Hilary Bagshaw

All Publications

• Deep learning applications in automatic needle segmentation in ultrasound-guided prostate brachytherapy. Medical physics Wang, F., Xing, L., Bagshaw, H., Buyyounouski, M., Han, B. 2020

  Abstract

  PURPOSE: High-Dose-Rate (HDR) brachytherapy is one of the most effective ways to treat the prostate cancer, which is the second most common cancer in men worldwide. This treatment delivers highly conformal dose through the transperineal needle implants and is guided by a real time ultrasound (US) imaging system. Currently, the brachytherapy needles in the US images are manually segmented by physicists during the treatment, which is time-consuming and error-prone. In this study, we propose a set of deep learning based algorithms to accurately segment the brachytherapy needles and locate the needle tips from the US images.METHODS: Two deep neural networks are developed to address this problem. First, a modified deep U-Net is used to segment the pixels belonging to the brachytherapy needles from the US images. Second, an additional VGG-16 based deep convolutional network is combined with the segmentation network to predict the locations of the needle tips. The networks are trained and evaluated on a clinical US images dataset with labeled needle trajectories collected in our hospital (Institutional Review Board approval (IRB 41755)).RESULTS: The evaluation results show that our method can accurately extract the trajectories of the needles with a resolution of 0.668 mm and 0.319 mm in x and y direction respectively. 95.4% of the x direction and 99.2% of the y direction have error ≤ 2 mm. Moreover, The position resolutions of the tips are 0.721 mm, 0.369 mm and 1.877 mm in x, y and z directions respectively, while 94.2%, 98.3% and 67.5% of the data have error ≤ 2 mm.CONCLUSIONS: This paper proposed a neural network based algorithm to segment the brachytherapy needles from the US images and locate the needle tip. It can be used in the HDR brachytherapy to help improve the efficiency and quality of the treatments.

  View details for DOI 10.1002/mp.14328

  View details for PubMedID 32542758

• A Deep Learning Framework for Prostate Localization in Cone Beam CT Guided Radiotherapy. Medical physics Liang, X., Zhao, W., Hristov, D. H., Buyyounouski, M. K., Hancock, S. L., Bagshaw, H., Zhang, Q., Xie, Y., Xing, L. 2020

  Abstract

  To develop a deep learning-based model for prostate planning target volume (PTV) localization on cone-beam CT (CBCT) to improve the workflow of CBCT-guided patient setup.A two-step task-based residual network (T2 RN) is proposed to automatically identify inherent landmarks in prostate PTV. The input to the T2 RN is the pre-treatment CBCT images of the patient, and the output is the deep learning-identified landmarks in the PTV. To ensure robust PTV localization, the T2 RN model is trained by using over thousand sets of CT images with labeled landmarks, each of the CTs corresponds to a different scenario of patient position and/or anatomy distribution generated by synthetically changing the planning CT (pCT) image. The changes, including translation, rotation, and deformation, represent vast possible clinical situations of anatomy variations during a course of radiation therapy (RT). The trained patient-specific T2 RN model is tested by using 240 CBCTs from six patients. The testing CBCTs consists of 120 original CBCTs and 120 synthetic CBCTs. The synthetic CBCTs are generated by applying rotation/translation transformations to each of the original CBCT.The systematic/random setup errors between the model prediction and the reference are found to be less than 0.25/2.46 mm and 0.14/1.41° in translation and rotation dimensions, respectively. Pearson's correlation coefficient between model prediction and the reference is higher than 0.94 in translation and rotation dimensions. The Bland-Altman plots show good agreement between the two techniques.A novel T2 RN deep learning technique is established to localize the prostate PTV for RT patient setup. Our results show that highly accurate marker-less prostate setup is achievable by leveraging the state-of-the-art deep learning strategy.

  View details for DOI 10.1002/mp.14355

  View details for PubMedID 32583418

• Timing of Adjuvant Radiation Therapy and Risk of Wound-Related Complications Among Patients With Spinal Metastatic Disease GLOBAL SPINE JOURNAL Azad, T. D., Varshneya, K., Herrick, D. B., Pendharkar, A., Ho, A. L., Stienen, M., Zygourakis, C., Bagshaw, H. P., Veeravagu, A., Ratliff, J. K., Desai, A. 2019

  View details for DOI 10.1177/2192568219889363

  View details for Web of Science ID 000498682300001

• Incorporating imaging information from deep neural network layers into image guided radiation therapy (IGRT). Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology Zhao, W., Han, B., Yang, Y., Buyyounouski, M., Hancock, S. L., Bagshaw, H., Xing, L. 2019; 140: 167–74

  Abstract

  BACKGROUND AND PURPOSE: To investigate a novel markerless prostate localization strategy using a pre-trained deep learning model to interpret routine projection kilovoltage (kV) X-ray images in image-guided radiation therapy (IGRT).MATERIALS AND METHODS: We developed a personalized region-based convolutional neural network to localize the prostate treatment target without implanted fiducials. To train the deep neural network (DNN), we used the patient's planning computed tomography (pCT) images with pre-delineated prostate target to generate a large amount of synthetic kV projection X-ray images in the geometry of onboard imager (OBI) system. The DNN model was evaluated by retrospectively studying 10 patients who underwent prostate IGRT. Three out of the ten patients who had implanted fiducials and the fiducials' positions in the OBI images acquired for treatment setup were examined to show the potential of the proposed method for prostate IGRT. Statistical analysis using Lin's concordance correlation coefficient was calculated to assess the results along with the difference between the digitally reconstructed radiographs (DRR) derived and DNN predicted locations of the prostate.RESULTS: Differences between the predicted target positions using DNN and their actual positions are (mean ± standard deviation) 1.58 ± 0.43 mm, 1.64 ± 0.43 mm, and 1.67 ± 0.36 mm in anterior-posterior, lateral, and oblique directions, respectively. Prostate position identified on the OBI kV images is also found to be consistent with that derived from the implanted fiducials.CONCLUSIONS: Highly accurate, markerless prostate localization based on deep learning is achievable. The proposed method is useful for daily patient positioning and real-time target tracking during prostate radiotherapy.

  View details for DOI 10.1016/j.radonc.2019.06.027

  View details for PubMedID 31302347

• Multi-institutional Analysis of Prostate-Specific Antigen Kinetics Following Stereotactic Body Radiotherapy (SBRT). International journal of radiation oncology, biology, physics Jiang, N. Y., Dang, A. T., Yuan, Y., Chu, F., Shabsovich, D., King, C. R., Collins, S. P., Aghdam, N., Suy, S., Mantz, C. A., Miszczyk, L., Napieralska, A., Namysl-Kaletka, A., Bagshaw, H., Prionas, N., Buyyounouski, M. K., Jackson, W. C., Spratt, D. E., Nickols, N. G., Steinberg, M. L., Kupelian, P. A., Kishan, A. U. 2019

  Abstract

  PURPOSE: Understanding prostate-specific antigen (PSA) kinetics after radiotherapy plays a large role in the management of prostate cancer (PCa) patients. This is particularly true regarding establishing expectations regarding PSA nadir (nPSA) and PSA bounces, which can be disconcerting. As increasingly more patients are being treated with stereotactic body radiotherapy (SBRT) for low- and intermediate-risk PCa, it is imperative to understand the PSA response to SBRT.METHODS&MATERIALS: PSA data from five institutions were retrospectively analyzed for localized PCa patients treated definitively with SBRT alone from 2004 to 2016. Patients received 35-40 Gy in five fractions per institutional standards. Patients who had less than 12 months of PSA data or received androgen deprivation therapy were excluded from this study. Linear and logistic multivariable analysis were performed to identify predictors of nPSA, bounce, and biochemical recurrence, and joint latent class models (JLCM) were developed to identify significant predictors of time to biochemical failure.RESULTS: 1062 patients were included in this study. Median follow-up was 66 months (interquartile range (IQR) 36.4 - 89.9 months). Biochemical failure per the Phoenix criteria occurred in 4% of patients. Median nPSA was 0.2 ng/ml, median time to nPSA was 40 months, 84% of patients had a nPSA <0.5 ng/ml, and 54% of patients had a nPSA <0.2 ng/ml. On multivariable analysis, nPSA was a significant predictor of biochemical failure. Benign PSA bounce was noted in 26% of patients. The median magnitude of PSA bounce was 0.52 ng/ml (IQR 0.3 - 1.0 ng/ml). Median time to PSA bounce was 18.1 months (IQR 12.0 - 31.1 months). On multivariable analysis, age and radiation dose were significantly associated with a lower incidence of bounce. JLCM modeling found that nPSA and radiation dose were significantly associated with longer time to biochemical failure.CONCLUSIONS: In this multi-institutional cohort of patients with long-term follow-up, we found that SBRT led to low nPSAs. In turn, lower nPSAs are associated with reduced incidence of, and longer time to, biochemical failure. Benign PSA bounces occurred in a quarter of patients, as late as several years after treatment. Further studies are needed to directly compare the PSA response of patients who receive SBRT versus other treatment modalities.

  View details for DOI 10.1016/j.ijrobp.2019.06.2539

  View details for PubMedID 31276777

• Evaluating Prostate-Specific Antigen (PSA) Nadir and Bounce After Stereotactic Body Radiotherapy (SBRT) in a Multi-Institutional Cohort Jiang, N., Dang, A., Yuan, Y., Chu, F., King, C., Collins, S., Aghdam, N., Suy, S., Miszczyk, L., Mantz, C., Bagshaw, H., Buyyounouski, M., Steinberg, M., Kupelian, P., Kishan, A. ELSEVIER SCIENCE INC. 2019: E17

  View details for DOI 10.1016/S0360-3016(19)30440-7

  View details for Web of Science ID 000463350600040

• Long-term Outcomes of Stereotactic Body Radiotherapy for Low-Risk and Intermediate-Risk Prostate Cancer. JAMA network open Kishan, A. U., Dang, A., Katz, A. J., Mantz, C. A., Collins, S. P., Aghdam, N., Chu, F., Kaplan, I. D., Appelbaum, L., Fuller, D. B., Meier, R. M., Loblaw, D. A., Cheung, P., Pham, H. T., Shaverdian, N., Jiang, N., Yuan, Y., Bagshaw, H., Prionas, N., Buyyounouski, M. K., Spratt, D. E., Linson, P. W., Hong, R. L., Nickols, N. G., Steinberg, M. L., Kupelian, P. A., King, C. R. 2019; 2 (2): e188006

  Abstract

  Importance: Stereotactic body radiotherapy harnesses improvements in technology to allow the completion of a course of external beam radiotherapy treatment for prostate cancer in the span of 4 to 5 treatment sessions. Although mounting short-term data support this approach, long-term outcomes have been sparsely reported.Objective: To assess long-term outcomes after stereotactic body radiotherapy for low-risk and intermediate-risk prostate cancer.Design, Setting, and Participants: This cohort study analyzed individual patient data from 2142 men enrolled in 10 single-institution phase 2 trials and 2 multi-institutional phase 2 trials of stereotactic body radiotherapy for low-risk and intermediate-risk prostate cancer between January 1, 2000, and December 31, 2012. Statistical analysis was performed based on follow-up from January 1, 2013, to May 1, 2018.Main Outcomes and Measures: The cumulative incidence of biochemical recurrence was estimated using a competing risk framework. Physician-scored genitourinary and gastrointestinal toxic event outcomes were defined per each individual study, generally by Radiation Therapy Oncology Group or Common Terminology Criteria for Adverse Events scoring systems. After central review, cumulative incidences of late grade 3 or higher toxic events were estimated using a Kaplan-Meier method.Results: A total of 2142 men (mean [SD] age, 67.9 [9.5] years) were eligible for analysis, of whom 1185 (55.3%) had low-risk disease, 692 (32.3%) had favorable intermediate-risk disease, and 265 (12.4%) had unfavorable intermediate-risk disease. The median follow-up period was 6.9 years (interquartile range, 4.9-8.1 years). Seven-year cumulative rates of biochemical recurrence were 4.5% (95% CI, 3.2%-5.8%) for low-risk disease, 8.6% (95% CI, 6.2%-11.0%) for favorable intermediate-risk disease, 14.9% (95% CI, 9.5%-20.2%) for unfavorable intermediate-risk disease, and 10.2% (95% CI, 8.0%-12.5%) for all intermediate-risk disease. The crude incidence of acute grade 3 or higher genitourinary toxic events was 0.60% (n=13) and of gastrointestinal toxic events was 0.09% (n=2), and the 7-year cumulative incidence of late grade 3 or higher genitourinary toxic events was 2.4% (95% CI, 1.8%-3.2%) and of late grade 3 or higher gastrointestinal toxic events was 0.4% (95% CI, 0.2%-0.8%).Conclusions and Relevance: In this study, stereotactic body radiotherapy for low-risk and intermediate-risk disease was associated with low rates of severe toxic events and high rates of biochemical control. These data suggest that stereotactic body radiotherapy is an appropriate definitive treatment modality for low-risk and intermediate-risk prostate cancer.

  View details for PubMedID 30735235

• Long-term Outcomes of Stereotactic Body Radiotherapy for Low-Risk and Intermediate-Risk Prostate Cancer JAMA NETWORK OPEN Kishan, A. U., Dang, A., Katz, A. J., Mantz, C. A., Collins, S. P., Aghdam, N., Chu, F., Kaplan, I. D., Appelbaum, L., Fuller, D. B., Meier, R. M., Loblaw, D., Cheung, P., Pham, H. T., Shaverdian, N., Jiang, N., Yuan, Y., Bagshaw, H., Prionas, N., Buyyounouski, M. K., Spratt, D. E., Linson, P. W., Hong, R. L., Nickols, N. G., Steinberg, M. L., Kupelian, P. A., King, C. R. 2019; 2 (2)

  View details for DOI 10.1001/jamanetworkopen.2018.8006

  View details for Web of Science ID 000465423800030

• Vulvar Recurrences After Intensity-modulated Radiation Therapy for Squamous Cell Carcinoma of the Anus. American journal of clinical oncology Bagshaw, H. P., Sause, W. T., Gawlick, U., Kim, H. T., Whisenant, J., Cannon, G. M. 2016: -?

  Abstract

  The objective is to determine localregional control (LRC), distant metastasis free survival, disease-free survival, overall survival (OS), and toxicity for patients with squamous cell carcinoma of the anus treated with definitive chemotherapy and intensity-modulated radiation therapy (IMRT).We conducted a retrospective review of patients treated using IMRT for squamous cell carcinoma of the anus at our institution since 2005. Patients with local recurrences were identified and reviewed. The Kaplan-Meier curves were used for LRC and OS.From 2005 to 2014, 52 patients were treated with IMRT-based chemoradiation for squamous cell carcinoma of the anus. Median dose to the primary tumor was 54 Gy. LRC, distant metastasis free survival, OS, and disease-free survival were 92.3%, 88.5%, 86.5%, and 84.6%, respectively, with a median follow-up of 20 months. Two local failures occurred at the anal primary site and 2 in the vulva. Despite subsequent palliative radiotherapy and chemotherapy, neither patient with a vulvar recurrence achieved disease control.In a cohort of patients treated with IMRT-based chemoradiation, 2 vulvar recurrences were identified within the avoided external genitalia despite limited recurrence rates within the cohort overall. This experience suggests that for patients with a locally advanced primary tumor and bulky bilateral inguinal or pelvic disease, the in-transit vulvar dermal lymphatics may be at risk for subclinical involvement and subsequent recurrence. If substantiated by a similar pattern of recurrence at other institutions, the external genitalia may need to be reclassified from an avoidance structure to a clinical treatment volume in patients with locally advanced anal cancer.

  View details for PubMedID 27438690

• Adjuvant radiotherapy for atypical meningiomas. Journal of neurosurgery Bagshaw, H. P., Burt, L. M., Jensen, R. L., Suneja, G., Palmer, C. A., Couldwell, W. T., Shrieve, D. C. 2016: 1–7

  Abstract

  OBJECTIVE The aim of this paper was to evaluate outcomes in patients with atypical meningiomas (AMs) treated with surgery alone compared with surgery and radiotherapy at initial diagnosis, or at the time of first recurrence. METHODS Patients with pathologically confirmed AMs treated at the University of Utah from 1991 to 2014 were retrospectively reviewed. Local control (LC), overall survival (OS), Karnofsky Performance Status (KPS), and toxicity were assessed. Outcomes for patients receiving adjuvant radiotherapy were compared with those for patients treated with surgery alone. Kaplan-Meier and the log-rank test for significance were used for LC and OS analyses. RESULTS Fifty-nine patients with 63 tumors were reviewed. Fifty-two patients were alive at the time of analysis with a median follow-up of 42 months. LC for all tumors was 57% with a median time to local failure (TTLF) of 48 months. The median TTLF following surgery and radiotherapy was 180 months, compared with 46 months following surgery alone (p = 0.02). Excluding Simpson Grade IV (subtotal) resections, there remained an LC benefit with the addition of radiotherapy for Simpson Grade I, II, and III resected tumors (median TTLF 180 months after surgery and radiotherapy compared with 46 months with surgery alone [p = 0.002]). Patients treated at first recurrence following any initial therapy (either surgery alone or surgery and adjuvant radiotherapy) had a median TTLF of 26 months compared with 48 months for tumors treated at first diagnosis (p = 0.007). There were 2 Grade 3 toxicities and 1 Grade 4 toxicity associated with radiotherapy. CONCLUSIONS Adjuvant radiotherapy improves LC for AMs. The addition of adjuvant radiotherapy following even a Simpson Grade I, II, or III resection was found to confer an LC benefit. Recurrent disease is difficult to control, underscoring the importance of aggressive initial treatment.

  View details for DOI 10.3171/2016.5.JNS152809

  View details for PubMedID 27611201

• Local Control of Melanoma Brain Metastases Treated with Stereotactic Radiosurgery Journal of Radiosurgery & SBRT Bagshaw, H. P., Ly, D., Suneja, G., Jensen, R. L., Shrieve, D. C. 2016; 4 (3): 181-190
• Palladium interstitial implant in combination with external beam radiotherapy and chemotherapy for the definitive treatment of a female urethral carcinoma. Gynecologic oncology reports Bagshaw, H. P., Williams, N. L., Huang, Y. J., Tward, J. D., Gaffney, D. K. 2015; 13: 40-43

  Abstract

  Primary urethral cancer is a rare diagnosis, especially in females. This report presents the utilization of a palladium interstitial implant and a review of the retrospective data published on the management of female urethral cancer. Excellent local control and survival has been obtained with the use of a palladium interstitial implant in combination with external beam radiotherapy and concurrent chemotherapy. This modality represents a novel and effective way to treat primary urethral cancer in females.

  View details for DOI 10.1016/j.gore.2015.06.001

  View details for PubMedID 26425719

• Local control after stereotactic radiosurgery for brain metastases in patients with melanoma with and without BRAF mutation and treatment JOURNAL OF NEUROSURGERY Ly, D., Bagshaw, H. P., Anker, C. J., Tward, J. D., Grossmann, K. F., Jensen, R. L., Shrieve, D. C. 2015; 123 (2): 395-401

  Abstract

  BRAF inhibitors improve progression-free and overall survival in patients with metastatic melanoma. Brain metastases are common, and stereotactic radiosurgery (SRS) has been used, resulting in excellent local control. Because BRAF inhibitors are associated with intracranial responses, the authors hypothesized that BRAF inhibitors would improve local control in patients with melanoma who are receiving SRS for brain metastases.The authors retrospectively identified patients with metastatic melanoma who had been tested for BRAF mutation and treated with SRS for brain metastases. Patients with previous resection, multiple brain metastases, or multiple courses of SRS were eligible. SRS was delivered in a single fraction to a median dose of 2000 cGy. Patients with a BRAF mutation were treated with a BRAF inhibitor on the basis of physician preference.The authors identified 52 patients who were treated in 82 treatment sessions for 185 brain metastases and 13 tumor beds. At a median follow-up of 10.5 months, the 1-year local control rate was 69.2%. At 1 year, the local control rate for brain metastases in patients with BRAF mutation with BRAF treatment was 85.0%, and the local control rate for brain metastases in those without BRAF treatment was 51.5% (p = 0.0077). The rates of distant brain failure, freedom from whole-brain radiation, and overall survival were not different on the basis of BRAF mutation status or inhibitor therapy. The number of new intratumoral hemorrhages after SRS was increased significantly in patients with BRAF treatment.Treatment with BRAF inhibitors was associated with improved local control after SRS in patients with melanoma and brain metastases. An increased number of intratumoral hemorrhages was associated with BRAF inhibitor therapy.

  View details for DOI 10.3171/2014.9.JNS141425

  View details for Web of Science ID 000362200800015

  View details for PubMedID 25768829

• Patterns of Care With Brachytherapy for Cervical Cancer INTERNATIONAL JOURNAL OF GYNECOLOGICAL CANCER Bagshaw, H. P., Pappas, L. M., Kepka, D. L., Tward, J. D., Gaffney, D. K. 2014; 24 (9): 1659-1664

  Abstract

  Concurrent chemotherapy with external beam radiotherapy (EBRT) and brachytherapy (BT) is critical to the curative treatment of locally advanced cervical cancer. Patterns of care and the use of EBRT and BT for locally advanced cervical cancer in the United States were analyzed with an emphasis on regional variation across the United States.A retrospective analysis was performed using the Surveillance, Epidemiology, and End Results Program database from 1988 to 2010 to identify women with locally advanced cervical carcinoma treated with definitive radiotherapy.Twelve thousand three hundred women were identified who met the inclusion criteria. From 1988 to 2010, percent use of EBRT and BT decreased from 68% to 45%; specifically, between 1988 and 2000, there was a decrease of 12% (P = 0.0003), and between 2000 and 2010, there was another decrease of 11% (P < 0.0001). When examined individually, 15 of the 16 registries displayed a decline in use of EBRT and BT with a significant decrease in 11 of the registries. No registry displayed an increased use of EBRT and BT, but the use of EBRT alone increased from 1988 to 2000 by 8% (P = 0.0055) and from 2000 to 2010 by 6% (P = 0.0095).Combination of EBRT and BT for locally advanced cervical cancer continues to decline, despite guidelines indicating the appropriateness of BT. This decline was seen for most regions across the United States, with a concomitant rise in the use of EBRT. EBRT alone is an inferior therapy and must be used in conjunction with BT to realize maximal patient benefit.

  View details for DOI 10.1097/IGC.0000000000000276

  View details for Web of Science ID 000344611800023

  View details for PubMedID 25251463

• Does family history of prostate cancer affect outcomes following radiotherapy? Radiotherapy and oncology Bagshaw, H., Ruth, K., Horwitz, E. M., Chen, D. Y., Buyyounouski, M. K. 2014; 110 (2): 229-234

  Abstract

  To examine family history (FH) as a prognostic factor following radiotherapy (RT).Between 1989 and 2007, 1711 men with clinically localized prostate cancer and complete family history who had received RT (median RT dose=74Gy) without androgen deprivation therapy were analyzed. FH was defined as any prostate cancer in a first degree relative. For the biochemical failure (BF) outcome, this sample size has 85% power to detect a hazard ratio of 1.56 for positive versus negative FH.With a median follow-up of 71 months, there was no significant difference in the distribution of Gleason score (GS) or prostate specific antigen (PSA) based on FH. A positive FH was not an independent predictor of BF, distant metastasis (DM), prostate cancer specific mortality (PCSM), or overall mortality (OM) in Cox proportional multivariable analysis. On further analysis in a Cox proportional multivariable analysis, men with two or more first degree relatives with prostate cancer had a significantly higher likelihood of BF and DM than those with no FH, although there was no difference in PCSM or OM. Men with a positive FH (23%) were more likely to be younger, have a lower PSA, and non-palpable disease. There was no interaction between a positive FH and neither race nor treatment era (pre-PSA vs. PSA era).A positive FH is not a prognostic factor following RT and should not alter standard treatment recommendations. Patients with two or more first degree relatives with prostate cancer had a higher likelihood of BF and DM, but there was no effect on survival. There was no interaction between a positive FH and African American race or treatment era. A positive FH was however, associated with more favorable PSA values and T-stage that may be the result of earlier screening.

  View details for DOI 10.1016/j.radonc.2013.11.014

  View details for PubMedID 24560758

• Radiation Therapy in Male Breast Cancer Oncology & Hematology Review Bagshaw, H. P., Cloyd, J. M., Poppe, M. M., Wapnir, I. L. 2014; 10: 61-65