Pretreatment Planning Research Articles

Pre-operative Assessment of Ablation Margins for Variable Blood Perfusion Metrics in a Magnetic Resonance Imaging Based Complex Breast Tumour Anatomy: Simulation Paradigms in Thermal Therapies

Background and objectivesImage-guided medical interventions facilitates precise visualization at treatment site. The conformal prediction for sparing healthy tissue fringes precisely in the vicinity of irregular tumour anatomy remains clinically challenging. Pre-clinical image-based computational modelling is imperative as it helps in enhancement of treatment quality, augmenting clinical-decision making, while planning, targeting, controlling, monitoring and assessing treatment response with an effective risk assessment before the onset of treatment in clinical settings. In this study, the influence of heat deposition rate (SAR), exposure duration, and variable blood perfusion metrics for a patient-specific breast tumour is quantified considering the tumour margins thereby suggesting need of geometrically accurate models. MethodsA three-dimensional realistic model mimicking dimensions of a female breast, comprising ~1.7 cm irregular tumour, was generated from patient specific two-dimensional DICOM format MRI images through image segmentation tools MIMICS 19.0® and 3-Matic 11.0® which is finally exported to COMSOL Multiphysics 5.2® as a volumetric mesh for finite element analysis. The Pennes bioheat transfer model and Arrhenius thermal damage model of cell-death are integrated to simulate a coupled biophysics problem. A comparative blood perfusion analysis is done to evaluate the response of tumour during heating considering thermal damage extent, including the tumour margins while sparing critical adjoining healthy tissues. ResultsThe evaluated thermal damage zones for 1 mm, 2 mm and 3 mm fringe heating region (beyond tumour boundary) reveals 0.09%, 0.21% and 0.34% thermal damage to the healthy tissue (which is <1%) and thus successful necrosis of the tumour. The iterative computational experiments suggests treatment margins < 5 mm are sufficient enough as heating beyond 3 mm fringe layer leads to higher damage surrounding the tumour approximately 1.5 times the tumour volume. Further, the heat-dosage requirements are 22% more for highly perfused tumour as compared to moderately perfused tumour with an approximate double time to ablate the whole tumour volume. ConclusionsDepending on the blood perfusion characteristics of a tumour, it is a trade-off between heat-dosage (SAR) and exposure/treatment duration to get desired thermal damage including the irregular tumour boundaries while taking into account, the margin of healthy tissue. The suggested patient-specific integrated multiphysics-model based on MRI-Images may be implemented for pre-treatment planning based on the tumour blood perfusion to evaluate the thermal ablation zone dimensions clinically and thereby avoiding the damage of off-target tissues. Thus, risks involving underestimation or overestimation of thermal coagulation zones may be minimised while preserving the surrounding normal breast parenchyma.

A practical pretreatment planning method of multiple puncturing for thermal ablation surgery

Thermal ablation surgery serves as one of the main approaches to treat liver tumors. The pretreatment planning, which highly demands the experience and ability of the physician, plays a vital role in thermal ablation surgery. The planning of multiple puncturing is necessary for avoiding the possible interference, destroying the tumor thoroughly and minimizing the damage to healthy tissue. A GPU-independent pretreatment planning method is proposed based on multi-objective optimization, which takes the most comprehensive constraints into consideration. An adaptive decision method of closing kernel size based on Jenks Natural Breaks is utilized to describe the final feasible region more accurately. It should be noted that the reasonable procedure of solving the feasible region and the use of KD tree based high dimensional search approach are used to enhance the computational efficiency. Seven constraints are handled within 7 s without GPU acceleration. The Pareto front points of nine puncturing tests are obtained in 5 s by using the NSGA-II algorithm. To evaluate the maximum difference and similarity between the planning results and the puncturing points recommended by the physician, Hausdorff distance and overlap rate are respectively developed, the Hausdorff distances are within 30 mm in seven out of nine tests and the average value of overlap rate is 73.0% for all the tests. The puncturing paths of high safety and clinical-practice compliance can be provided by the proposed method, based on which the pretreatment planning software developed can apply to the interns’ training and ability evaluating for thermal ablation surgery.

Impact of hydrogel peri-rectal spacer insertion on prostate gland intra-fraction motion during 1.5\u2009T MR-guided stereotactic body radiotherapy

BackgroundThe assessment of organ motion is a crucial feature for prostate stereotactic body radiotherapy (SBRT). Rectal spacer may represent a helpful device in order to outdistance rectal wall from clinical target, but its impact on organ motion is still a matter of debate. MRI-Linac is a new frontier in radiation oncology as it allows a superior visualization of the real-time anatomy of the patient and the current highest level of adaptive radiotherapy.MethodsWe present data regarding a total of 100 fractions in 20 patients who underwent MRI-guided prostate SBRT for low-to-intermediate risk prostate cancer with or without spacer. Translational and rotational shifts were computed on the pre- and post-treatment MRI acquisitions referring to the delivery position for antero-posterior, latero-lateral and cranio-caudal direction, and assessed using the Mann-Whitney U-Test.ResultsAll patients were treated with a five sessions schedule (35 Gy/5fx) using MRI-Linac for a median fraction treatment time of 50 min (range, 46–65). In the entire study sample, median rotational displacement was 0.1° in cranio-caudal, − 0.002° in latero-lateral and 0.01° in antero-posterior direction; median translational shift was 0.11 mm in cranio-caudal, − 0.24 mm in latero-lateral and − 0.22 mm in antero-posterior. A significant difference between spacer and no-spacer patients in terms of rotational shifts in the antero-posterior direction (p = 0.033) was observed; also for translational shifts a positive trend was detected in antero-posterior direction (p = 0.07), although with no statistical significance. We observed statistically significant differences in the pre-treatment planning phase in favor of the spacer cohort for several rectum dose constraints: rectum V32Gy < 5% (p = 0.001), V28 Gy < 10% (p = 0.001) and V18Gy < 35% (p = 0.039). Also for bladder V35 Gy < 1 cc, the use of spacer provided a dosimetric advantage compared to the no-spacer subpopulation (p = 0.04). Furthermore, PTV V33.2Gy > 95% was higher in the spacer cohort compared to the no-spacer one (p = 0.036).ConclusionIn our experience, the application of rectal hydrogel spacer for prostate SBRT resulted in a significant impact on rotational antero-posterior shifts contributing to limit prostate intra-fraction motion. Further studies with larger sample size and longer follow-up are required to confirm this ideally favorable effect and to assess any potential impact on clinical outcomes.

Experimental assessment of microwave ablation computational modeling with MR thermometry.

Computational models are widely used during the design and characterization of microwave ablation (MWA) devices, and have been proposed for pretreatment planning. Our objective was to assess three-dimensional (3D) transient temperature and ablation profiles predicted by MWA computational models with temperature profiles measured experimentally using magnetic resonance (MR) thermometry in ex vivo bovine liver. We performed MWA in ex vivo tissue under MR guidance using a custom, 2.45GHz water-cooled applicator. MR thermometry data were acquired for 2min prior to heating, during 5-10min microwave exposures, and for 3min following heating. Fiber-optic temperature sensors were used to validate the accuracy of MR temperature measurements. A total of 13 ablation experiments were conducted using 30-50W applied power at the applicator input. MWA computational models were implemented using the finite element method, and incorporated temperature-dependent changes in tissue physical properties. Model-predicted ablation zone extents were compared against MRI-derived Arrhenius thermal damage maps using the Dice similarity coefficient (DSC). Prior to heating, the observed standard deviation of MR temperature data was in the range of 0.3-0.7°C. Mean absolute error between MR temperature measurements and fiber-optic temperature probes during heating was in the range of 0.5-2.8°C. The mean DSC between model-predicted ablation zones and MRI-derived Arrhenius thermal damage maps for 13 experimental set-ups was 0.95. When comparing simulated and experimentally (i.e. using MRI) measured temperatures, the mean absolute error (MAE %) relative to maximum temperature change was in the range 5%-8.5%. We developed a system for characterizing 3D transient temperature and ablation profiles with MR thermometry during MWA in ex vivo liver tissue, and applied the system for experimental validation of MWA computational models.

Numerical Sensitivity Analysis for Dielectric Characterization of Biological Samples by Open-Ended Probe Technique.

Dielectric characterization of biological tissues has become a fundamental aspect of the design of medical treatments based on electromagnetic energy delivery and their pre-treatment planning. Among several measuring techniques proposed in the literature, broadband and minimally-invasive open-ended probe measurements are best-suited for biological tissues. However, several challenges related to measurement accuracy arise when dealing with biological tissues in both ex vivo and in vivo scenarios such as very constrained set-ups in terms of limited sample size and probe positioning. By means of the Finite Integration Technique in the CST Studio Suite® software, the numerical accuracy of the reconstruction of the complex permittivity of a high water-content tissue such as liver and a low water-content tissue such as fat is evaluated for different sample dimensions, different location of the probe, and considering the influence of the background environment. It is found that for high water-content tissues, the insertion depth of the probe into the sample is the most critical parameter on the accuracy of the reconstruction. Whereas when low water-content tissues are measured, the probe could be simply placed in contact with the surface of the sample but a deeper and wider sample is required to mitigate biasing effects from the background environment. The numerical analysis proves to be a valid tool to assess the suitability of a measurement set-up for a target accuracy threshold.

Imaging modalities in the assessment of presacral recurrent rectal cancer

Imaging plays a key role in the diagnosis and decision-making process including pre-treatment planning, surgical strategy, and follow-up. The critical point in diagnosis of presacral recurrent rectal cancer by imaging modalities is to distinguish the recurrent tumor from nonmalignant tissues induced by operation or radiotherapy. The practice guideline recommends CT as surveillance imaging modality for recurrent rectal cancer. MRI shows higher accuracy, sensitivity, and specificity in diagnosis of presacral recurrent rectal cancer compared with CT. If CT or MRI can not make final diagnosis in challenging cases, 18-fluorodeoxyglucose positron emission tomography ((18)FDG PET) is recommended to aid diagnosis with high sensitivity and specificity, though false-positivity and negativity should be considered. If new or enlarging soft tissue are shown in the follow-up examination, tumor recurrence should be suspected. In addition, tumor-related high risk factors, treatment protocol, surgery, quality of specimen and pathological stages should also be considered when presacral recurrent rectal cancer is to be diagnosed.

A simple endoscopic method with radial endobronchial ultrasonography for low-migration rate coil-tailed fiducial marker placement.

BackgroundFiducial markers (FMs) are useful for tracking small peripheral lung nodules (PLN) before stereotactic radiotherapy, but migration over the course of treatment may result in inaccurate dosing to the tumor. To minimize FM migration, coil-tailed FMs have been designed. Our objective was to assess both the feasibility of radial endobronchial ultrasonography (r-EBUS) placement and the migration rate of coil-tailed FMs.MethodsIn this retrospective study, we included patients who received r-EBUS guided placement of coil-tailed FMs for PLN <25 mm from June 2015 to May 2018. We introduced the FM into the nodule with the use of bronchial brush, without fluoroscopy.ResultsThirty patients had r-EBUS guided placement of a coil-tailed FM before stereotactic radiation therapy. Nodule’s median long- and short-axis diameters were 15 mm (8–25 mm) and 8 mm (5–20 mm), respectively; short diameter of 27 nodules (90%) was less than 15 mm. All nodules were reached and visualized with r-EBUS, with an ultrasound (US) signal showing a centered or tangential probe in 26 and 4 cases, respectively. No immediate complication was reported. Twenty-three patients had stereotactic radiation therapy within a median time of 29 days (14–126 days). No FM migration occurred between r-EBUS placement and radiotherapy. Pre-treatment planning and 3-month follow-up CT scans showed that all FMs stayed in direct contact with the lesions.Conclusionsr-EBUS is a safe procedure for the placement of nitinol coil FMs, which have a low migration rate.

Accuracy evaluation of computer-guided implant surgery associated with prototyped surgical guides

Statement of problemCorrect implant placement by means of a pretreatment planning software program is still subject to deviations between the planned and achieved implant positions. Inaccuracy at this level may have drastic consequences, including neurovascular trauma. Further data are therefore needed to evaluate the accuracy of such computer-guided implant planning software programs. PurposeThe purpose of this clinical study was to evaluate the accuracy of computer-guided implant surgery associated with prototyped surgical guides. Material and methodsCone beam computed tomography (CBCT) scans were made of the participants with a tomographic guide to merge anatomic and prosthetic data. This allowed virtual planning with a prosthetically guided approach respecting the anatomy of the participant. A prototyped surgical guide was then fabricated from the virtual plan, determining the intrasurgical position of the implants. Flapless guided implant surgery was carried out according to the manufacturer’s instructions. A second CBCT was made 30 days after the surgery, to enable overlapping of the data from before and after the implant placement. The angular, coronal, central, and apical deviations of the placed implants were measured and compared with those virtually planned. The data were submitted to descriptive statistic and intraclass correlation coefficient (ICC), analysis of variance, and the Student t test (α=.05). ResultsA total of 61 implants were analyzed. The mean angular deviation was 2.04 degrees. The mean coronal, central, and apical linear deviations were 0.68 mm, 0.72 mm, and 0.82 mm, respectively. No statistically significant difference was found between the virtual and the real position of the implants inserted. A tendency toward a greater absolute error was observed in the mandible than in the maxilla in terms of angular (P=.047), central (P=.043), and coronal error (P=.031). ConclusionsFlapless computer-guided implant surgery with virtual planning had some angular and linear deviations; nevertheless, this technique should be acceptably safe and accurate.

Fertility Preservation Discussion in Adult Male Cancer Patients Is Underutilized and Varies Based on Age

Introduction. Oncofertility is often omitted in pretreatment discussion and planning. This study seeks to evaluate the prevalence and influencing factors of cryopreservation discussions. Methods. A retrospective review was performed on 3,133 male patients aged 18–60 years with a cancer diagnosis at a single institution. Data was collected for cryopreservation discussion and if cryopreservation of sperm was performed. Age, race, cancer location, primary treatment, and chemotherapy status were also recorded. Results. A total of 2,504 patients were included for analysis. There was documentation of counseling in 14.1% and 5.0% underwent cryopreservation. Logistic regression indicated a significant effect of age, race, site of cancer, primary treatment, and chemotherapy treatment on whether cryopreservation was discussed (χ2 P Conclusion. Our study indicates that cryopreservation is vastly under-discussed, especially in older patients and those who do not receive chemotherapy. It is important to include preservation options and counseling for all patients.

The role of imaging in head and neck cancer: An overview of different imaging modalities in primary diagnosis and staging of the disease.

Radiology has an essential role in diagnosis, staging, and management in all subspecialties related to oncology. It has a broad utility from its use as an initial screening tool for cancer detection, followed by staging and surveillance of disease as well as the delivery of appropriate treatment regimens.In order to plan a treatment, the use of different modalities of radiological imaging are a key factor for pre-treatment planning and staging of the extent of disease in accordance with tumor-node-metastasis (TNM) system for highly conformal treatments, such as brachytherapy (BT).In this work, we present an overview of main modalities of imaging and principles of their application especially in head and neck cancer, with its diagnosis, treatment, follow-up with post-treatment changes, and overall management.

Singlet oxygen model evaluation of interstitial photodynamic therapy with 5-aminolevulinic acid for malignant brain tumor.

.Interstitial photodynamic therapy (iPDT) with 5-aminolevulinic acid (ALA) is a possible alternative treatment for malignant brain tumors. Further evaluation is, however, required before it can be clinically applied. Computational simulation of the photophysical process in ALA-iPDT can offer a quantitative tool for understanding treatment outcomes, which depend on various variables related to clinical treatment conditions. We propose a clinical simulation method of ALA-iPDT for malignant brain tumors using a singlet oxygen () model and threshold to induce cell death. In this method, the amount of generated is calculated using a photosensitizer photobleaching coefficient and quantum yield, which have been measured in several previous studies. Results of the simulation using clinical magnetic resonance imaging data show the need to specify the insertion positions of cylindrical light diffusers and the level of light fluence. Detailed analysis with a numerical brain tumor model demonstrates that ALA-iPDT treatment outcomes depend on combinations of photobleaching and threshold values. These results indicate that individual medical procedures, including pretreatment planning and treatment monitoring, will greatly benefit from simulation of ALA-iPDT outcomes.

An image-based computational modeling approach for prediction of temperature distribution during photothermal therapy

Nanoparticle-assisted photothermal therapy (NPTT) has recently renewed the interest of using hyperthermia in cancer therapy due to selective heating of tumor by utilizing light-responsive nanoparticles such as gold nanoparticles (AuNPs). Pre-treatment planning of NPTT can help to predict temperature distribution within the body in order to optimize the treatment parameters before the actual heating operation. The use of actual tumor geometry and nanoparticle distribution are key requirements for accurate prediction of temperature distribution during numerical calculations of the heat transfer process. This study attempts to develop a numerical modeling strategy for NPTT based on computed tomography (CT) imaging. To this end, CT26 colon tumor-bearing mice were injected with alginate-coated AuNPs (Au@Alg) and then underwent CT imaging. The tumor geometry and nanoparticle distribution map were obtained directly from CT image of the tumor and exported into a finite element simulation software for subsequent heat transfer modeling. The predicted temperature of the tumor from numerical modeling was found to be in reasonable agreement with the measured data from in vivo thermometry. This model has the potential to be used as a pre-treatment planning tool to design an individualized heating protocol for various tumor geometry before the actual heating treatment.

Fertility preservation discussion in cancer patients is underutilized and varies based on age

The American Society of Clinical Oncology recommends that all patients with a cancer diagnosis be counseled on the impact of their disease and treatment on fertility. Despite this, oncofertility is often omitted in pretreatment discussion and planning. The reason for lack of adequate counselling and referrals for fertility preservation is unclear. This study seeks to evaluate the prevalence of cryopreservation discussions. A retrospective review was performed on 3133 male patients aged 18-60 years with a cancer diagnosis at a single institution. Patient’s charts were queried for “vasectomy”, “semen”, “sperm”, “fertility” and “preservation”. Patients were excluded if they had a vasectomy or sperm banking prior to their cancer diagnosis, as well as if they had no treatment for their cancer diagnosis. Data was collected for cryopreservation discussion, discussion prior to treatment, discussion before chemotherapy, and if cryopreservation of sperm was performed. Age, race, cancer location, primary treatment, and chemotherapy status was also recorded. A total of 2504 patients were included for analysis. Mean age was 49 years. There was documentation of counseling on cryopreservation in 353 (14.1%)—280 (79.3%) before primary treatment and 298 (84.4%) before chemotherapy. 126 (5.0%) patients underwent cryopreservation. A logistic regression indicated a significant effect of age, race, site of cancer, primary treatment, and chemotherapy treatment on whether cryopreservation was discussed (chi2 <0.001, pseudo R2=0.29). Chemotherapy at any time of treatment (OR 6.99, p<0.001) was significantly associated with cryopreservation counseling. Testicular and prostate cancer patients were significantly more likely to be offered cryopreservation (p<0.001 and p=0.005, respectively). Patients aged 30-39, 40-49, and 50-60 were significantly less likely to receive counseling when compared to patients aged 18-29 while controlling for other variables (OR 0.41, 0.12 and 0.05 respectively, p<0.001 for all 3 groups). Reproductive side effects are not as commonly discussed as other systemic side effects when a patient receives a cancer diagnosis or when they start treatment. Our study indicates that cryopreservation is vastly underdiscussed. Younger patients, those undergoing chemotherapy during their treatment period, and a diagnosis of testicular and prostate cancer were more likely to receive cryopreservation counseling. As assisted reproductive techniques have become more successful and readily available, it is important to include options and counseling for all patients.

Spinal metastasis: diagnosis, management and follow-up.

Spine metastatic disease is an increasingly common occurrence in cancer patients due to improved patient survival. Close proximity of the bony spinal column to the spinal cord limits many conventional treatments for metastatic disease. In the past decade, we have witnessed dramatic advancements in therapies, with improvements in surgical techniques and recent adoption of spine stereotactic radiotherapy techniques leading to improved patient outcomes. Multidisciplinary approaches to patient evaluation, treatment and follow-up are essential. Imaging plays an ever increasing role in disease detection, pre-treatment planning and assessment of patient outcomes. It is important for the radiologist to be familiar with imaging algorithms, best practices for surgery and/or radiotherapy and imaging findings in the post-treatment period that may indicate disease recurrence. In this review, we present a multidisciplinary discussion of spine metastases, with specific focus on pre-treatment imaging, planning, current treatment approaches, and post-treatment assessment.

2018 FIGO Staging System for Uterine Cervical Cancer: Enter Cross-sectional Imaging.

Imaging plays a central role in the 2018 International Federation of Gynecology and Obstetrics staging system for uterine cervical cancer. The revision calls for a more precise measurement of primary tumor size, best assessed with imaging. Evaluation for abdominopelvic retroperitoneal lymphadenopathy, either with imaging alone or with pathologic analysis, is now also part of staging. Choice of modality depends on the technology available within the practice setting. In high-resource settings, pelvic MRI (to assess tumor size and central pelvic spread) and torso fluorodeoxyglucose PET/CT (to assess lymphadenopathy and distant metastases) are used to assign stage and to plan therapy. In lower-resource settings, analogous modalities are pelvic US and chest radiography. Although imaging is already a part of pretreatment planning in some high-resource settings, its incorporation into assigning stage is a new development.

Cinematic rendering of focal liver masses

Cinematic rendering (CR) is a recently described three-dimensional (3D) rendering technique that generates photorealistic images based on a new lighting model. This review illustrates the potential application of CR in the evaluation of focal liver masses. CR shows promise in improving the visualization of enhancement pattern and internal architecture, local tumor extension, and global disease burden, which may be helpful in focal liver mass characterization and pretreatment planning.

Feasibility of ultrashort echo time images using full-wave acoustic and thermal modeling for transcranial MRI-guided focused ultrasound (tcMRgFUS) planning

Transcranial MRI-guided focused ultrasound (tcMRgFUS) is increasingly used to non-invasively treat a wide variety of neurological disorders including essential tremors, Parkinson’s disease, and neuropathic pain. Although this treatment is an MRI-guided procedure, the current pre-treatment screening and planning involve a CT of the head to obtain 3D skull images. These images are necessary for estimating the proportion of absorbed energy and the acoustic phase shift associated with the skull and determining the transmit energy of ultrasonic waves to create thermal lesions at a desired focal spot. Ultrashort echo time (UTE) MR sequences can capture signals from tissues such as bone which have a very short transverse relaxation time. In this manuscript, we assess the use of a UTE based sequence to image the calvarium and test the feasibility of obviating the need for CT based imaging during an MR-guided focused ultrasound therapy. We demonstrate that the segmentation of bone using UTE images leads to similar skull density ratio values as determined from CT with high correlation (r = 0.88; p < 0.0001). Furthermore, through treatment specific modeling we demonstrate that the thermal profiles and focal locations are in concordance with the actual treatment plan when using the UTE based skull intensity information suggesting the possibility of replacing the CT scans with UTE based skull imaging in all tcMRgFUS procedures, potentially eliminating unnecessary radiation exposure. Overall, the results indicate that UTE MR imaging may serve as an effective and accurate alternative to CT imaging for both screening and pre-treatment planning on patients undergoing the tcMRgFUS procedure.

Imaging Appearance and Nonsurgical Management of Pelvic Venous Congestion Syndrome.

Pelvic venous congestion syndrome (PVCS) is a challenging and complex cause of chronic pelvic pain in female patients. PVCS due to incompetent vein valves is the combination of gonadal vein reflux and pelvic venous engorgement in patients with chronic pelvic pain without other causes. However, pelvic venous engorgement and gonadal vein reflux can be seen in patients without pelvic pain, which makes obtaining a detailed history and physical examination important for workup and diagnosis. The underlying cause of PVCS may be incompetent gonadal vein valves or structural causes such as left renal vein compression with an incompetent gonadal vein valve (nutcracker syndrome) or iliac vein compression (May-Thurner configuration) with reflux into the ipsilateral internal iliac vein. Venography is considered the criterion standard for imaging diagnosis; however, more recently, US and MRI have been shown to provide adequate accuracy for diagnosis. Noninvasive imaging studies aid in the diagnosis of PVCS and also aid in pretreatment planning. When PVCS is caused by incompetent gonadal vein valves, treatment typically is performed by means of embolization via a minimally invasive catheter with excellent technical and clinical success rates. When PVCS is caused by venous obstruction, the obstruction must be treated first before gonadal vein embolization and sclerotherapy are considered. ©RSNA, 2019 Online supplemental material is available for this article.

Convolutional Neural Networks for the Detection and Measurement of Cerebral Aneurysms on Magnetic Resonance Angiography.

Aneurysm size correlates with rupture risk and is important for treatment planning. User annotation of aneurysm size is slow and tedious, particularly for large data sets. Geometric shortcuts to compute size have been shown to be inaccurate, particularly for nonstandard aneurysm geometries. To develop and train a convolutional neural network (CNN) to detect and measure cerebral aneurysms from magnetic resonance angiography (MRA) automatically and without geometric shortcuts. In step 1, a CNN based on the U-net architecture was trained on 250 MRA maximum intensity projection (MIP) images, then applied to a testing set. In step 2, the trained CNN was applied to a separate set of 14 basilar tip aneurysms for size prediction. Step 1-the CNN successfully identified aneurysms in 85/86 (98.8% of) testing set cases, with a receiver operating characteristic (ROC) area-under-the-curve of 0.87. Step 2-automated basilar tip aneurysm linear size differed from radiologist-traced aneurysm size on average by 2.01mm, or 30%. The CNN aneurysm area differed from radiologist-derived area on average by 8.1mm2 or 27%. CNN correctly predicted the area trend for the set of aneurysms. This approach is to our knowledge the first using CNNs to derive aneurysm size. In particular, we demonstrate the clinically pertinent application of computing maximal aneurysm one-dimensional size and two-dimensional area. We propose that future work can apply this to facilitate pre-treatment planning and possibly identify previously missed aneurysms in retrospective assessment.

Optimizing interstitial photodynamic therapy with custom cylindrical diffusers.

Interstitial photodynamic therapy (iPDT) has shown promise recently as a minimally invasive cancer treatment, partially due to the development of non-toxic photosensitizers in the absence of activation light. However, a major challenge in iPDT is the pre-treatment planning process that specifies the number of diffusers needed, along with their positions and allocated powers, to confine the light distribution to the target volume as much as possible. In this work, a new power allocation algorithm for cylindrical light diffusers including those that can produce customized longitudinal (tailored) emission profiles is introduced. The proposed formulation is convex to guarantee the minimum over-dose possible on the surrounding organs-at-risk. The impact of varying the diffuser lengths and penetration angles on the quality of the plan is evaluated. The results of this study are demonstrated for different photosensitizers activated at different wavelengths and simulated on virtual tumors modeling virtual glioblastoma multiforme cases. Results show that manufacturable cylindrical diffusers with tailored emission profiles can significantly outperform those with conventional flat profiles with an average damage reduction on white matter of 15% to 55% and on gray matter of 23% to 58%.