Engineering Papers⌕ Search

Engineering topics

Park, Ji Won

Publications and source records attributed to Park, Ji Won.

Latent Stochastic Differential Equations for Modeling Quasar Variability and Inferring Black Hole Properties

Quasars are bright and unobscured active galactic nuclei (AGN) thought to be powered by the accretion of matter around supermassive black holes at the centers of galaxies. The temporal variability of a quasar’s brightness contains valuable information about its physical properties. The UV/optical variability is thought to be a stochastic process, often represented as a damped random walk described by a stochastic differential equation (SDE). Upcoming wide-field telescopes such as the Rubin Observatory Legacy Survey of Space and Time (LSST) are expected to observe tens of millions of AGN in multiple filters over a ten year period, so there is a need for efficient and automated modeling techniques that can handle the large volume of data. Latent SDEs are machine learning models well suited for modeling quasar variability, as they can explicitly capture the underlying stochastic dynamics. In this work, we adapt latent SDEs to jointly reconstruct multivariate quasar light curves and infer their physical properties such as the black hole mass, inclination angle, and temperature slope. Our model is trained on realistic simulations of LSST ten year quasar light curves, and we demonstrate its ability to reconstruct quasar light curves even in the presence of long seasonal gaps and irregular sampling across different bands, outperforming a multioutput Gaussian process regression baseline. Our method has the potential to provide a deeper understanding of the physical properties of quasars and is applicable to a wide range of other multivariate time series with missing data and irregular sampling.

79 ASTRONOMY AND ASTROPHYSICS↗

Large-scale Gravitational Lens Modeling with Bayesian Neural Networks for Accurate and Precise Inference of the Hubble Constant

We investigate the use of approximate Bayesian neural networks (BNNs) in modeling hundreds of time delay gravitational lenses for Hubble constant (H 0 ) determination. Here, our BNN was trained on synthetic Hubble Space Telescope quality images of strongly lensed active galactic nuclei with lens galaxy light included. The BNN can accurately characterize the posterior probability density functions (PDFs) of model parameters governing the elliptical power-law mass profile in an external shear field. We then propagate the BNN-inferred posterior PDFs into an ensemble H 0 inference, using simulated time delay measurements from a plausible dedicated monitoring campaign. Assuming well-measured time delays and a reasonable set of priors on the environment of the lens, we achieve a median precision of 9.3% per lens in the inferred H 0 . A simple combination of a set of 200 test lenses results in a precision of 0.5 km s –1 Mpc –1 (0.7%), with no detectable bias in this H 0 recovery test. The computation time for the entire pipeline—including the generation of the training set, BNN training and H0 inference—translates to 9 minutes per lens on average for 200 lenses and converges to 6 minutes per lens as the sample size is increased. Being fully automated and efficient, our pipeline is a promising tool for exploring ensemble-level systematics in lens modeling for H 0 inference.

79 ASTRONOMY AND ASTROPHYSICS↗

DESC DC2 Data Release Note

In preparation for cosmological analyses of the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST), the LSST Dark Energy Science Collaboration (LSST DESC) has created a 300 deg$^2$ simulated survey as part of an effort called Data Challenge 2 (DC2). The DC2 simulated sky survey, in six optical bands with observations following a reference LSST observing cadence, was processed with the LSST Science Pipelines (19.0.0). In this Note, we describe the public data release of the resulting object catalogs for the coadded images of five years of simulated observations along with associated truth catalogs. We include a brief description of the major features of the available data sets. To enable convenient access to the data products, we have developed a web portal connected to Globus data services. We describe how to access the data and provide example Jupyter Notebooks in Python to aid first interactions with the data. We welcome feedback and questions about the data release via a GitHub repository.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Preoperative Short-Course Concurrent Chemoradiation Therapy Followed by Delayed Surgery for Locally Advanced Rectal Cancer: A Phase 2 Multicenter Study (KROG 10-01)

Purpose: A prospective phase 2 multicenter trial was performed to investigate the efficacy and safety of preoperative short-course concurrent chemoradiation therapy (CRT) followed by delayed surgery for patients with locally advanced rectal cancer. Methods and Materials: Seventy-three patients with cT3-4 rectal cancer were enrolled. Radiation therapy of 25 Gy in 5 fractions was delivered over 5 consecutive days using helical tomotherapy. Concurrent chemotherapy was administered on the same 5 days with intravenous bolus injection of 5-fluorouracil (400 mg/m{sup 2}/day) and leucovorin (20 mg/m{sup 2}/day). After 4 to 8 weeks, total mesorectal excision was performed. The primary endpoint was the pathologic downstaging (ypStage 0-I) rate, and secondary endpoints included tumor regression grade, tumor volume reduction rate, and toxicity. Results: Seventy-one patients completed the planned preoperative CRT and surgery. Downstaging occurred in 20 (28.2%) patients, including 1 (1.4%) with a pathologic complete response. Favorable tumor regression (grade 4-3) was observed in 4 (5.6%) patients, and the mean tumor volume reduction rate was 62.5 ± 21.3%. Severe (grade ≥3) treatment toxicities were reported in 27 (38%) patients from CRT until 3 months after surgery. Conclusions: Preoperative short-course concurrent CRT followed by delayed surgery for patients with locally advanced rectal cancer demonstrated poor pathologic responses compared with conventional long-course CRT, and it yielded considerable toxicities despite the use of an advanced radiation therapy technique.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Long-Term Follow-Up of Preoperative Pelvic Radiation Therapy and Concomitant Boost Irradiation in Locally Advanced Rectal Cancer Patients: A Multi-Institutional Phase II Study (KROG 04-01)

Purpose: To perform a prospective phase II study to investigate the efficacy and safety of preoperative pelvic radiation therapy and concomitant small-field boost irradiation with 5-fluorouracil and leucovorin for 5 weeks in locally advanced rectal cancer patients. Methods and Materials: Sixty-nine patients with locally advanced, nonmetastatic, mid-to-lower rectal cancer were prospectively enrolled. They had received preoperative chemoradiation therapy and total mesorectal excision. Pelvic radiation therapy of 43.2 Gy in 24 fractions plus concomitant boost radiation therapy of 7.2 Gy in 12 fractions was delivered to the pelvis and tumor bed for 5 weeks. Two cycles of 5-fluorouracil and leucovorin were administered for 3 days in the first and fifth week of radiation therapy. The pathologic response, survival outcome, and treatment toxicity were evaluated for the study endpoints. Results: Of 69 patients, 8 (11.6%) had a pathologically complete response. Downstaging rates were 40.5% for T classification and 68.1% for N classification. At the median follow-up of 69 months, 36 patients have been followed up for more than 5 years. The 5-year disease-free survival (DFS) and overall survival rates were 66.0% and 75.3%, respectively. Higher pathologic T (P = .045) and N (P = .032) classification were significant adverse prognostic factors for DFS, and high-grade histology was an adverse prognostic factor for both DFS (P = .025) and overall survival (P = .031) on the multivariate analysis. Fifteen patients (21.7%) experienced grade 3 or 4 acute toxicity, and 7 patients (10.1%) had long-term toxicity. Conclusion: Preoperative pelvic radiation therapy with concomitant boost irradiation with 5-fluorouracil and leucovorin for 5 weeks showed acceptable acute and long-term toxicities. However, the benefit of concomitant small-field boost irradiation for 5 weeks in rectal cancer patients was not demonstrated beyond conventional irradiation for 6 weeks in terms of tumor response and survival.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Stage-to-Stage Comparison of Preoperative and Postoperative Chemoradiotherapy for T3 Mid or Distal Rectal Cancer

Purpose: To investigate, in a comparative analysis, the prognostic implications of postchemoradiotherapy (post-CRT) pathologic stage (ypStage) vs. postoperative pathologic stage (pStage) in rectal cancer. Methods and Materials: Between May 2001 and December 2006, 487 patients with T3 mid or distal rectal cancer were analyzed retrospectively. Concurrent CRT was administered preoperatively (n = 364, 74.7%) or postoperatively (n = 123, 25.3%). The radiation dose was 50.4 Gy in 28 fractions. All patients underwent a total mesorectal excision and received adjuvant chemotherapy. Disease-free survival (DFS) was estimated using the Kaplan-Meier method. Differences in DFS, stratified by ypStage and pStage, were compared using the log-rank test. Results: For surviving patients, the median follow-up period was 68 months (range, 12-105 months). The 5-year local recurrence-free survival rate was not different, at 95.3% and 92.1% in preoperative and postoperative CRT groups, respectively (p = 0.402), but the 5-year distant metastasis-free survival rate was significantly different, at 81.6% (preoperative CRT) vs. 65.4% (postoperative CRT; p = 0.001). The 5-year DFS rate of 78.8% in the preoperative CRT group was significantly better than the 63.0% rate in the postoperative CRT group (p = 0.002). Post-CRT pathologic Stage 0-I occurred in 42.6% (155 of 364) of the patients with preoperative CRT. The 5-year DFS rates were 90.2% (ypStage 0-I), 83.5% (ypStage II), 77.3% (pStage II), 58.6% (ypStage III), and 54.7% (pStage III). The DFS rate of ypStage 0-I was significantly better than that of ypStage II or pStage II. Post-CRT pathologic Stage II and III had similar DFS, compared with pStage II and III, respectively. Conclusions: Disease-free survival predicted by each ypStage was similar to that predicted by the respective pStage. Improved DFS with preoperative vs. postoperative CRT was associated with the ypStage 0-I group that showed a similarly favorable outcome to pStage I rectal cancer.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Tumor Volume Reduction Rate After Preoperative Chemoradiotherapy as a Prognostic Factor in Locally Advanced Rectal Cancer

Purpose: To investigate the prognostic significance of tumor volume reduction rate (TVRR) after preoperative chemoradiotherapy (CRT) in locally advanced rectal cancer (LARC). Methods and Materials: In total, 430 primary LARC (cT3-4) patients who were treated with preoperative CRT and curative radical surgery between May 2002 and March 2008 were analyzed retrospectively. Pre- and post-CRT tumor volumes were measured using three-dimensional region-of-interest MR volumetry. Tumor volume reduction rate was determined using the equation TVRR (%) = (pre-CRT tumor volume - post-CRT tumor volume) Multiplication-Sign 100/pre-CRT tumor volume. The median follow-up period was 64 months (range, 27-99 months) for survivors. Endpoints were disease-free survival (DFS) and overall survival (OS). Results: The median TVRR was 70.2% (mean, 64.7% {+-} 22.6%; range, 0-100%). Downstaging (ypT0-2N0M0) occurred in 183 patients (42.6%). The 5-year DFS and OS rates were 77.7% and 86.3%, respectively. In the analysis that included pre-CRT and post-CRT tumor volumes and TVRR as continuous variables, only TVRR was an independent prognostic factor. Tumor volume reduction rate was categorized according to a cutoff value of 45% and included with clinicopathologic factors in the multivariate analysis; ypN status, circumferential resection margin, and TVRR were significant prognostic factors for both DFS and OS. Conclusions: Tumor volume reduction rate was a significant prognostic factor in LARC patients receiving preoperative CRT. Tumor volume reduction rate data may be useful for tailoring surgery and postoperative adjuvant therapy after preoperative CRT.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Tumor Volume Reduction Rate Measured by Magnetic Resonance Volumetry Correlated With Pathologic Tumor Response of Preoperative Chemoradiotherapy for Rectal Cancer

Purpose: To determine whether the tumor volume reduction rate (TVRR) measured using three-dimensional region-of-interest magnetic resonance volumetry correlates with the pathologic tumor response after preoperative chemoradiotherapy (CRT) for locally advanced rectal cancer. Methods and Materials: The study included 405 patients with locally advanced rectal cancer (cT3-T4) who had undergone preoperative CRT and radical proctectomy. The tumor volume was measured using three-dimensional region-of-interest magnetic resonance volumetry before and after CRT but before surgery. We analyzed the correlation between the TVRR and the pathologic tumor response in terms of downstaging and tumor regression grade (TRG). Downstaging was defined as ypStage 0-I (ypT0-T2N0M0), and the TRG proposed by Dworak et al. was used. Results: The mean TVRR was 65.0% {+-} 22.3%. Downstaging and complete regression occurred in 167 (41.2%) and 58 (14.3%) patients, respectively. The TVRRs according to ypT classification (ypT0-T2 vs. ypT3-T4), ypN classification (ypN0 vs. ypN1-N2), downstaging (ypStage 0-I vs. ypStage II-III), good regression (TRG 3-4 vs. TRG 1-2), and complete regression (TRG 4 vs. TRG 1-3) were all significantly different (p <.05). When the TVRR was categorized into three groups (<60%, 60-80%, and >80%), the rates of ypT0-T2, ypN0, downstaging, and good regression were all significantly greater for patients with a TVRR of {>=}60%, as was the complete regression rate for patients with a TVRR >80% (p <.05). Conclusion: The TVRR measured using three-dimensional region-of-interest magnetic resonance volumetry correlated significantly with the pathologic tumor response in terms of downstaging and TRG after preoperative CRT for locally advanced rectal cancer.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Carcinoembryonic Antigen as a Predictor of Pathologic Response and a Prognostic Factor in Locally Advanced Rectal Cancer Patients Treated With Preoperative Chemoradiotherapy and Surgery

Purpose: To evaluate the role of serum carcinoembryonic antigen (CEA) as a predictor of response to preoperative chemoradiotherapy (CRT) and prognostic factor for rectal cancer. Materials and Methods: The study retrospectively evaluated 352 locally advanced rectal cancer patients who underwent preoperative CRT followed by surgery. Serum CEA levels were determined before CRT administration (pre-CRT CEA) and before surgery (post-CRT CEA). Correlations between pre-CRT CEA levels and rates of good response (Tumor regression grade 3/4) were explored. Patients were categorized into three CEA groups according to their pre-/post-CRT CEA levels (ng/mL) (Group A: pre-CRT CEA {<=} 3; B: pre-CRT CEA >3, post-CRT CEA {<=}3; C: pre- and post-CRT CEA >3 ng/mL), and their oncologic outcomes were compared. Results: Of 352 patients, good responses were achieved in 94 patients (26.7%). The rates of good response decreased significantly as the pre-CRT CEA levels became more elevated (CEA [ng/mL]: {<=}3, 36.4%; 3-6, 23.6%; 6-9, 15.6%; >9, 7.8%; p < 0.001). The rates of good response were significantly higher in Group A than in Groups B and C (36.4% vs. 17.3% and 14.3%, respectively; p < 0.001). The 3-year disease-free survival rate was significantly better in Groups A and B than in Group C (82% and 79% vs. 57%, respectively; p = 0.005); the CEA grouping was identified as an independent prognostic factor (p = 0.025). Conclusions: In locally advanced rectal cancer patients, CEA levels could be of clinical value as a predictor of response to preoperative CRT and as an independent prognostic factor after preoperative CRT and curative surgery.

62 RADIOLOGY AND NUCLEAR MEDICINE↗