Medical Specialties Commons^™

Renal Cell Carcinoma With Fibromyomatous Stroma Has A Distinctive Staining Pattern With Gata3 And Cd133 From Clear Cell Rcc And Clear Cell Papillary Rcc (Ccp-Rcc), Brandon Metcalf, Jining Feng, Leilei Chen, Vaishali Pansare, Ping L. Zhang

Conference Presentation Abstracts

Context: Renal cell carcinoma with fibromyomatous stroma (RCC-FMS) is a provisional entity of renal cell neoplasm that can be difficult to distinguish from low-grade clear cell RCC and clear cell papillary renal cell carcinoma (CCP-RCC) owing to its fibromyomatous stroma, low-grade nuclei, and clear cell cytoplasm.

Design: In this pilot study, we evaluated the utility of GATA3 and CD133 (a stem cell marker) as immunostain markers for the standard panel of cytokeratin 7 (CK7), CD10, and CAIX in distinguishing RCC-FMS from clear cell RCC and CCP-RCC. We stained 6 RCC-FMSs, 5 clear cell RCCs, and 5 CCP-RCC samples for all …

Monoclonal Immunoglobulin-Associated Glomerulopathy: Single Health Center Experience With 16 Cases, Brandon Metcalf, Yazan A. Al-Othman, James Huang, Wei Li, Hassan D. Kanaan, Ping L. Zhang

Conference Presentation Abstracts

Context: In addition to monoclonal cryoglobulinemic glomerulopathy, nonorganized monoclonal immunoglobulin (Ig) can deposit in glomeruli, causing either proliferative glomerulonephritis or membranous glomerulopathy. We present our experience related to this phenomenon.

Design: Sixteen cases of monoclonal immunoglobulin-associated glomerulopathy (MIAGN) from 14 patients were identified during the past 13 years and their pathologic and clinical features were collected and presented.

Results: One case of membranous glomerulopathy with monoclonal immunoglobulin deposits (MGNMID) (also called membranous glomerulopathy with light-chain–restricted deposits) was found. The patient was a 71-year-old woman with serum creatinine at 0.7 mg/dL and nephrotic syndrome (urine protein to creatinine ratio at 5.41), …

Igm-Dominant Immune Complex-Mediated Glomerulonephritis Should Be Classified As Membranoproliferative Glomerulonephritis, Igm Subtype, Sarang Khan, Brandon Metcalf, Wei Li, Hassan Kanaan, Ping L. Zhang

Conference Presentation Abstracts

Context: Immunoglobulin M (IgM) glomerulopathy usually presents as drug-resistant nephrotic syndrome and is characterized by diffuse effacement of foot processes and scattered mesangial IgM deposits, which is considered within the spectrum of minimal change disease and primary focal segmental glomerulosclerosis. However, IgM-dominant immune complex–mediated glomerulonephritis (IgM-ICMGN) is a rare renal entity and usually characterized by a membranoproliferative pattern with light microscopy; dominant IgM staining by immunofluorescent staining; and subendothelial, mesangial, and occasional subepithelial deposits by electron microscopy.

Design: We report our experience of 5 IgM-ICMGN cases with their pathologic phenotypes and clinical scenarios.

Results: Patient ages ranged from 47 to …

Combined Smad4 And Imp3 Immunohistochemical Stains To Increase The Probability Of Diagnosing Pancreatic Adenocarcinomas In Fine-Needle Biopsies, Reenal Patel, Ashbita Pokharel, Ping L. Zhang, Zhenhong H. Qu

Conference Presentation Abstracts

Context: SMAD4 and IMP3 are useful stains in diagnosing pancreatic cancer but not perfect to reach that goal individually. We attempted to determine if using both SMAD4 and IMP3 can increase the diagnostic specificity of fine-needle biopsies (FNBs) for pancreatic adenocarcinomas, as FNBs often showed limited tissue for the morphologic evaluation.

Design: Fifteen FNBs with or without prior cytology were evaluated with both SMAD4 and IMP3 stains. Nuclear loss of SMAD4 was graded as 0 (no loss), 1–2+ (partial loss), and 3+ (total loss), whereas IMP3 was graded 0–3+ by staining intensity. Staining results were correlated with morphologic changes for …

Doac Therapy Associated With Macroscopic Hematuria: Investigation Of The Underlying Urologic Etiologies, Catherine Xie, Mary Wyrzykowski, Subhash Edupuganti, Ming Xie

Conference Presentation Abstracts

Context: Direct oral anticoagulants (DOACs) have gained wide use in prevention and treatment of thromboembolic diseases. Both major and nonmajor bleeding events have been reported. Hematuria is a common bleeding complication in patients receiving DOACs.

Design: This study retrospectively investigates the underlying urologic etiologies responsible for DOAC therapy–related macroscopic hematuria. During a 1-year period, hospitalized patients with prolonged PTT and DOAC therapy–related hemorrhage were identified. Medical chart review focused on urology consult and related laboratory and imaging findings.

Results: Among 93 patients with DOAC therapy–related hemorrhage, 15 had macroscopic hematuria (16%), including 11 men and 4 women with a median …

Using An Artificial Intelligence (Ai) Tool (Chatgpt) To Identify The Main Indications And Side Effects Of Drugs In Patient Medical Records, Mai Elzieny, Zhenhong Qu

Conference Presentation Abstracts

Context: Drug history is important clinical information for the correct histologic diagnosis of a wide variety of nonneoplastic diseases, such as hepatitis, gastritis, and colitis. As pathologists, however, keeping a strong knowledge of all drugs (and variations of drug names) can be challenging. Using online databases to find relevant drug information one at a time is very time-consuming.

Design: Sixty-seven drugs were collected from the medical records of 5 randomly selected patients. ChatGPT (https://chat.openai.com/chat) was instructed to identify (1) the main clinical indications, and (2) the major side effects of each drug, using the prompt “List the major indications and …

Audit Tumor Reports In Pathology For Completeness And Compliance By An Artificial Intelligence Tool, Chatgpt, Ashbita Pokharel, Keran Zhao, Reenal Patel, Jack W. Qu, Zhenhong Qu

Conference Presentation Abstracts

Context: A tumor synoptic report (TSR) audit is an important quality control tool for compliance with reporting requirements; however, no simple audit tool is readily available to most pathologists. ChatGPT (https://chat.openai.com/chat) is an emerging artificial intelligence tool with wide applications. Its ability to extract data from text content has been widely reported but lacks evidence-based validation for its application in medicine. We assess the effectiveness of ChatGPT to identify and extract key data elements of TSRs and detect missing elements.

Design: Forty synoptic reports of malignant tumors of the lung, kidney, colon, and uterus were collected randomly. Key data elements …

Enhancing Pathology Report Analysis With Artificial Intelligence: A Validation Study Of Chatgpt's Effectiveness In 444 Cases, Zhenhong Qu, Megan Wang, Christian Santiago, Jack W. Qu, Ping Zhang

Conference Presentation Abstracts

Context: Despite years of efforts, the pathology report lacks standardization for its gross description. This inevitably leads to the occasional omission of key data elements, and difficulty to identify key elements in a lengthy unstructured report. A novel artificial intelligence (AI) tool, ChatGPT by OpenAI (San Francisco, California), was reportedly proficient in extracting keywords from text content. We assess the effectiveness of ChatGPT to quickly find multiple data elements in reports and detect missing data elements.

Design: Gross description of 444 pathology reports of several common organ types was submitted to ChatGPT (https://chat.openai.com/chat) that extracted and listed key data elements …

Refractive Outcomes Of Phacoemulsification Wtih Endocyclophotocoagulation Vs. Phacoemulsification Alone, Truman Silvasi

Conference Presentation Abstracts

No abstract provided.

Richter Syndrome And The Retina, Sahal Saleh

Conference Presentation Abstracts

No abstract provided.

Growth Hormone Deficiency: When Is Mri Imaging Indicated?, Haya Azouz, Kurt Tech, M Bassem Dekelbab

Conference Presentation Abstracts

Objectives The diagnosis of Growth Hormone Deficiency (GHD) is made with a combination of clinical assessment, measurement of insulin-like growth factor 1 (IGF-1) and insulin-like growth factor binding protein-3 (IGFBP-3) levels, and GH stimulation tests. Provocative (stimulation) GH testing is indicated for most patients to confirm a diagnosis of GHD. Traditionally, most pediatric endocrinologists defined a "normal" response by a serum GH concentration of >10 ng/ml. After the clinical and biochemical diagnosis of GHD is made, it is typical to obtain an MRI of the brain with narrow cuts through the hypothalamic-pituitary area. The aim of our study is to …

Calculating Differences In Surrogate Lung Perfusion Measurements Between Smokers And Nonsmokers As A Biomarker To Assess Their Potentials For Developing Chronic Obstructive Pulmonary Disease, A Prakash, A Nowacki, Y K. Liu, J Cisneros-Paz, Girish Nair, E Castillo

Conference Presentation Abstracts

Purpose: COPD (chronic obstructive pulmonary disease) is a collection of lung diseases that complicates breathing and causes irreversible lung damage. Given that it is a progressive disease, early detection is crucial for managing its symptoms. We recently developed a method for calculating pulmonary blood mass change (PBMC) apparent on paired inhale/exhale non-contrast computed tomography (CT) scans as a quantitative surrogate for lung perfusion. Our long term objective is to develop a PBMC biomarker for predicting COPD progression. As a first step, we hypothesize that PBMC is significantly lower for non-COPD smokers versus non-COPD non-smokers. Methods: CT scans from 77 healthy …

57 Yo Man With High Grade Astrocytoma With Piloid Features And 2 Other Cns Lesions, Joseph Fullmer

Conference Presentation Abstracts

Patient is a 57 yo man who presented with periodic headaches and lower back pain, initially attributed to lifting heavy weights but symptoms persisted. Eventual imaging was performed and revealed a 4 cm heterogeneously enhancing left cerebellar mass, 1 cm enhancing lateral right frontal subcortical lesion, and enhancing T12-L1 spinal cord lesion and multiple smaller spinal cord lesions suggestive of drop metastasis. The cerebellar lesion was first resected and final results, including methylome profiling, were consistent with high grade astrocytoma with piloid features (HGAP). The T12-L1 spinal lesion was later resected and had somewhat distinct histology although intraoperatively surgeon felt …

Does Proton Arc Therapy Has Potential Clinical Benefits For Hepatocellular Carcinoma?, Peilin Liu, R Dalfsen, M Soukup, Shupeng Chen, Lewei Zhao, Z Wang, X S. Gao, Thomas Quinn, Craig Stevens, Rohan Deraniyagala, Xiaoqiang Li, Xuanfeng Ding

Conference Presentation Abstracts

Purpose: To investigate the potential clinical benefits and dose-averaged Linear Energy Transfer (LETd) sparing, utilizing proton arc plan for Hepatocellular carcinoma (HCC) patients in comparison with Intensity Modulated Proton Therapy (IMPT). Methods: Ten HCC patients have been retrospectively selected. Two planning groups were created: Proton Arc plans using Monaco ver. 6 and the clinical IMPT plan. Both planning groups used the same robustness parameters. The prescription dose is 67.5 Gy (RBE) in 15 fractions/ 66 GyE in 10 fractions of the Clinical Target Volume (CTV) following clinical IMPT plan. Robustness evaluations were performed to ensure target dose coverage. Normal Tissue …

A Novel Planning Framework For The Efficient Spot-Scanning Proton Arc Therapy Via The Particle Swarm Optimization (Sparc-Particle Swarm), Y Qian, R Dao, G Liu, Xiaoqiang Li, H Quan, Xuanfeng Ding

Conference Presentation Abstracts

Purpose: The delivery efficiency is the bottleneck of spot-scanning proton arc therapy (SPArc) because of the numerous energy layers (EL) ascending switches. This study aims to develop a new algorithm to mitigate the need of EL ascending via water equivalent thickness (WET) sector selection followed by the particle swarm optimization (SPArc- particle swarm). Methods: SPArc- particle swarm divided the full arc trajectory into the optimal sectors based on the K-means clustering analysis of the relative mean WET. Within the sector, particle swarm optimization was used to minimize the total energy switch time, in which it optimized the energy selection integrated …

Application Of A Novel Optical Tool For Linac Calibrations, Cory S. Knill, Brian Loughery, Raminder P. Sandhu, Lifeng Lin, Mark Stefani

Conference Presentation Abstracts

Purpose: Maintaining a linac within tight stereotactic isocentricity tolerances requires minimizing errors in individual components of the linac. However, while isocentricity can be measured using a Winston-Lutz test, contributions from individual components can be difficult to discern. The disconnect between how individual components contribute to isocentricity can lead to lengthy troubleshooting exercises when isocentricity is outside tolerance. To provide more info rmation to physicists and increase the speed of isocentricity adjustments, a new tool that uses optical tracking to investigate and adjust individual linac components that contribute to isocentricity was tested on three clinical linacs. Methods: The optical tracking tool …

Individual Vmat Beam Complexity Indices Have Limited Predictive Value For Qa Pass-Rates, Jian Liang, Q Liu, David T. To, Michael G. Snyder

Conference Presentation Abstracts

Purpose: Previous studies with sample sizes <400 have indicated that VMAT beam complexity indices may have a relationship with measured QA pass-rates. The aim of this study is to evaluate such relationships on a large scale to further investigate their strength. Methods: Total of 17,273 VMAT QA results across 8 Elekta Agility machines were used in this study. All beams were flattened 6MV or 10MV. Sun Nuclear ArcCheck was used for QA measurements with γ-analysis parameters of 10% threshold, 3% dose difference, 2mm distance to agreement, and 95% passing threshold. A total 214 Beams (1.23%) failed QA. QA distributions in Pinnacle were calculated with 2 degree gantry intervals and a dose grid of 2mm. Four complexity indices (MCSv, MU weighted Segment Area, Segment Area/Perimeter, and Average Leaf Gap) were computed. Spearman’s correlations were evaluated for each, and cut-off values for the indices were sought to indicate values above which no QA failures would be expected. Results: Evaluation of the 4 complexity metrics showed weak correlation with QA pass rates (< ±0.42) with high significance (p < 0.001). When metrics were evaluated independently, cut-off values from correlation plots separated only 0.67%, 0.72%, 2.2%, 6.3% guaranteed passing beams from the total sample, respectively. When evaluating whether anymetric for a given beam exceeded its cut-off, the fraction of guaranteed passing beams rose to only 6.6%. However, when cut-off values for one metric were allowed to be functionally dependent upon another, the fraction of guaranteed passing beams increased. For MU weighted Segment Area andAverage Leaf Gap, a dependent approach separated 19.1% guaranteed passing beams from the sample. Conclusion: Individual complexity metrics have weak predictive utility in assessing plan performance. However, using multiple complexity cut-off functions with dependence upon more than one metric increased the fraction of predicted passing beams by around 300%.

A Comprehensive Investigation Of Linear Energy Transfer Optimization Effectiveness In Intensity-Modulated Proton Therapy Via Alternating Direction Method Of Multipliers (Impt Let-Admm), Q Fan, G Liu, Lewei Zhao, Xiaoqiang Li, Xuanfeng Ding, S Dai

Conference Presentation Abstracts

Linear energy transfer (LET)-guided optimization in intensity-modulated proton therapy (IMPT) via alternating direction method of multipliers (ADMM) has the potential to improve its biological effectiveness (IMPTLET- ADMM), in which the LET distribution in target can be escalated, while the LET distribution in the organs at risk can be mitigated. This study aims to quantitatively investigate the effectiveness of LET optimization in IMPT via ADMM with different solvers in its iteration loop. Methods: The clinical dose-volume-histogram (DVH) constraint noted dose sub-problem and clinical LET -volume-histogram (LVH) constraint noted LET sub-problem are combined to generate a composite objective function, which is available …

Integrative Analysis Of Metabolome And Methylome Identifies The Correlation Of Aminoacyl-Trna Biosynthesis Metabolitesand Sepsecs Gene Methylation In Huntington’S Disease Brain Tissue, Sangeetha Vishweswaraiah, Ali Yilmaz, Nadia Ashrafi, Nazia Saiyed, Abdulla Khalid, Purvesh R. Koladiya, Xiaobei Pan, Shirin Macias, Uppala Radhakrishna, Stewart F. Graham

Conference Presentation Abstracts

No abstract provided.

Particle Arc Single-Isocenter Stereotactic Radiosurgery Of Multiple Brain Metastases, L Volz, S Mein, T Tessonnier, Peilin Liu, A Mairani, M Durante, Xuanfeng Ding, C Graeff, T Li

Conference Presentation Abstracts

Purpose: To develop spot-scanning proton (SPArc) and heavy ion (SHArc) arc therapy for single-isocenter stereotatic radiosurgery (SISRS) of multiple brain metastases (MBMs). First results achieved with dedicated planning infrastructure are compared to state-of-the -art volumetric modulated photon arc therapy (HA-VMAT). Methods: A multi-institutional collaboration was established to develop bespoke SPArc and SHArc SISRS infrastructure with available research and commercial treatment planning systems. Devised planning and delivery methods considered dedicated energy, spot, and multi-arc selection strategies. SPArc and SHArc (carbon and helium ions) plans were generated for MBM patients exhibiting 3-12 intracranial lesions with GTV volumes between 0.03ccm and 19.8ccm, at …

Urinary Cytokines As Potential Biomarkers Of Mild Cognitive Impairment And Alzheimer’S Disease: A Pilot Study, Nazia Saiyed, Ali Yilmaz, Sangeetha Vishweswaraiah, Abhishek Maiti, Ilyas Ustun, Sarah Bartolone, Travonia Brown-Hughes, Roland Thorpe Jr, Tammy Osentoski, Stacey Ruff, Amita Pai, Michael E. Maddens, Khaled Imam, Stewart Graham

Conference Presentation Abstracts

No abstract provided.

An Analytic Model Of The External Beam Dose Rate Effect On Immune Cells In Blood Dosimetry, Lewei Zhao, Peilin Liu, G Liu, Xuanfeng Ding

Conference Presentation Abstracts

Purpose: Radiation therapy is a double-edged sword with regard to its immunomodulatory effect. It is critical to find how it is affected by dose rate. Thus, we developed an analytic model for calculating the number of immune cells receiving radiation. Methods: Suppose the blood is received radiation by constant dose rate. The blood is one-way flow without circulation in the radiation delivery at a constant velocity. Two kinds of lung treatment regimens were considered standard delivery 2Gy per fx total 30fx and hypo-fractionation SBRT 12 Gy per fx total 5 fx. Two different dose rates were simulated for standard treatment: …

Investigate The Delivery Efficiency Improvement For Sparc Compared To Impt, Peilin Liu, B A. De Jong, S Chang, Lewei Zhao, G Liu, Craig Stevens, Rohan Deraniyagala, E W. Korevaar, J A. Langendijk, Xiaoqiang Li, Xuanfeng Ding

Conference Presentation Abstracts

Purpose: To assess the efficiency improvement of SPArc compared to IMPT in proton clinical operation.Methods: A random clinical operation day at a proton therapy center with a two room system. Aa total of 10 Head & Neck cancer cases treated on Jan 6th, 2021were selected and SPArc plans were generated for all the cases with similar target coverage and dose constraints of organ at risk (OARs). 3% range uncertainties and 3mm setup uncertainties were applied for robust optimization. The log files of clinical IMPT were exported to calculate true treatment delivery time. The machine-specific delivery sequence model was utilized to …

Develop A Novel Beam Orientation Optimization Algorithm For Let (Linear Energy Transfer) Incorporated Impt (Intensity Modulated Proton Therapy) (Imptlet, R Dao, Y Qian, Xiaoqiang Li, H Quan, Xuanfeng Ding, G Liu

Conference Presentation Abstracts

Purpose: To improve the dose and LET distribution in the IMPTLET, we introduce a new optimization algorithm to search for the optimal beam orientation and count through the arc trajectory (BOC-IMPTLET). Methods: We introduce kmeans clustering analysis in the BOC-IMPTLET algorithm by analyzing the distribution of LET in the target from the full arc trajectory. The optimal of cluster count was used as beam count, which was determined based on the SSE (Sum of Square Error) value and the angle of the corresponding centroid of each cluster was used as beam orientation, respectively. The BOC-IMPTLET algorithm was implemented in the …

Predicting Tumor Voxel Dose-Response Using Deep Learning On Serial Fdg-Pet Feedback Images, Shupeng Chen, Di Yan, An Qin, Rohan Deraniyagala, Daniel Krauss, Peter Chen, Craig Stevens

Conference Presentation Abstracts

Purpose: Tumor voxel dose-response matrix (DRM) can be assessed using a series of FDG-PET/CT feedback images acquired during radiotherapy. Predicting the tumor voxel DRM earlier is crucial for treatment adaptation. However, it is also challenging due to FDG fluctuation in tumor cells. This study investigated the feasibility of predicting DRM during early treatment weeks using deep learning (DL). Methods: FDG-PET images were acquired at the pretreatment (preTx), the 2nd and 4th weeks during chemo-radiotherapy (35×2Gy) from 50 head and neck patients. The reference value of tumor voxel DRM (DRMref), representing the average metabolic change ratio during the treatment, was determined …

Linac-Based Srs Treatments Slow Down Unnecessarily Due To Parked Leaf Optimization, Brian Loughery, Cory S. Knill, Raminder P. Sandhu, Lifeng Lin, Jian Liang

Conference Presentation Abstracts

Purpose: It was observed on a VersaHD that clinical SRS deliveries created in the Elements treatment planning system (TPS) slowed gantry rotation and dose rate to allow parked MLC leaves to move into position for use in an upcoming control point. Gantry rotation slowed suddenly, visibly backtracking to previous gantry angles. The purpose of this work is to study the effect of parked MLC positioning parameters in Elements on cranial SRS delivery speed and gantry motion on a VersaHD. Methods: To study the effect of parked MLC leaf positions on VersaHD deliverability, cranial SRS plans were generated with different TPS …

Internal Lead Shielding For Clinical Electron Treatments, Patrick N. Mcdermott

Conference Presentation Abstracts

Purpose: Electron beams are often used to treat superficial lesions of the lip, cheek, nose and ear. Lead is frequently used to block distal structures. It is customary to place an internal bolus of low atomic number in between the tissue and the lead to reduce electron backscatter from the lead. Space for the lead and the internal bolus is quite limited. The current method, recommended by TG25, for estimating the thickness of the lead plus internal bolus is not self-consistent and leads to a larger than necessary thickness. Furthermore, this method uses an older expression for the electron backscatter …

Quantifying Disease Progression In Copd Patients By Forecasting Pft Scores Using Extracted Features From Non-Contrast Ct, A T. Luong, C J. Herrera, E Young, Y K. Liu, A Nowacki, J Cisneros-Paz, Girish Nair, E Castillo

Conference Presentation Abstracts

Purpose: Currently there is no validated method for quantifying risk of disease progression in chronic obstructive pulmonary disease (COPD). We aim to address this by predicting whether a patient will worsen using a machine learning model trained on basic patient information, current pulmonary function values, FEV1 and FVC, and extracted features from a non-contrast CT scan. Disease severity is characterized by these pulmonary function values, and by extension disease progression can be determined by the change in these values between future timepoints. Methods: XGBoost, a popular classification library that utilizes gradient boosted trees, was used to define an ensemble model …

Quantifying Disease Progression In Copd Patients By Forecasting Pft Scores Using Extracted Features From Non-Contrast Ct, A T. Luong, C J. Herrera, E Young, Y K. Liu, A Nowacki, J Cisneros-Paz, Girish Nair, E Castillo

Conference Presentation Abstracts

Purpose: Currently there is no validated method for quantifying risk of disease progression in chronic obstructive pulmonary disease (COPD). We aim to address this by predicting whether a patient will worsen using a machine learning model trained on basic patient information, current pulmonary function values, FEV1 and FVC, and extracted features from a non-contrast CT scan. Disease severity is characterized by these pulmonary function values, and by extension disease progression can be determined by the change in these values between future timepoints. Methods: XGBoost, a popular classification library that utilizes gradient boosted trees, was used to define an ensemble model …

The Clinical Usability, Efficiency, And Accuracy Of Combining Deep-Learning-Based Synthetic Ct And Auto-Contouring For Daily Cbct-Based Online Adaptive Radiation Therapy, An Qin, J Xu, N O'Connell, Xuanfeng Ding, Shupeng Chen, D Thill

Conference Presentation Abstracts

Purpose: To evaluate the feasibility of deep-learning and CBCT-based online adaptive radiotherapy(ART) in thorax, Head&Neck, and pelvic regions. Methods: The study utilized ADMIRE(Advanced Medical Image Registration Engine), a research software designed to enable fast online ART through: (1)synthetic CT(sCT) generation from CBCT, (2)automatic segmentation of critical organs(OARs), and(3)deformable image registration(DIR) to transfer the target volume. ADMIRE trained a modified Cycle-Generative-Adversarial-Networks with Structural Similarity Index Map(SSIM) loss term to generate sCT, while the auto-segmentation model was based on a 3D-ResUnet with modification. Site-specific models were trained for three anatomic sites: thorax, pelvic, and head&neck, and for both networks. The study included …