University of Tsukuba

About: University of Tsukuba is a education organization based out in Tsukuba, Ibaraki, Japan. It is known for research contribution in the topics: Population & Gene. The organization has 36352 authors who have published 79483 publications receiving 1934752 citations. The organization is also known as: Tsukuba daigaku & Tsukuba University.

TAS-102 monotherapy for pretreated metastatic colorectal cancer: a double-blind, randomised, placebo-controlled phase 2 trial

Abstract: Summary Background Treatments that confer survival benefit are needed in patients with heavily pretreated metastatic colorectal cancer. The aim of this trial was to investigate the efficacy and safety of TAS-102—a novel oral nucleoside antitumour agent. Methods Between August 25, 2009, and April 12, 2010, we undertook a multicentre, double-blind, randomised, placebo-controlled phase 2 trial in Japan. Eligible patients were 20 years or older; had confirmed colorectal adenocarcinoma; had a treatment history of two or more regimens of standard chemotherapy; and were refractory or intolerant to fluoropyrimidine, irinotecan, and oxaliplatin. Patients had to be able to take oral drugs; have measurable lesions; have an Eastern Cooperative Oncology Group performance status of between 0 and 2; and have adequate bone-marrow, hepatic, and renal functions within 7 days of enrolment. Patients were randomly assigned (2:1) to either TAS-102 (35 mg/m 2 given orally twice a day in a 28-day cycle [2-week cycle of 5 days of treatment followed by a 2-day rest period, and then a 14-day rest period]) or placebo; all patients received best supportive care. Randomisation was done with minimisation methods, with performance status as the allocation factor. The randomisation sequence was generated with a validated computer system by an independent team from the trial sponsor. Investigators, patients, data analysts, and the trial sponsor were masked to treatment assignment. The primary endpoint was overall survival in the intention-to-treat population. Safety analyses were done in the per-protocol population. The study is in progress and is registered with Japan Pharmaceutical Information Center, number JapicCTI-090880. Findings 112 patients allocated to TAS-102 and 57 allocated to placebo made up the intention-to-treat population. Median follow-up was 11·3 months (IQR 10·7–14·0). Median overall survival was 9·0 months (95% CI 7·3–11·3) in the TAS-102 group and 6·6 months (4·9–8·0) in the placebo group (hazard ratio for death 0·56, 80% CI 0·44–0·71, 95% CI 0·39–0·81; p=0·0011). 57 (50%) of 113 patients given TAS-102 in the safety population had neutropenia of grade 3 or 4, 32 (28%) leucopenia, and 19 (17%) anaemia. No patient given placebo had grade 3 or worse neutropenia or leucopenia; three (5%) of 57 had grade 3 or worse anaemia. Serious adverse events occurred in 21 (19%) patients in the TAS-102 group and in five (9%) in the placebo group. No treatment-related deaths occurred. Interpretation TAS-102 has promising efficacy and a manageable safety profile in patients with metastatic colorectal cancer who are refractory or intolerant to standard chemotherapies. Funding Taiho Pharmaceutical.

Deep reverse tone mapping

Abstract: Inferring a high dynamic range (HDR) image from a single low dynamic range (LDR) input is an ill-posed problem where we must compensate lost data caused by under-/over-exposure and color quantization. To tackle this, we propose the first deep-learning-based approach for fully automatic inference using convolutional neural networks. Because a naive way of directly inferring a 32-bit HDR image from an 8-bit LDR image is intractable due to the difficulty of training, we take an indirect approach; the key idea of our method is to synthesize LDR images taken with different exposures (i.e., bracketed images) based on supervised learning, and then reconstruct an HDR image by merging them. By learning the relative changes of pixel values due to increased/decreased exposures using 3D deconvolutional networks, our method can reproduce not only natural tones without introducing visible noise but also the colors of saturated pixels. We demonstrate the effectiveness of our method by comparing our results not only with those of conventional methods but also with ground-truth HDR images.