6 additional institutions, encompassing 8 additional evaluators for a total of 11 physician evaluators across 7 institutions. One hundred fifty-six patients and 1366 contours were prospectively evaluated. The 5 most commonly contoured organs were the lung (136 contours, average rating = 4.0), spinal cord (106 contours, average rating = 3.1), eye globe (80 contours, average rating = 3.9), lens (77 contours, average rating = 3.9), and optic nerve (75 contours, average rating = 4.0). The average rating per evaluator per contour was 3.6. On average 124 contours were evaluated by each evaluator. 65% of the contours were rated as 4 and 31% were rated as 3. Only 4% of contours were rated as 1 or 2. 33 organs were evaluated in the study, with 19 structures having a 3.5 or above average rating (ribs, abdominopelvic cavity, skeleton, larynx, lung, aorta, brachial plexus, lens, eye globe, glottis, heart, parotid glands, bladder, kidneys, supraglottic larynx, submandibular glands, esophagus, optic nerve, oral cavity) and the remaining organs having a rating of 3.0 or greater (female breast, proximal femur, seminal vesicles, rectum, sternum, brainstem, prostate, brain, lips, mandible, liver, optic chiasm, spinal cord, spleen). No organ had an average rating below 3. AIRC performed well with greater than 95% of contours accepted by treating physicians with no or minor edits. It supported a fully automated workflow with the potential for time savings and increased standardization with the use of AI-powered algorithms for high quality OAR contouring.

Reappraisal of Treatment Volume Determination for Parametrial Boosting in Patients with Locally Advanced Cervical Cancer

Objective: In this study, we evaluated treatment volume determination for parametrial boosting in patients with locally advanced cervical cancer. Materials and Methods: Primary goal of this study has been to evaluate treatment volume determination for parametrial boosting based on CT only or fused CT-MRI in patients with locally advanced cervical cancer. We have carried out a comparative analysis of treatment volume determination by CT simulation images only or by integration of MRI. While we primarily focused on evaluation of incorporated multimodality imaging for treatment volüme determination, we also assessed critical organ contouring along with interobserver and intraobserver variations. Ground truth target volume has been utilized for comparative analysis, and it was determined by board certified radiation oncologists after detailed evaluation of all imaging and relevant data with thorough colleague peer review and consensus.

Prevalence and risk factors for retropharyngeal and retro-styloid lymph node metastasis in hypopharyngeal carcinoma

BackgroundWe evaluated the prevalence and identified the risk factors for retropharyngeal and retro-styloid lymph node metastasis (LNM) in patients with hypopharyngeal carcinoma (HPC). This was achieved using a combination of magnetic resonance (MR) and [18 F]-fluoro-2-deoxy-D-glucose (FDG)–positron emission tomography (PET)/computed tomography (CT) images.MethodsTwo board-certified radiation oncologists retrospectively reviewed pretreatment FDG–PET/CT images and contrast-enhanced thin-slice CT and MR images of 155 patients with HPC who underwent radiotherapy. Fisher’s exact tests and logistic regression analyses were performed to assess the risk factors for LNM.ResultsRetropharyngeal LNM (RPLNM) was confirmed in 20 (13%) patients. Posterior wall (PW) tumors (odds ratio [OR]: 4.128, 95% confidence interval [CI]: 1.339–12.727; p = 0.014) and bilateral or contralateral cervical LNM (OR: 11.577, 95% CI: 2.135–62.789; p = 0.005) were significantly correlated with RPLNM. The RPLNM was found in 9 (32%) of the 28 patients with PW tumors. Of these 9 patients, 2 (7%) had ipsilateral RPLNM, 3 (11%) had contralateral RPLNM, and 4 (14%) had bilateral RPLNM. The PW tumors were significantly associated with contralateral RPLNM (p < 0.001). Retro-styloid LNM (RSLNM) was confirmed in two (1%) patients, both of whom had ipsilateral RSLNM with lymph nodes (LNs) of ≥ 15 mm in the upper limit of ipsilateral level II. A significant association was found between LNs of ≥ 15 mm in the upper limit of ipsilateral level II and ipsilateral RSLNM (p = 0.001).ConclusionsThe RPLNM was identified in 13% of patients with HPC. The PW tumors and bilateral or contralateral cervical LNM were risk factors for RPLNM; particularly, PW tumors were a specific risk factor for contralateral RPLNM. Although the RSLNM was rare, LNs of ≥ 15 mm in the upper limit of ipsilateral level II were a risk factor for ipsilateral RSLNM.

Performance Analysis of a Radiation Oncology Educational Podcast

PurposeAsynchronous podcast education is a popular supplementary tool, with up to 88% of medical residents reporting its use. Radiation oncology podcasts remain scarce. The authors analyzed the early performance, listenership, and engagement of the first education-specific radiation oncology medical podcast. MethodsEpisode data and listener demographics were gathered from Spotify and Apple Podcasts. Episodes were case based, categorized by disease subsite, and reviewed by a board-certified radiation oncologist. Listenership was defined by the number of plays per day (ppd) on unique devices, averaged up to 60 days from publication. Episode engagement was defined as a percentage of plays on unique devices playing >40% of an episode within a single session. Quantitative end points included episode engagement and listenership. Pearson’s correlation coefficient calculations were used for analysis. ResultsFrom July 2022 to March 2023, 20 total episodes had 13,078 total plays over 227 days. The median episode length was 13.8 min (range, 9.2-20.1 min). Listener demographics were as follows: 54.4% men, 44.0% women, 1.3% not specified, and 0.3% nonbinary, with ages 18 to 22 (1%), 23 to 27 (13%), 28 to 34 (58%), 35 to 44 (22%), 45 to 59 (4%), and ≥60 (2%) years. Episodes were played in 53 countries, with the most plays in North America (71.5%), followed by Asia (10.2%), Europe (8.2%), Oceania (8.0%), Africa (1.5%), and South America (0.5%). There was a 585.2% increase in listenership since initiation, with median growth of 46.0% per month. Median listenership and engagement were 11.3 ppd (interquartile range, 10.3-13.8 ppd) and 81.4% (interquartile range, 72.0%-84.2%) for all episodes, respectively. A significant negative relationship between episode length and engagement was observed (r[20] = −0.51, P = .02). There was no statistically significant relationship between ppd and episode length (r[20] = −0.19, P = .42). ConclusionsThe significant rise in listenership, high episode engagement, and large international audience support a previously unmet need in radiation oncology medical education that may be supplemented by podcasts.

Shape reconstruction for undetectable regions of abdominal organs based on a graph convolutional network

Although computed tomography (CT) and magnetic resonance imaging (MRI) produce high-resolution images, during surgery or radiotherapy, only low-resolution cone-beam CT and low-dimensional X-ray images are usually obtained. Furthermore, because the duodenum and stomach are filled with air, it may be hard to accurately segment their contours, even on high-resolution images. Although denoising and image enhancement can be used to assist in the segmentation and reconstruction of 3D images, some regions of organs may remain difficult to detect. To approach this issue, we propose a graph convolutional network (GCN) to reconstruct organs that are hard to detect on medical images. The method uses a framework to transform an initial template into an estimated shape of the patient’s organs based on previous information attainted from different detectable organs from different patients. To overcome the inaccurate estimation results of triangular surface mesh data with the GCN, we propose a method to add internal vertices and edges, which has the effect of improving the calculation accuracy by enhancing the topological structure of the mesh data. In this study, the organ data of 124 patients who underwent pancreatic tumor treatment were used to verify the predictive effect of the proposed method. The 3D organ contours were defined by board-certified radiation oncologists and used as the ground truth surface meshes. For 100 randomly selected detectable feature vertices, the average Euclidean distance error of the liver was 3.72 mm. The topological enhancement proposed in this study reduced the prediction error of the liver by an average of 0.84 mm.

Patterns and Longitudinal Changes in the Practice of Breast Cancer Radiotherapy in Korea: Korean Radiation Oncology Group 22-01.

We aimed to analyze contemporary practice patterns in breast cancer radiotherapy (RT) and assess longitudinal changes over five years in Korea. In 2022, a nationwide survey was conducted among board-certified radiation oncologists. The survey consisted of 44 questions related to six domains: hypofractionated (HypoFx) whole breast RT, accelerated partial breast RT (APBI), regional nodal irradiation (RNI), RT for ductal carcinoma in situ (DCIS), postmastectomy RT (PMRT), and tumor bed boost. Seventy radiation oncologists from 61 (out of 101; 60%) institutions participated in the survey. HypoFx RT was used by 62 respondents (89%), a significant increase from 36% in 2017. HypoFx RT is commonly administered at 40-42.5 Gy in 15-16 fractions. APBI was used by 12 respondents (17%), an increase from 5% in 2017. The use of RNI did not change significantly: ≥ pN2 (6%), ≥ pN1 (33%), and ≥ pN1 with pathological risk factors (61%). However, indications for internal mammary lymph node (IMN) irradiation have expanded. In particular, the rates of routine treatment of IMN (11% from 6% in 2017) and treatment in cases of ≥ pN2 (27% from 14% in 2017) have doubled; however, the rate of treatment for only IMN involvement, identified on imaging, has decreased from 47% in 2017 to 31%. For DCIS, the use of HypoFx RT increased from 25% in 2017 to 75%, and the rate of RT omissions after breast-conserving surgery (BCS) decreased from 48% in 2017 to 38%. The use of HypoFx RT for PMRT increased from 8% in 2017 to 36%. The adoption of HypoFx RT after BCS for invasive breast cancer and DCIS has increased significantly, whereas the use of HypoFx PMRT has increased moderately since 2017. However, further studies are required to determine the optimal use of RNI.

Prevalence and Risk Factors of Retro-Styloid Lymph Node Metastasis in Oropharyngeal Carcinoma

<h3>Purpose/Objective(s)</h3> The assessment of the prevalence of retro-styloid lymph node metastasis (RSLNM) and associated risk factors may guide the selection criteria of level VIIb for inclusion in the target volume for high-precision radiotherapy (RT). However, data clarifying the selection criteria have not yet been established. We evaluated the prevalence of level VIIb RSLNM and associated risk factors in patients with oropharyngeal carcinoma (OPC). <h3>Materials/Methods</h3> Two board-certified radiation oncologists with an experience of 15 and 17 years in diagnosing and treating head and neck cancers retrospectively reviewed pretreatment [¹⁸F]-fluoro-2-deoxy-D-glucose–positron emission tomography/computed tomography (CT) along with contrast-enhanced thin slice CT and magnetic resonance (MR) images of 137 patients pathologically confirmed as having OPC who underwent RT. Observers independently evaluated the images without prior knowledge regarding the clinical information of the patients, and disagreements were resolved by consensus. The location of LN was confirmed on the planning CT images. In addition to RSLNM assessment, which includes the short-axis diameter and the maximum standardized uptake value (SUV_max), long-axis diameter of the largest LN in the upper limit of ipsilateral level II (caudal edge of the C1 lateral process) was recorded. Fisher's exact test and logistic regression analyses were made to determine the risk factors of RSLNM. <h3>Results</h3> RSLNM was confirmed in 18 (13%) patients. All RSLNMs were located within level VIIb on the planning CT images. The median short-axis diameter and SUV_max of RSLNM were 11 mm (range 8–15) and 5.2 (range, 1.9–9.4), respectively. No patients exhibited LNM in contralateral level VIIb. Furthermore, no patients with negative or single ipsilateral cervical LNM had RSLNM. Fisher's exact test revealed that smoking status (p = 0.027), multiple ipsilateral cervical LNM (p = 0.045), and LN ≥ 15 mm in the upper limit of ipsilateral level II (p < 0.001) were significantly associated with RSLNM. Logistic regression analyses revealed that the only factor significantly associated with RSLNM was LN ≥15 mm in the upper limit of ipsilateral level II (odds ratio: 977.297, 95% CI: 57.629–16573.308; p < 0.001). <h3>Conclusion</h3> RSLNM is relatively common in patients with OPC with a prevalence rate of approximately 10%. The prevalence of RSLNM in patients with negative or single ipsilateral cervical LNM and contralateral RSLNM is extremely low; therefore, level VIIb can be excluded from the target volume in such patients. LNs ≥ 15 mm in the upper limit of ipsilateral level II is a risk factor for RSLNM. Ipsilateral level VIIb should be included in the target volume for patients with this risk factor.

Impact of metal artifact reduction algorithm on gross tumor volume delineation in tonsillar cancer: reducing the interobserver variation

BackgroundPatients with tonsillar cancer (TC) often have dental fillings that can significantly degrade the quality of computed tomography (CT) simulator images due to metal artifacts. We evaluated whether the use of the metal artifact reduction (MAR) algorithm reduced the interobserver variation in delineating gross tumor volume (GTV) of TC.MethodsEighteen patients with TC with dental fillings were enrolled in this study. Contrast-enhanced CT simulator images were reconstructed using the conventional (CTCONV) and MAR algorithm (CTMAR). Four board-certified radiation oncologists delineated the GTV of primary tumors using routine clinical data first on CTCONV image datasets (GTVCONV), followed by CTCONV and CTMAR fused image datasets (GTVMAR) at least 2 weeks apart. Intermodality differences in GTV values and Dice similarity coefficient (DSC) were compared using Wilcoxon’s signed-rank test.ResultsGTVMAR was significantly smaller than GTVCONV for three observers. The other observer showed no significant difference between GTVCONV and GTVMAR values. For all four observers, the mean GTVCONV and GTVMAR values were 14.0 (standard deviation [SD]: 7.4) cm3 and 12.1 (SD: 6.4) cm3, respectively, with the latter significantly lower than the former (p < 0.001). The mean DSC of GTVCONV and GTVMAR was 0.74 (SD: 0.10) and 0.77 (SD: 0.10), respectively, with the latter significantly higher than that of the former (p < 0.001).ConclusionsThe use of the MAR algorithm led to the delineation of smaller GTVs and reduced interobserver variations in delineating GTV of the primary tumors in patients with TC.