Variations In Patient Anatomy Research Articles

A method framework of semi-automatic knee bone segmentation and reconstruction from computed tomography (CT) images.

Accurate delineation of knee bone boundaries is crucial for computer-aided diagnosis (CAD) and effective treatment planning in knee diseases. Current methods often struggle with precise segmentation due to the knee joint's complexity, which includes intricate bone structures and overlapping soft tissues. These challenges are further complicated by variations in patient anatomy and image quality, highlighting the need for improved techniques. This paper presents a novel semi-automatic segmentation method for extracting knee bones from sequential computed tomography (CT) images. Our approach integrates the fuzzy C-means (FCM) algorithm with an adaptive region-based active contour model (ACM). Initially, the FCM algorithm assigns membership degrees to each voxel, distinguishing bone regions from surrounding soft tissues based on their likelihood of belonging to specific bone regions. Subsequently, the adaptive region-based ACM utilizes these membership degrees to guide the contour evolution and refine segmentation boundaries. To ensure clinical applicability, we further enhance our method using the marching cubes algorithm to reconstruct a three-dimensional (3D) model. We evaluated the method on six randomly selected knee joints. We evaluated the method using quantitative metrics such as the Dice coefficient, sensitivity, specificity, and geometrical assessment. Our method achieved high Dice scores for the femur (98.95%), tibia (98.10%), and patella (97.14%), demonstrating superior accuracy. Remarkably low root mean square distance (RSD) values were obtained for the tibia and femur (0.5±0.14 mm) and patella (0.6±0.13 mm), indicating precise segmentation. The proposed method offers significant advancements in CAD systems for knee pathologies. Our approach demonstrates superior performance in achieving precise and accurate segmentation of knee bones, providing valuable insights for anatomical analysis, surgical planning, and patient-specific prostheses.

Multi-institutional experimental validation of online adaptive proton therapy workflows

Objective. To experimentally validate two online adaptive proton therapy (APT) workflows using Gafchromic EBT3 films and optically stimulated luminescent dosimeters (OSLDs) in an anthropomorphic head-and-neck phantom. Approach. A three-field proton plan was optimized on the planning CT of the head-and-neck phantom with 2.0 Gy(RBE) per fraction prescribed to the clinical target volume. Four fractions were simulated by varying the internal anatomy of the phantom. Three distinct methods were delivered: daily APT researched by the Paul Scherrer Institute (DAPTPSI), online adaptation researched by the Massachusetts General Hospital (OAMGH), and a non-adaptive (NA) workflow. All methods were implemented and measured at PSI. DAPTPSI performed full online replanning based on analytical dose calculation, optimizing to the same objectives as the initial treatment plan. OAMGH performed Monte-Carlo-based online plan adaptation by only changing the fluences of a subset of proton beamlets, mimicking the planned dose distribution. NA delivered the initial plan with a couch-shift correction based on in-room imaging. For all 12 deliveries, two films and two sets of OSLDs were placed at different locations in the phantom. Main results. Both adaptive methods showed improved dosimetric results compared to NA. For film measurements in the presence of anatomical variations, the [min-max] gamma pass rates (3%/3 mm) between measured and clinically approved doses were [91.5%–96.1%], [94.0%–95.8%], and [67.2%–93.1%] for DAPTPSI, OAMGH, and NA, respectively. The OSLDs confirmed the dose calculations in terms of absolute dosimetry. Between the two adaptive workflows, OAMGH showed improved target coverage, while DAPTPSI showed improved normal tissue sparing, particularly relevant for the brainstem. Significance. This is the first multi-institutional study to experimentally validate two different concepts with respect to online APT workflows. It highlights their respective dosimetric advantages, particularly in managing interfractional variations in patient anatomy that cannot be addressed by non-adaptive methods, such as internal anatomy changes.

Enhancing adaptive proton therapy through CBCT images: Synthetic head and neck CT generation based on 3D vision transformers.

Proton therapy is a form of radiotherapy commonly used to treat various cancers. Due to its high conformality, minor variations in patient anatomy can lead to significant alterations in dose distribution, making adaptation crucial. While cone-beam computed tomography (CBCT) is a well-established technique for adaptive radiation therapy (ART), it cannot be directly used for adaptive proton therapy (APT) treatments because the stopping power ratio (SPR) cannot be estimated from CBCTimages. To address this limitation, Deep Learning methods have been suggested for converting pseudo-CT (pCT) images from CBCT images. In spite of convolutional neural networks (CNNs) have shown consistent improvement in pCT literature, there is still a need for further enhancements to make them suitable for clinicalapplications. The authors introduce the 3D vision transformer (ViT) block, studying its performance at various stages of the proposed architectures. Additionally, they conduct a retrospective analysis of a dataset that includes 259 image pairs from 59 patients who underwent treatment for head and neck cancer. The dataset is partitioned into 80% for training, 10% for validation, and 10% for testingpurposes. The SPR maps obtained from the pCT using the proposed method present an absolute relative error of less than 5% from those computed from the planning CT, thus improving the results ofCBCT. We introduce an enhanced ViT3D architecture for pCT image generation from CBCT images, reducing SPR error within clinical margins for APT workflows. The new method minimizes bias compared to CT-based SPR estimation and dose calculation, signaling a promising direction for future research in this field. However, further research is needed to assess the robustness and generalizability across different medical imaging applications.

Fusion extracted features from deep learning for identification of multiple positioning errors in dental panoramic imaging.

Dental panoramic imaging plays a pivotal role in dentistry for diagnosis and treatment planning. However, correctly positioning patients can be challenging for technicians due to the complexity of the imaging equipment and variations in patient anatomy, leading to positioning errors. These errors can compromise image quality and potentially result in misdiagnoses. This research aims to develop and validate a deep learning model capable of accurately and efficiently identifying multiple positioning errors in dental panoramic imaging. This retrospective study used 552 panoramic images selected from a hospital Picture Archiving and Communication System (PACS). We defined six types of errors (E1-E6) namely, (1) slumped position, (2) chin tipped low, (3) open lip, (4) head turned to one side, (5) head tilted to one side, and (6) tongue against the palate. First, six Convolutional Neural Network (CNN) models were employed to extract image features, which were then fused using transfer learning. Next, a Support Vector Machine (SVM) was applied to create a classifier for multiple positioning errors, using the fused image features. Finally, the classifier performance was evaluated using 3 indices of precision, recall rate, and accuracy. Experimental results show that the fusion of image features with six binary SVM classifiers yielded high accuracy, recall rates, and precision. Specifically, the classifier achieved an accuracy of 0.832 for identifying multiple positioning errors. This study demonstrates that six SVM classifiers effectively identify multiple positioning errors in dental panoramic imaging. The fusion of extracted image features and the employment of SVM classifiers improve diagnostic precision, suggesting potential enhancements in dental imaging efficiency and diagnostic accuracy. Future research should consider larger datasets and explore real-time clinical application.

Dosimetric Benefits of Online Adaptive Radiotherapy in Nasopharyngeal Carcinoma

Online adaptive radiotherapy (ART) has the advantage of compensating for potential underdosing to targets and overdosing to organs-at-risk (OARs) caused by variations in patient anatomy and tumor geometry. Artificial intelligence (AI)-assisted rapid generation of new plans makes online ART possible. We aimed to evaluate the dosimetric benefits of online ART on tumor coverage and OARs sparing in nasopharyngeal carcinoma (NPC). Twenty patients diagnosed with NPC (19 with stage III and 1 with stage II according to the 8th edition of the AJCC/UICC staging system) who underwent definitive radiotherapy or concurrent chemoradiotherapy and received online ART on CT-Linac between April 2022 and December 2022 were included in this study, consisting of 14 males and 6 females with a median age of 48 years (range: 29-68 years). The prescription dose was 6996 cGy/33 fractions for primary gross tumor volume (GTVp), 6600-6996 cGy/33 fractions for gross tumor volume of nodes (GTVn), 6006 cGy/33 fractions for high-risk clinical tumor volume (CTV1), 5412 cGy/33 fractions for low-risk clinical tumor volume (CTV2). The majority of the patients (15/20) received online ART during the fourth to fifth week of their radiotherapy treatment The auto-segmented contours and auto-plan generated by AI were manually reviewed and edited by radiotherapists and physicists. The paired samples t-test was used to compare the dose and volumes metrics of targets and OARs between scheduled plan and online ART plan. The results of this study showed that compared to the scheduled plan, the online ART plan resulted in significant reductions in the volumes of all targets and 8/12 OARs (temporal lobes, optic nerves, lenses, eyes, parotids, submandibulars, mandibles, and thyroid) (P<0.05). The online ART plan also improved target coverage, with D98% for GTVp in the scheduled plan compared to the online ART plan being 7063.4 ± 76.1 cGy and 7096.1 ± 53.9 cGy (P = 0.1), CTV1 being 6266.7 ± 114.9 cGy and 6208.7 ± 54.7 cGy (P<0.05), and CTV2 being 4142.5 ± 1700.9 cGy and 5416.4 ± 23.8 cGy (P<0.01), respectively. The dose to all 12 OARs was reduced with the use of online ART, with 5/12 OARs showing statistical significance. The D0.03cm3 for the spinal cord in the scheduled plan and online ART plan were 3630.9 ± 197.6 and 3454.1 ± 132.0 cGy; for the temporal lobes were 7075.2 ± 303.0 and 6994.2 ± 345.1 cGy; and 4396.0 ± 2575.0 and for the pituitary were 4214.5 ± 2499.2 cGy. Meanwhile the Dmean for the eyes in the scheduled plan and online ART plan was 769.0 ± 232.0 and 714.8 ± 200.1 cGy; and for the mandibles were 3187.7 ± 211.5 and 3066.0 ± 152.1 cGy. Online ART was effective in protecting most of the OARs in NPC patients, while simultaneously indicating a trend towards enhancing target coverage. This study demonstrated the promising potential of online ART for patients with NPC. This approach will be tested in an upcoming phase III trial.

Abstract P1-10-12: AI-based cardiac sub-structures segmentation for safer radiotherapy planning

Abstract Purpose: Whilst radiotherapy increases cure rates in breast cancer, lung cancer, among others, it may also involve some cardiac exposure, which in turn may increase the risk of different heart diseases. The heart is a complex anatomical organ that involves many different structures making it difficult to contour cardiac sub-structures reproducibly. Contouring, especially for these cases, suffers from inter- and intra-expert variability while being time consuming. Cardiac atlases have been developed to aid in the delineation of cardiac substructures. However, these methods have many shortcomings, including the inability to overcome variations in patient anatomy. In this study, a deep learning based commercial solution for automatic OAR delineation was trained following international guidelines for heart substructures delineation and tested on an unseen cohort of lung and breast patients to evaluate its clinical acceptability. Methods: ART-Net, a CE-marked, FDA-cleared anatomically preserving deep-learning ensemble architecture for automatic annotation of OAR was evaluated using data of 20 breast/lung patients from 2 centers. Automatic annotation of 27 different structures (Ventricles (left and right), atria (left and right), left ventricle (anterior, apical, inferior, lateral, septal), LAD coronary (mid, proximal, distal, total), circumflex coronary (distal, proximal, total), RCA (distal, mid, proximal, total), coronary sinus, left main coronary artery, ascending aorta, pulmonary arteries, vena cava inferior, vena cava superior and the heart) was performed and submitted to 2 experts across 2 centers for qualitative evaluation. Contours were scored as A/acceptable, B/acceptable after minor corrections, and C/not acceptable for clinical use. To avoid any bias, experts were blind to whether the contour were manually, or AI delineated. The DSC between automatic and manual (ground truth) contours of the heart sub-structures were evaluated and compared with interobserver variability from the literature [1,2] using average and min DSC scores. Results: Automatic contours were generated in a mean time of 0.5s per scan slice. Out of the 27 structures, 20 were considered clinically acceptable in the qualitative study. In the inter-expert variability study, 12 structures passed the test successfully using initial acceptance criterion over an acceptable sample size and 9 other structures demonstrated performances above the minimal threshold of inter-expert variability, sometimes on smaller datasets due to lack of manual data. Overall, 16 structures were included in the final model. 13 structures were considered clinically acceptable in 100% of the cases with AI contours rated at the same level as manual contours. For the other 3 structures (coronary sinus, left main coronary artery and vena cava inferior), the performance of the AI contours was slightly below that of the manual contours (within 3.4% difference), with the least performing structure being the coronary sinus (84% for AI vs 87% manual). Conclusion: We show first results for the evaluation of AI-based auto-contouring tool for annotation of the substructures of the heart. The results show very good clinical acceptance, highlighting the high usability of the commercial tool for cardiac cases and its clinical implementation feasibility. The use of this AI tool can facilitate and accelerate future research studies investigating relationships between substructure doses and cardiac outcomes. Future work will include improvement of the sub-structures (mid, proximal, distal) and a retrospective meta-analysis to assess heart sub-structures degree of importance in terms of toxicity.

Subject-specific computational modeling of acromioclavicular and coracoclavicular ligaments.

Disruption of the acromioclavicular joint is a common injury. Despite the different surgical procedures described for treating this injury, complications such as loss of reduction and failure of fixation remain unacceptably high. We developed a computer model of the acromioclavicular joint to better understand the biomechanical contributions of the ligaments that are typically injured. Six fresh frozen human cadaveric shoulders were tested on an AMTI VIVO 6-degree of freedom test platform to measure force-displacement in inferior translation, anteroposterior translation, and internal rotation before and after sequentially transecting the coracoclavicular and acromioclavicular ligaments. These data were used to construct computer models of each specimen. Three-dimensional computed tomographic scans were used to generate a rigid-body dynamics model using the AnyBody Modeling System. The scapula and clavicle were connected by the acromioclavicular joint capsule, the conoid ligament, and the trapezoid ligament. Subject-specific ligament properties were calculated by matching computer predictions to experimental force-displacement data. The calculated free lengths of the conoid, trapezoid, and acromioclavicular ligaments were 13.5 (±3.2), 11.8 (±2.4), and 11.0 (±2.7) mm, respectively. The calculated stiffnesses of the conoid, trapezoid, and acromioclavicular ligaments were 34.3 (±6.3), 28.4(±3.2), and 33.8 (±8.2) N/mm, respectively. Root mean square deviation (RMSD) of predicted force-displacement curves relative to experimental force-displacement curves (during inferior and anteroposterior translation) was less than 1 mm. For validation of subject-specific models, after ligament properties were calculated, the RMSD of the predicted torque over 15° of internal rotation was 12% of maximum rotational torque (average for 6 specimens). Acromioclavicular disruption results in multidirectional instability, which requires careful consideration of the individual contributions of the injured ligaments. In addition, variations in patient anatomy can significantly affect the biomechanical stability of the reconstruction. Subject-specific models can enhance our understanding of the individual and collective biomechanical contributions of the injured soft tissues to the multiaxial stability of the acromioclavicular joint. These models may also be useful for analyzing and assessing biomechanical stability after various types of surgical reconstruction.

Validation of prompt gamma-ray spectroscopy for proton range verification in tissue-mimicking and porcine samples

Objective. Proton therapy of cancer improves dose conformality to the target and sparing of surrounding healthy tissues compared to conventional photon treatments. However, proton therapy’s advantage could be even larger if proton range uncertainties were reduced. Sources of range uncertainties include computed tomography treatment planning images and variations in patient anatomy and setup. To reduce range uncertainties, we have developed a system for real-time in vivo range monitoring. The system is based on spectroscopy of prompt gamma-rays emitted through proton-nuclear interactions during irradiation. We validated the performance of our prompt gamma-ray spectroscopy detector prototype using tissue-mimicking and porcine samples. Approach. Measurements were performed in water, four tissue-mimicking samples (spongiosa, muscle, adipose tissue, and cortical bone), and two porcine samples (liver and brain). A dose of 0.9 Gy was delivered to a target at a depth of 12.5–17.5 cm. Multi-layer ionization chamber measurements were performed to determine stopping power ratios relative to water and ground truth proton ranges. Ground truth elemental compositions were determined using combustion analysis. Proton ranges and elemental compositions measured using prompt gamma-ray spectroscopy were compared to the ground truth. Main results. For all samples, the mean measured range over all pencil-beam spots differed from the ground truth by less than 1.2 mm. The mean standard deviation was 0.9 mm (range: 0.4–1.6 mm). The mean difference between ground truth and measured elemental compositions was 0.06 (range: 0.00 to 0.12 ). Significance. We verified the performance of our prompt gamma-ray spectroscopy detector prototype for proton range verification using tissue-mimicking and porcine samples. Measured proton ranges and elemental sample compositions were in good agreement with the ground truth. These measurements confirm the system’s reliability for a variety of tissues and bridge the gap between previously-reported experiments and ongoing in vivo patient measurements.

Participatory action imaging in increasing clinician confidence of trigger point dry needling or trigger point injection of muscles requiring use of intercostal blocking: A case report

Some clinicians privately report a lack of confidence of being able to safely perform trigger point dry needling (TrPDN) or trigger point injection (TrPI) on muscles that require intercostal blocking, presumably resulting in procedural underutilization. Participatory action imaging (PAI), combined with procedural training and literature review of adverse event incidence, can be a useful tool in enhancing clinician confidence. A 6’ 2”, 185 lbs, 53-year-old male clinician subject with a latissimus dorsi trigger point (TrP) and privately reported high anxiety of performing TrPDN using the intercostal blocking technique, was examined to determine how clinician subject confidence could be improved through PAI and education regarding the degree of coverage of the intercostal space and rib during intercostal blocking. The clinician subject was placed in a left lateral decubitus position and rib imaged with a Siemens Acuson S2000 Ultrasound system with an 18L6 16 Hz high definition linear probe without intercostal blocking, and subsequently with intercostal blocking during maximal exhalation and inhalation, respectively. During intercostal blocking with maximal exhalation, the pleural space and an additional 13% of each side of the rib’s superior and inferior borders were completely blocked by the examiner’s fingers. During intercostal blocking with maximal inhalation, the pleural space and an additional 3% of each side of the rib’s superior and inferior borders were completely blocked by the fingers of the clinician. On visual inspection and interpretation of the images (e.g., PAI), the clinician subject reported a “significant” decrease in self-reported anxiety in performing the intercostal blocking technique (STAI Y-1 score improvement to “low to no anxiety range”). While more study is needed to investigate how variation in patient anatomy (weight, height, and morphology), clinician anatomy (hand size), needle direction, and specific rib being used for influences patient safety, this case report presents PAI as a previously undescribed means for future research and clinician education regarding risk assessment of TrPDN or TrPI of muscles requiring intercostal blocking.

Total knee arthroplasty: implant selection and surgical considerations

Choosing a total knee arthroplasty implant can be confusing, with implant manufacturers all providing a wide range of implants even within the same brand. These may vary by design philosophy, fixation, constraint, materials and construct. A surgeon will need to have considered the evidence before making a choice of implant, and this evidence can be in the form of large cohorts of registry-based data, that may lack detail, or small studies that may be too small to be valid or have wider application. This article reviews the sources of data available to surgeons and discusses the categories of implant available to the surgeon. Variations in patient anatomy and pathology may require a surgeon to be familiar with a range of implant options. Surgeons should be aware of how their chosen implant performs, not just in studies from the designer, but in independent series and on national registries, and should know the UK Orthopaedic Data Evaluation Panel (ODEP) rating of the specific detailed construct that they are regularly using.

Zygoma Anatomy-Guided Approach "Scarf Graft" for Prevention of Soft Tissue Dehiscence Around Zygomatic Implants: Technical Note.

Different surgical approaches including the slot and the extrasinus techniques have been described to overcome disadvantages of the original Brånemark technique for the placement of zygomatic implants. A new concern associated with zygomatic implants placed externally to the maxillary wall is the possibility of disturbing buccal soft tissues, ending up with a dehiscence and a potential infective problem. Recently, a new methodology known as the Zygoma Anatomy-Guided Approach (ZAGA) has been described based on the concept of delivering specific therapy for each patient. ZAGA involves a variety of possibilities of implant trajectory from the intrasinus to an eventual extrasinus passage according to variations in patient anatomy. ZAGA methodology includes a rationale of how to prevent most of the reported complications of zygomatic implants. The objective of this technical note is to introduce the "Scarf Graft" as a part of the ZAGA protocol intended to prevent soft tissue dehiscence around extramaxillary zygomatic implants. A pediculated connective tissue graft is placed around the neck of the extramaxillary zygomatic implants. The increased connective tissue thickness consistently gives stable gingival tissue for prevention of recession. Currently, the treatment of soft tissue dehiscence around zygomatic implants does not have predictable results. Protocols for its prevention, such as the proposed ZAGA Scarf Graft, should be incorporated if an eventual dehiscence is foreseen.

Computational framework for population-based evaluation of TKR-implanted patellofemoral joint mechanics.

Differences in patient anatomy are known to influence joint mechanics. Accordingly, intersubject anatomical variation is an important consideration when assessing the design of joint replacement implants. The objective of this study was to develop a computational workflow to perform population-based evaluations of total knee replacement implant mechanics considering variation in patient anatomy and to assess the potential for an efficient sampling strategy to support design phase screening analyses. The approach generated virtual subject anatomies using a statistical shape model of the knee and performed virtual implantation to size and align the implants. A finite-element analysis simulated a deep knee bend activity and predicted patellofemoral (PF) mechanics. The study predicted bounds of performance for kinematics and contact mechanics and investigated relationships between patient factors and outputs. For example, the patella was less flexed throughout the deep knee bend activity for patients with an alta patellar alignment. The results also showed the PF range of motions in AP and ML were generally larger with increasing femoral component size. Comparison of the 10-90% bounds between sampling strategies agreed reasonably, suggesting that Latin Hypercube sampling can be used for initial screening evaluations and followed up by more intensive Monte Carlo simulation for refined designs. The platform demonstrated a functional workflow to consider variation in joint anatomy to support robust implant design.

A simple algorithm to predict non-compliance with organ at risk dose-volume constraints when planning intensity modulated post-prostatectomy radiation treatment: 'Why we should put the CART before the horse'.

It is not always apparent when the optimal IMRT/VMAT plan for post-prostatectomy radiotherapy (PPRT) has been achieved. Individual variation in patient anatomy is a key contributor. This study aimed to create a model to determine the probability of rectum and/or bladder doses exceeding planning goals based on individual patient anatomy prior to planning. The IMRT/VMAT PPRT plans from 200 men were randomly and evenly allocated into the Training Cohort and the Validation Cohort. Univariate and multivariate analysis of the Training Cohort identified variables which impacted bladder and rectal doses. Significant variables were included in a Classification and Regression Tree (CART) analysis. The resultant algorithm was then applied to the Validation Cohort. On multivariate analysis, prescription dose; bladder and rectal volume; lymph node treatment; and percentage of bladder and rectal overlap with the PTV were significant variables. Following CART analysis, the overlap volume (OV) for both rectum (rectum overlap>20%) and bladder (bladder overlap>20%) were the key drivers of meeting planning goals. Treatment of pelvic lymph nodes was included as the secondary driving factor for bladder (but not rectal) dose. On application to the Validation Cohort, CART analysis predicted 95% and 87% of patients who would meet bladder and rectal planning goals respectively. A simple algorithm was developed to predict plan quality by using the OV of the bladder and rectum with the PTV. This algorithm may be used a priori to assess the planning process in the context of variable anatomy, and to optimise planning quality and efficiency.

Virtual surgical planning, flow simulation, and 3-dimensional electrospinning of patient-specific grafts to optimize Fontan hemodynamics

BackgroundDespite advances in the Fontan procedure, there is an unmet clinical need for patient-specific graft designs that are optimized for variations in patient anatomy. The objective of this study is to design and produce patient-specific Fontan geometries, with the goal of improving hepatic flow distribution (HFD) and reducing power loss (Ploss), and manufacturing these designs by electrospinning. MethodsCardiac magnetic resonance imaging data from patients who previously underwent a Fontan procedure (n = 2) was used to create 3-dimensional models of their native Fontan geometry using standard image segmentation and geometry reconstruction software. For each patient, alternative designs were explored in silico, including tube-shaped and bifurcated conduits, and their performance in terms of Ploss and HFD probed by computational fluid dynamic (CFD) simulations. The best-performing options were then fabricated using electrospinning. ResultsCFD simulations showed that the bifurcated conduit improved HFD between the left and right pulmonary arteries, whereas both types of conduits reduced Ploss. In vitro testing with a flow-loop chamber supported the CFD results. The proposed designs were then successfully electrospun into tissue-engineered vascular grafts. ConclusionsOur unique virtual cardiac surgery approach has the potential to improve the quality of surgery by manufacturing patient-specific designs before surgery, that are also optimized with balanced HFD and minimal Ploss, based on refinement of commercially available options for image segmentation, computer-aided design, and flow simulations.

System events: readily accessible features for surgical phase detection.

Surgical phase recognition using sensor data is challenging due to high variation in patient anatomy and surgeon-specific operating styles. Segmenting surgical procedures into constituent phases is of significant utility for resident training, education, self-review, and context-aware operating room technologies. Phase annotation is a highly labor-intensive task and would benefit greatly from automated solutions. We propose a novel approach using system events-for example, activation of cautery tools-that are easily captured in most surgical procedures. Our method involves extracting event-based features over 90-s intervals and assigning a phase label to each interval. We explore three classification techniques: support vector machines, random forests, and temporal convolution neural networks. Each of these models independently predicts a label for each time interval. We also examine segmental inference using an approach based on the semi-Markov conditional random field, which jointly performs phase segmentation and classification. Our method is evaluated on a data set of 24 robot-assisted hysterectomy procedures. Our framework is able to detect surgical phases with an accuracy of 74% using event-based features over a set of five different phases-ligation, dissection, colpotomy, cuff closure, and background. Precision and recall values for the cuff closure (Precision: 83%, Recall: 98%) and dissection (Precision: 75%, Recall: 88%) classes were higher than other classes. The normalized Levenshtein distance between predicted and ground truth phase sequence was 25%. Our findings demonstrate that system events features are useful for automatically detecting surgical phase. Events contain phase information that cannot be obtained from motion data and that would require advanced computer vision algorithms to extract from a video. Many of these events are not specific to robotic surgery and can easily be recorded in non-robotic surgical modalities. In future work, we plan to combine information from system events, tool motion, and videos to automate phase detection in surgical procedures.

SU-E-J-72: Dosimetric Study of Cone-Beam CT-Based Radiation Treatment Planning Using a Patient-Specific Stepwise CT-Density Table

Purpose: To assess dose calculation accuracy of cone-beam CT (CBCT) based treatment plans using a patient-specific stepwise CT-density conversion table in comparison to conventional CT-based treatment plans. Methods: Unlike CT-based treatment planning which use fixed CT-density table, this study used patient-specific CT-density table to minimize the errors in reconstructed mass densities due to the effects of CBCT Hounsfield unit (HU) uncertainties. The patient-specific CT-density table was a stepwise function which maps HUs to only 6 classes of materials with different mass densities: air (0.00121g/cm3), lung (0.26g/cm3), adipose (0.95g/cm3), tissue (1.05 g/cm3), cartilage/bone (1.6g/cm3), and other (3g/cm3). HU thresholds to define different materials were adjusted for each CBCT via best match with the known tissue types in these images. Dose distributions were compared between CT-based plans and CBCT-based plans (IMRT/VMAT) for four types of treatment sites: head and neck (HN), lung, pancreas, and pelvis. For dosimetric comparison, PTV mean dose in both plans were compared. A gamma analysis was also performed to directly compare dosimetry in the two plans. Results: Compared to CT-based plans, the differences for PTV mean dose were 0.1% for pelvis, 1.1% for pancreas, 1.8% for lung, and −2.5% for HN in CBCT-based plans. The gamma passing ratemore » was 99.8% for pelvis, 99.6% for pancreas, and 99.3% for lung with 3%/3mm criteria, and 80.5% for head and neck with 5%/3mm criteria. Different dosimetry accuracy level was observed: 1% for pelvis, 3% for lung and pancreas, and 5% for head and neck. Conclusion: By converting CBCT data to 6 classes of materials for dose calculation, 3% of dose calculation accuracy can be achieved for anatomical sites studied here, except HN which had a 5% accuracy. CBCT-based treatment planning using a patient-specific stepwise CT-density table can facilitate the evaluation of dosimetry changes resulting from variation in patient anatomy.« less

Neurological Examination and Diagnostic Testing in Birds and Reptiles

Neurological dysfunction is a frequently presenting sign in avian and reptile patients. Clinical neurological signs are rarely pathognomonic, often requiring the attending veterinary surgeon to perform more involved diagnostic tests to achieve a diagnosis. Variations in patient anatomy, physiology and demeanour present unique challenges to the veterinarian evaluating the bird or reptile that has neurological disease signs. Despite these challenges, a well-structured physical and neurological examination can often be accomplished with minimal equipment. A structured neurological examination is essential to formulate an appropriate investigative plan and therapeutic regime for these difficult cases and to provide the owner with an accurate prognosis.

SU‐F‐BRD‐12: Prostate Cancer Dose Escalation Including Lymph Nodes: A Planning Study

Purpose:Dose escalation improves local control for prostate cancer but not overall survival (OS), only systemic therapy such as ADT can improve OS. One explanation is that occult diseases in the lymph nodes (LNs) are either not treated, or treated at inadequate level. This study explores dose escalation using IMRT with reduced margins.Methods:14 previously treated patients were selected. The CTV1 includes prostate and seminal vesicles, and CTV2 includes pelvic LNs. OARs include rectum, bladder, small bowel, femoral heads, and penile bulb. Reference plan (RP) was designed to follow RTOG‐0521, with 10 mm margin for PTVs. In phase I, both PTVs are prescribed to 1.8 Gy/fx for 26 fractions. In phase II, PTV1 is boosted for another 16 fractions. In dose escalation plans (DEPs), margins from 0 to 10 mm are added to form PTVs, which are prescribed to 1.8 Gy/fx for 42 fractions. 95% of PTVs are covered by prescription.Results:The CTV1/CTV2 has volume of 63±20 and 415±132 cc. To keep bladder Dmean same as in RP, margin must be ≤6 mm in DEPs. Similarly, margin needs to be ≤7 mm for rectum Dmean. For small bowel, even at margin of 0, Dmean and EUDs on DEPs are higher than RP, because RP was also IMRT. Instead, QUANTEC criteria of “45Gy volume&lt;195cc” was adopted. 6/14 patients fail this criteria in RP. For the remaining 8 patients, margins of ≤7(3) mm are necessary for 6(2) patients, respectively.Conclusion:To escalate LNs to full dose, planning margins need to be ≤6 mm to achieve same tolerance for rectum and bladder. A large variation in patient anatomy exists for small bowel and dose escalation was possible for half of patients with ≤6 mm margin. Investigations of whether and how ≤6 mm margin can be achieved through image guidance are needed.

Impact of Patient Anatomy Variations on a Fourier Technique for Markerless 4D-CBCT

Impact of Patient Anatomy Variations on a Fourier Technique for Markerless 4D-CBCT

Rotation Radius Dependence of 123I-FP-CIT and 123I-IBZM SPECT Uptake Ratios: A Monte Carlo Study

In dopamine brain SPECT, semiquantitative techniques are in use, mostly for research purposes, to calculate activity uptake in the striatum relative to the background. The measured uptake ratios depend on both acquisition and reconstruction, and one important parameter is the rotation radius of the γ-camera detectors, which affects spatial resolution. In brain SPECT research studies, the rotation radius is typically set to a constant value to maintain a constant resolution, but because of variations in patient anatomy and compliance, this is not always possible. In this study, correction factors as a function of rotation radius are developed to correct the uptake ratios where the rotation radius has deviated from the reference value, 15 cm. Monte Carlo simulations of a digital brain phantom were used to produce images with a high and a low uptake ratio, and for both studies the rotation radius was varied between 14 and 23 cm. Two different methods, one based on 2-dimensional (2D) regions of interest of constant shape and size, and one based on predefined 3-dimensional (3D) volumes of interest, were used to calculate the semiquantitative uptake ratios. For the 2D method, the change in uptake ratio was 1.2%/cm for the high uptake ratio and 0.9%/cm for the low uptake ratio. The corresponding results for the 3D method were 2.1% and 1.7%, respectively. The 3D method was found to be more dependent on rotation radius than the 2D method, which was expected because of the 3D nature of the partial-volume effect. The correction factors were, however, less dependent on which of the 2 uptake ratios was simulated, which is positive for the application of the correction equations on patient data.