Search

This article presents a computing-in-memory (CIM) structure aimed at improving the energy efficiency of edge devices running multi-bit multiply-and-accumulate (MAC) operations. The proposed scheme includes a 6T SRAM-based CIM (SRAM-CIM) macro capable of: 1) weight-bitwise MAC (WbwMAC) operations to expand the sensing margin and improve the readout accuracy for high-precision MAC operations; 2) a compact 6T local computing cell to perform multiplication with suppressed sensitivity to process variation; 3) an algorithm-adaptive low MAC-aware readout scheme to improve energy efficiency; 4) a bitline header selection scheme to enlarge signal margin; and 5) a small-offset margin-enhanced sense amplifier for robust read operations against process variation. A fabricated 28-nm 64-kb SRAM-CIM macro achieved access times of 4.1–8.4 ns with energy efficiency of 11.5–68.4 TOPS/W, while performing MAC operations with 4- or 8-b input and weight precision.

The association between hepatitis C virus (HCV) infection and chronic kidney disease (CKD) still remains controversial. We aimed to investigate whether HCV really affects renal function, and to analyze the association between clinical effects of CHC and decreased kidney function (assessed by glomerular filtration rate (eGFR) level).An estimated 3360 patients with HCV infection and 3360 age- and sex-matched community-based control individuals without HCV were enrolled (1:1, case and control ratio) in this study between 2004 and 2016. We used the modification of diet in renal diseases to calculate eGFR. Demographic and laboratory parameters were assessed, and appropriate statistical methods were performed for the analysis.Multivariate logistic regression analysis revealed that serum alanine aminotransferase level (odds ratio [OR] 0.998; 95% confidence interval [CI] 0.997-0.999; P = 0.001), platelet count (OR 0.997; 95% CI 0.995-0.999; p = 0.002), and hypertension (OR 1.31; 95% CI 1.03-1.66; P = 0.027) were significantly associated with HCV infection and serum triglyceride levels (OR 1.001; 95% CI 1.00-1.002; p = 0.005), platelet count (OR 0.996; 95% CI 0.995-0.997; p0.001), body mass index (BMI)25 (OR 1.43; 95% CI 1.23-1.67; p0.001), hypertension (OR 1.69; 95% CI 1.42-1.99; p0.001), hyperlipidemia (OR 1.32; 95% CI 1.02-1.71; p = 0.035), and diabetes (OR 1.33; 95% CI 1.03-1.71; p = 0.032) were significantly associated with a low eGFR (90 mL/min/m3) in control subjects. The BMI25 kg/m2, hypertension, and diabetes were found to be associated with low eGFR interaction with the HCV infection, via a multivariate analysis.Our study found that the patients with HCV infection are associated with a low eGFR compared with non-HCV-infected patients. This association is consistent in obese, diabetic, and hypertensive patients.

The spreading of viral hepatitis among injecting drug users (IDU) is an emerging public health concern. This study explored the prevalence and the risks of hepatitis B virus (HBV), hepatitis C virus (HCV) and hepatitis D virus (HDV) among IDU-dominant prisoners in Taiwan. HBV surface antigen (HBsAg), antibodies to HCV (anti-HCV) and HDV (anti-HDV), viral load and HCV genotypes were measured in 1137(67.0%) of 1697 prisoners. 89.2% of participants were IDUs and none had HIV infection. The prevalence of HBsAg, anti-HCV, dual HBsAg/anti-HCV, HBsAg/anti-HDV, and triple HBsAg/anti-HCV/anti-HDV was 13.6%, 34.8%, 4.9%, 3.4%, and 2.8%, respectively. HBV viremia rate was significantly lower in HBV/HCV-coinfected than HBV mono-infected subjects (66.1% versus 89.9%, adjusted odds ratio/95% confidence intervals [aOR/CI] = 0.27/0.10–0.73). 47.5% anti-HCV-seropositive subjects (n = 396) were non-viremic, including 23.2% subjects were antivirals-induced. The predominant HCV genotypes were genotype 6(40.9%), 1a(24.0%) and 3(11.1%). HBsAg seropositivity was negatively correlated with HCV viremia among the treatment naïve HCV subjects (44.7% versus 72.4%, aOR/CI = 0.27/0.13–0.58). Anti-HCV seropositivity significantly increased the risk of anti-HDV-seropositivity among HBsAg carriers (57.1% versus 7.1%, aOR/CI = 15.73/6.04–40.96). In conclusion, IUDs remain as reservoirs for multiple hepatitis viruses infection among HIV-uninfected prisoners in Taiwan. HCV infection increased the risk of HDV infection but suppressed HBV replication in HBsAg carriers. An effective strategy is mandatory to control the epidemic in this high-risk group.

Background/Aims: Growth hormone (GH) is the main regulator of somatic growth, metabolism, and gender dimorphism in the liver. GH receptor (GHR) signaling in cancer is derived from a large body of evidence, although the GHR signaling pathway involved in the prognosis of hepatocellular carcinoma (HCC) in patients with hepatitis C virus (HCV)-related HCC, remains unclear. We aimed to explore the expression of GHR and analyze its association with clinicopathologic features and prognosis of patients with chronic hepatitis C and HCC.Methods: The expression of GHR mRNA was investigated by quantitative real-time polymerase chain reaction in paired tumors and adjacent non-tumorous (ANT) liver tissues of 200 patients with chronic hepatitis C and HCC. Western blotting and immunofluorescence assays using the HCV-infected Huh7.5.1 cell model was performed.Results: GHR mRNA was significantly lower in HCV-HCC tissues than in corresponding ANT liver tissues. GHR mRNA and protein levels also decreased in the HCV-infected Huh7.5.1 cell model. Notably, lower GHR expression was associated with age of >60 years (P=0.0111) and worse clinicopathologic characteristics, including alpha-fetoprotein >100 ng/mL (P=0.0403), cirrhosis (P=0.0075), vascular invasion (P=0.0052), pathological stage II–IV (P=0.0002), and albumin ≤4.0 g/dL (P=0.0055), which were linked with poor prognosis of HCC. Most importantly, the high incidence of recurrence and poor survival rates in patients with a low ratio of tumor/ANT GHR (≤0.1) were observed, indicating that low expression levels of GHR had great risk for development of HCC in patients with chronic hepatitis C.Conclusions: Our study demonstrates a significant down-regulation of GHR expression as a new unfavorable independent prognostic factor in patients with chronic hepatitis C and HCC.

Hepatitis D virus (HDV) infection increases the risk of hepatocellular carcinoma (HCC) in the natural course of chronic hepatitis B (CHB) patients. Its role in patients treated with nucleotide/nucleoside analogues (NAs) is unclear. We aimed to study the role of hepatitis D in the development of HCC in CHB patients treated with NAs. Altogether, 1349 CHB patients treated with NAs were tested for anti-HDV antibody and RNA. The incidence and risk factors of HCC development were analyzed. Rates of anti-HDV and HDV RNA positivity were 2.3% and 1.0%, respectively. The annual incidence of HCC was 1.4 per 100 person-years after a follow-up period of over 5409.5 person-years. The strongest factor association with HCC development was liver cirrhosis (hazard ratio [HR]/95% confidence interval [CI] 9.98/5.11–19.46, P P = 0.02), age > 50 years old (HR/CI 3.64/2.03–6.54, P P: 0.01), and body mass index (BMI, HR/CI 1.11/1.03–1.18, P = 0.004). The 5-year cumulative incidence of HCC was 7.3% for patients with HDV RNA negativity compared to that of 22.2% for patients with HDV RNA positivity (P = 0.01). In the subgroup of cirrhotic patients, the factors associated with HCC development were HDV RNA positivity (HR/CI 4.45/1.04–19.09, P = 0.04) and BMI (HR/CI 1.11/1.03–1.19, P = 0.01). HDV viremia played a crucial role in HCC development in CHB patients who underwent NA therapy.

The development of small, energy-efficient artificial intelligence edge devices is limited in conventional computing architectures by the need to transfer data between the processor and memory. Non-volatile compute-in-memory (nvCIM) architectures have the potential to overcome such issues, but the development of high-bit-precision configurations required for dot-product operations remains challenging. In particular, input–output parallelism and cell-area limitations, as well as signal margin degradation, computing latency in multibit analogue readout operations and manufacturing challenges, still need to be addressed. Here we report a 2 Mb nvCIM macro (which combines memory cells and related peripheral circuitry) that is based on single-level cell resistive random-access memory devices and is fabricated in a 22 nm complementary metal–oxide–semiconductor foundry process. Compared with previous nvCIM schemes, our macro can perform multibit dot-product operations with increased input–output parallelism, reduced cell-array area, improved accuracy, and reduced computing latency and energy consumption. The macro can, in particular, achieve latencies between 9.2 and 18.3 ns, and energy efficiencies between 146.21 and 36.61 tera-operations per second per watt, for binary and multibit input–weight–output configurations, respectively. Commercial complementary metal–oxide–semiconductor and resistive random-access memory technologies can be used to create multibit compute-in-memory circuits capable of fast and energy-efficient inference for use in small artificial intelligence edge devices.

BACKGROUND AND AIM The serial serologic changes of hepatitis D virus (HDV) infection among chronic hepatitis B virus (HBV) infected patients who received oral nucleotide/nucleoside analogues are elusive. METHODS Serum anti-HDV and HDV RNA among chronic hepatitis B (CHB) patients were tested at the time of initiating anti-HBV therapy and subsequently during the follow-up period. RESULTS The seropositive rate of anti-HDV and HDV RNA among 2850 CHB patients, was 2.7% and 0.9%, respectively. Factors associated with anti-HDV seropositivity were platelet counts (odds ratio [OR]/95% confidence intervals [CI]: 0.995/0.992-0.999; P = 0.006), HBV DNA levels (OR/CI: 0.81/0.70-0.94; P = 0.005), and hepatitis B e-antigen (HBeAg) seropositivity (OR/CI: 0.22/0.05-0.95; P = 0.04). The only factor associated with HDV RNA positivity among anti-HDV seropositive patients was age (OR/CI: 0.95/0.90-1.00; P = 0.03). The spontaneous clearance rate of serum anti-HDV antibody was 3.0 per 100 person-years with a median follow-up period of 3.5 years (range 2-12 years), whereas the seroclearance rate of HDV RNA was 4.3 per 100 person-years among anti-HDV seropositive patients after a median follow-up period of 6.0 years (range 2-11 years). A baseline anti-HDV titer

Recent SRAM-based computation-in-memory (CIM) macros enable mid-to-high precision multiply-and-accumulate (MAC) operations with improved energy efficiency using ultra-small/small capacity (0.4-8KB) memory devices. However, advanced CIM-based edge-AI chips favor multiple mid/large capacity SRAM-CIM macros: with high input (IN) and weight (W) precision to reduce the frequency of data reloads from external DRAM, and to avoid the need for additional SRAM buffers or ultra-large on-chip weight buffers. However, enlarging memory capacity and throughput increases the delay parasitics on WLs and BLs, and the number of parallel computing elements; resulting in longer compute latency (t AC ), lower energy-efficiency (EF), degraded signal margin, and larger fluctuations in power consumption across data-patterns (see Fig. 16.3.1). Recent SRAM-CIM macros tend to not use in-lab SRAM cells, with a logic-based layout, in favor of foundry provided compact-layout 8T [2], 3, [5] or 6T cells with local-computing cells (LCCs) [4], [6] to reduce the cell-array area and facilitate manufacturing. This paper presents a SRAM-CIM structure using (1) a segmented-BL charge-sharing (SBCS) scheme for MAC operations, with low energy consumption and a consistently high signal margin across MAC values (MACV); (2) An new LCC cell, called a source-injection local-multiplication cell (SILMC), to support the SBCS scheme with a consistent signal margin against transistor process variation; and (3) A prioritized-hybrid-ADC (Ph-ADC) to achieve a small area and power overhead for analog readout. A 28nm 384kb SRAM-CIM macro was fabricated using a foundry compact-6T cell with support for MAC operations with 16 accumulations of 8b-inputs and 8b-weights with near-full precision output (20b). This macro achieves a 7.2ns t AC and a 22.75TOPS/W EF for 8b-MAC operations with an FoM (IN-precision × W-precision × output-ratio × output-channel × EF/t AC ) 6× higher than prior work.

Battery-powered tiny-AI edge devices require large-capacity nonvolatile compute-in-memory (nvCIM), with multibit input (IN), weight (W), and output (OUT) precision to support complex applications, high-energy efficiency (EF$_{{\mathrm {MAC}}})$, and short computing latency $(t_{{\mathrm {AC}}})$ for multiply-and-accumulate (MAC) operations. Due to the low read-disturbfree voltage of nonvolatile memory (NVM) devices and the large parasitic load on the bitline, most existing Mb-level nvCIM macros use a current-mode read scheme [1–5] and only achieve a low IN-W precision (binary to 4b).

BACKGROUND The World Health Organization (WHO) set out to eliminate hepatitis C virus (HCV) infection by 2030, a goal Taiwan might achieve before 2025. Using effective direct antiviral agents (DAAs) against chronic hepatitis C (CHC) in Taiwan, the treatment of CHC has been initiated in rural areas. Here, we aimed to elucidate the clinical and virological characteristics of HCV infection, and the treatment efficacy of DAAs in patients from Pingtung county in southern Taiwan. METHODS A total of 152 chronic hepatitis patients treated with DAAs were consecutively enrolled. Baseline characteristics and therapeutic efficacy were evaluated. RESULTS HCV genotype 2 was the most common viral genotype (39.5%), followed by 1b (36.8%), 6 (10.5%), and 1a (9.2%). The sustained virological response (SVR) rate was 98.7%. Hakka patients accounted for 22.4% of the study cohort, of which 14.7% had HCV genotype 6. There were no differences in clinical characteristics between Hakka and non-Hakka patients. Patients with HCV genotype 6 were younger in age (OR/CI: 0.95/0.91-1.00, p = 0.04) and composed of more people who inject drugs (PWID) (OR/CI: 17.6/3.6-85.5, p

Background/Aims Undetectable HCV RNA 12 weeks after the end of treatment (SVR12) has been the valid efficacy endpoint in the era of direct-acting antivirals (DAAs). Its concordance with SVR4 and SVR24 and long-term durability is unknown in Taiwanese chronic hepatitis C (CHC) patients. Methods A total of 1080 CHC patients who received all-oral DAAs and an achieved end-of-treatment virological response (EOTVR), defined as undetectable HCV RNA at the end of therapy, were consecutively enrolled. HCV RNA was monitored 4, 12, and 24 weeks after EOT. Patients who achieved SVR24, defined as undetectable HCV RNA 24 weeks after EOT, were followed annually for assessing SVR durability. Results Eleven (1.02%) patients experienced HCV RNA reappearance after EOT. The most frequent timing of RNA reappearance was observed at SVR4 (n = 7), followed by SVR12 (n = 3) and SVR 24 (n = 1). The positive predictive value (PPV) and negative predictive value (NPV) of SVR4 in predicting SVR12 were 99.7% and 100%, respectively, whereas the PPV and NPV of SVR12 in predicting SVR24 were 99.9% and 100%, respectively. Pyrosequencing confirmed delayed relapse rather than reinfection for the patient who had detectable HCV RNA at SVR24. Among 978 patients who achieved SVR24, after a median follow-up period of 17.3±8.2 months, the SVR durability is 100% up to a 4-year follow-up. Conclusion Achievement of SVR12 provides excellent durability of HCV seroclearance after DAA therapy. On-demand HCV RNA beyond SVR12 should be recommended for patients with unexplainable abnormal liver function or high-risk behaviors.

ObjectiveHCV prevails in uremic haemodialysis patients. The current study aimed to achieve HCV microelimination in haemodialysis centres through a comprehensive outreach programme.DesignThe ERASE-C Campaign is an outreach programme for the screening, diagnosis and group treatment of HCV encompassing 2323 uremic patients and 353 medical staff members from 18 haemodialysis centres. HCV-viremic subjects were linked to care for directly acting antiviral therapy or received on-site sofosbuvir/velpatasvir therapy. The objectives were HCV microelimination (>80% reduction of the HCV-viremic rate 24 weeks after the end of the campaign in centres with ≥90% of the HCV-viremic patients treated) and ‘No-C HD’ (no HCV-viremic subjects at the end of follow-up).ResultsAt the preinterventional screening, 178 (7.7%) uremic patients and 2 (0.6%) staff members were HCV-viremic. Among them, 146 (83.9%) uremic patients received anti-HCV therapy (41 link-to-care; 105 on-site sofosbuvir/velpatasvir). The rates of sustained virological response (SVR12, undetectable HCV RNA 12 weeks after the end of treatment) in the full analysis set and per-protocol population were 89.5% (94/105) and 100% (86/86), respectively, in the on-site treatment group, which were comparable with the rates of 92.7% (38/41) and 100% (38/38), respectively, in the link-to-care group. Eventually, the HCV-viremic rate decreased to 0.9% (18/1,953), yielding an 88.3% reduction from baseline. HCV microelimination and ‘No-C HD’ were achieved in 92.3% (12/13) and 38.9% (7/18) of the haemodialysis centres, respectively.ConclusionOutreach strategies with mass screenings and on-site group treatment greatly facilitated HCV microelimination in the haemodialysis population.ClinicalTrials.gov identifierNCT03803410 and NCT03891550

Hepatitis C virus (HCV) infections can cause permanent liver-related diseases, including hepatocellular carcinoma (HCC). Low mortality and incidence of HCC have been observed in patients with chronic hepatitis C undergoing direct-acting antiviral therapy. Tumor suppressive let-7 family members are down-regulated in HCC. The present study, therefore, aimed to investigate whether expression levels for the full spectrum of let-7 family members (let-7a, 7b, 7c, 7d, 7e, 7f, 7g, 7i, and miR-98) in the circulatory system are useful as surveillance biomarkers for liver-related diseases to monitor treatment efficacy during HCV infection. To this end, we measured the levels of mature circulating let-7 family members using quantitative reverse transcription-PCR in 236 patients with HCV infection, and 147 age- and sex-matched controls. Using hierarchical cluster analysis and principal component analysis, three clusters were obtained after measuring expression levels of let-7 family members in the patients and controls. Cluster 1 included let-7a/d/e/g, Cluster 2 comprised let-7b and let-7i, and Cluster 3 comprised let-7c/f/miR-98. Let-7b/c/g represented the three clusters and showed the best survival response to liver cancer when analyzed with respect to patient data. Therefore, considering the circulating levels of let7 b/c/g as representatives of the let-7 family may facilitate effective monitoring of liver-related disease.

On-chip training of large-scale deep neural networks (DNNs) is challenging due to computational complexity and resource limitation. Compute-in-memory (CIM) architecture exploits the analog computation inside the memory array to speed up the vectormatrix multiplication (VMM) and alleviate the memory bottleneck. However, existing CIM prototype chips, in particular, SRAM-based accelerators target at implementing low-precision inference engine only. In this work, we propose a two-way SRAM array design that could perform bi-directional in-memory VMM with minimum hardware overhead. A novel solution of signed number multiplication is also proposed to handle the negative input in backpropagation. We taped-out and validated proposed two-way SRAM array design in TSMC 28nm process. Based on the silicon measurement data on CIM macro, we explore the hardware performance for the entire architecture for DNN on-chip training. The experimental data shows that proposed accelerator can achieve energy efficiency of ~3.2 TOPS/W, >1000 FPS and >300 FPS for ResNet and DenseNet training on ImageNet, respectively.

Background: The biochemical response is a crucial indicator of prognosis in chronic hepatitis B (CHB) patients treated with nucleotide/nucleoside analogues (NAs). The impact of hepatitis D virus (HDV) infection on alanine aminotransferase normalization is elusive. Methods: The longitudinal study recruited 1185 CHB patients who received NAs. These patients were tested for anti-HDV antibody and HDV RNA at the initiation of anti-hepatitis B virus (HBV) therapy and annually for patients who were HDV-seropositive. ALT levels were examined at the first and second year of anti-HBV therapy. ALT abnormality was defined as ALT levels above 40 IU/mL in both male and female, and the risk factors associated with ALT abnormality were analysed. Results: Rates of seropositivity for anti-HDV and HDV RNA were 2.0% and 0.8% among 1185 NA-treated CHB patients, respectively. The strongest factor associated with ALT abnormality (>40 IU/mL) after first year treatment with NAs was HDV RNA seropositivity at year 1 (odds ratio [OR]/95% confidence interval [CI]: 31.44/3.49–283.56, P = 0.002), followed by liver cirrhosis (2.18/1.51–3.15, P

Background We aimed to investigate the long-term outcomes in hepatitis B (HBV)/hepatitis C virus (HCV) dual-infected patients after anti-HCV therapy. Methods A total of 192 HBV/HCV dual-infected patients who had received pegylated interferon treatment were recruited. The investigation outcomes included HBV DNA ≥2000 IU/mL, with or without alanine aminotransferase (ALT) ≥2-fold the upper limit of normal, and hepatitis B surface antigen (HBsAg) seroclearance. Results Four (2.1%) patients developed early HBV reactivation before the end of treatment. Fifty (26.6%) of the remaining patients had an episode of HBV DNA ≥2000 IU/mL in a mean follow-up of 68.8 months. The risk was 4.6 per 100 person years. Only 19 (10.1%) patients developed concomitant ALT flare with oral HBV antiviral therapy; the risk was 1.7 per 100 person years. Despite HBV flare, 67 (34.9%) patients had a favorable outcome of HBsAg seroclearance. The probability was 5.7 per 100 person years. A pretreatment HBV DNA level of 300 IU/mL served as an independent predictor for all the outcomes. The combined pretreatment HBV DNA level and HCV response further enhanced the prediction of HBV flare and HBsAg seroclearance. Conclusions A pretreatment HBV DNA level of 300 IU/mL predicts HBV flare and HBsAg seroclearance after anti-HCV therapy.

Background: A direct-acting antiviral (DAA) era in hepatitis C virus (HCV) treatment is fast approaching; unfortunately, the availability and affordability of DAAs in Asia–Pacific areas vary, making it difficult to develop universal HCV practice guidelines appropriate for the all Asian populations. This study aimed to evaluate the real-world cost-effectiveness of IFN-based therapy according to the current strategies with PegIFN/RBV for “easy-to-treat” to provide a reference for application of future DAA development for IFN-eligible, treatment naïve HCV patients. Methods: A total of 1032 chronic hepatitis C treatment-naïve patients who corresponded to response-guided therapy (RGT) guidelines of PegIFN/RBV regimens were linked to the entire population of expenditures and order in the National Health Insurance Research Database of Taiwan. The average total cost per SVR achieved was calculated as the summation of the total cost for all treated patients/number of SVR cases. Results: Current RGT suggested 24 weeks of PegIFN/RBV for G1 naïve patients with baseline LVL and RVR at treatment week 4 achieved an average treatment cost per SVR of $5090 ± 2400. This was of superior cost-effectiveness compared with those other subgroups of G1 patients. In terms of G2 patients, according to current RGT of 16 weeks of treatment duration, PegIFN/RBV treatment with RVR achieved was of a very competitive cost per SVR ($3237 ± 488). Conclusion: For a naïve patient in the new DAA era, the PegIFN/RBV treatment might be conserved for those with all favorable risk parameters, considering the treatment duration and cost per SVR, in the resource-constrained countries. Keywords: Chronic hepatitis C, Cost-effectiveness analysis, Naïve, Pegylated interferon, Response-guided therapy

Background/purpose: The role of directly-acting antivirals (DAA)-containing regimens in the treatment of patients dually-infected with hepatitis C virus (HCV) and hepatitis B virus (HBV) remains unclear. The pilot study aimed to explore the safety and efficacy of a protease inhibitor, boceprevir, in combination with peginterferon/ribavirin for HCV genotype 1 (HCV-1)/HBV dually-infected patients refractory to prior peginterferon/ribavirin. Methods: Twelve peginterferon-experienced patients dually-infected with HCV-1/HBV were assigned to receive boceprevir 800 mg, twice a day, plus peginterferon-α 2b 1.5 μg/kg/week and ribavirin 800-1400 mg/day for 36 or 48 weeks. The primary endpoint was HCV sustained virological response (SVR, HCV RNA undetectable 24 weeks after end-of-treatment). Results: Five patients terminated treatment early due to adverse events (one at week 4, one at week 46), virological failures (one non-response and one breakthrough), and patient request (n = 1). Eight patients achieved HCV SVR (66.7% in full-analysis set and 72.7% in modified intention-to-treat population). The HCV SVR rate was 71.4% (5/7) in prior relapsers, 60.0% (3/5) in prior null responder; 75% in non-cirrhotic and 50% in cirrhotic patients. All four patients of prior non-cirrhotic relapsers received 36-week regimen and achieved HCV SVR. There was no HBV-related hepatic flare. All patients experienced at least one adverse event. Two had serious adverse events. Conclusion: Boceprevir plus peginterferon/ribavirin is effective in the treatment of HCV-1/HBV dually infected patients' refractory to prior peginterferon/ribavirin combination therapy. Keywords: Boceprevir, HBV, HCV, Taiwan

// Chia-Yen Dai 1, 2, 3, 4, 5, 6, * , Yi-Shan Tsai 1, * , Wen-Wen Chou 1 , Tawei Liu 1 , Chung-Feng Huang 1, 2, 4 , Shu-Chi Wang 3 , Pei-Chien Tsai 1 , Ming-Lun Yeh 1, 4, 5 , Ming-Yen Hsieh 1 , Ching-I Huang 1, 5 , Shang-Yin Vanson Liu 7 , Jee-Fu Huang 1, 4 , Wan-Long Chuang 1, 4, 6 and Ming-Lung Yu 1, 2, 3, 4, 5, 6, 8 1 Hepatobiliary Division, Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan 2 Department of Occupational and Environmental Medicine, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan 3 Health Management Center, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan 4 Faculty of Internal Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan 5 Graduate Institute of Clinical Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan 6 Center for Infectious Disease and Cancer Research, Kaohsiung Medical University, Kaohsiung, Taiwan 7 Department of Marine Biotechnology and Resources, National Sun Yat-Sen University, Kaohsiung, Taiwan 8 Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung, Taiwan * These authors contributed equally to this work Correspondence to: Ming-Lung Yu, email: fish6069@gmail.com Keywords: miR-125b; HCV; IL-6; STAT3; HCV Received: July 28, 2017 Accepted: December 01, 2017 Published: January 10, 2018 ABSTRACT Background/Aims: MicroRNA-125b (miR-125b) has been found to regulate inflammation and acts as an oncogene in many cancers. The mechanisms of miR-125b expression during hepatitis C virus (HCV) infection remain to be clarified. The present study aims to identify the factors that might regulate miR-125b expression in HCV infection. Results: High expression of miR-125b was found to correlate with HCV infection in replicon cells and in sera from HCV-infected patients, whereas the miR-125b inhibitor reduced HCV gene expression. The interleukin 6 (IL-6)/signal transducer and activator of transcription 3 (STAT3) pathway plays an inducible effect on miR-125b gene expression. STAT3 siRNA or inhibitor could reduce HCV replication. Materials and Methods: HCV replicon cells Con1 (type 1b) and Huh7/Ava5 (type 1b) were treated with 17-hydroxy-jolkinolide B (HJB) or STAT3 siRNA. Cell viability assay and Renilla Luciferase Assay were used. Fragments of the miR-125b-1 promoter were constructed for the luciferase reporter assay. PSMB8, PSMB9, miR-125b-1, and miR-125b-2 expression was determined using TaqMan ® Gene Expression Assays. Western blot analysis was performed to assess protein abundance. Conclusions: This study elucidates a novel pathway for miR-125b in the pathogenesis of chronic HCV infection and suggests it as a possible target for treating HCV infection.

Background/Purpose Pegylated interferon (PegIFN) plus ribavirin (RBV) combination therapy has been the standard of care since 2002. Although a better viral response has been achieved among chronic hepatitis C (CHC) patients in Taiwan, approximately 25% of hepatitis C virus (HCV) genotype 1 (G1) patients and 15% of G2 patients failed to achieve a sustained virological response (SVR) at the first therapy. The actual cost-effectiveness of the retreatment remains elusive. The present study conducted a real-world cost-effectiveness analysis of a large cohort among different pre-specified subgroups of treatment-experienced CHC patients. Methods A total of 117 patients with CHC who failed to achieve SVR at the first IFN-based therapy and received a second IFN-based therapy were enrolled. The inpatient and outpatient costs were acquired from National Health Insurance Research Database of Taiwan. The related medical care costs per treatment and per SVR were calculated. Results We demonstrated that the average cost per SVR achieved was $13,722 in treatment-experienced CHC patients. Especially, patients with HCV G1 infection, baseline viral loads > 400,000 IU/mL, advanced hepatic fibrosis, not achieving a rapid viral response at week 4 or complete early viral response at week 12, had poorer cost-effectiveness for PegIFN/RBV retherapy, ranging from around $15,520 to as high as $72,546 per SVR achieved. Conclusion In the current study, we explored the real-world cost-effectiveness data of PegIFN/RBV for different subgroups of treatment-experienced HCV patients. These findings provide information for policy-makers for making decisions on treatment strategies of costly direct-acting antiviral agents for retreating CHC patients.

Background and Aim Hepatitis B virus (HBV) may reactivate when treating chronic hepatitis C (CHC) with direct acting antivirals (DAA). We aim to investigate the risk of HBV reactivation during DAA therapy. Methods Chronic hepatitis C patients receiving pan-oral DAA therapy from December 2013 to August 2016 were evaluated. Fifty-seven patients that had a past HBV infection (negative hepatitis B surface antigen [HBsAg] and positive hepatitis B core antibody) and seven patients that had a current HBV infection (positive HBsAg) were enrolled. Serum HBV and hepatitis C virus (HCV) markers were regularly measured. The endpoints were the HCV sustained virological response (SVR) and the HBV virological/clinical reactivation. Results The overall SVR12 rate was 96.9%, and two patients, one with positive HBsAg, had a relapse of HCV. No episodes of HBV virological reactivation were observed among the patients with a past HBV infection. For the seven patients with a current HBV infection, HBV virological reactivation was found in four (57.1%) of the seven patients. Clinical reactivation of HBV was observed in one patient with pretreatment detectable HBV DNA and recovered after entecavir administration. For the other three patients with HBV virological reactivation, the reappearance of low level HBV DNA without clinical reactivation was observed. HBsAg levels demonstrated only small fluctuations in all the patients. Conclusions There was a minimal impact of hepatitis B core antibody seropositivity on HCV efficacy and safety. For CHC patients with current HBV infection, the risk of HBV reactivation was present, and monitoring the HBV DNA level during therapy is warranted.

Dynamic partial reconfigurable systems with a processor and a field-programmable gate array are promising innovations for meeting the requirement of mobile embedded devices. These systems demonstrate low power consumption and high performance. However, with limited battery life and chip size, energy saving and area allocation are critical concerns for such systems. In this study, an energy-aware hardware/software partition is presented to minimize the system energy consumption, and a contention-aware task allocation is presented to minimize the area requirement and response time. The energy efficiency and schedulability of the proposed methodology were evaluated using a series of workloads, and impressive results were obtained.

Many Al edge devices require local intelligence to achieve fast computing time (t AC ), high energy efficiency (EF), and privacy. The transfer-learning approach is a popular solution for Al edge chips, wherein data used to re-train the Al in the cloud is used to fine-tune (re-train) a few of the neural layers in edge devices. This enables the dynamic incorporation of data from in-situ environments or private information. Computing-in-memory (CIM) is a promising approach to improve EF for Al edge chips, existing CIM schemes support inference [1]–[5] with forward (FWD) propagation; however, they do not support training, requiring both FWD and backward (BWD) propagation, due to differences in weight-access flow for FWD and BWD propagation. As Fig. 15.2.1 shows, efforts to increase the precision of the input (IN), weight (W), and/or output (OUT) tend to degrade r AC and EF for training operations irrespective of scheme: digital FWD and BWD (DF-DB) or CIM-FWD-digital-BWD (CiMF-DB). This work develops a two-way transpose (TWT) SRAM-CIM macro supporting multibit MAC operations for FWD and BWD propagation with fast r AC and high EF within a compact area. The proposed scheme features (1) A TWT multiply cell (TWT-MC) with a high resistance to process variation; and (2) a small-offset gain-enhancement sense amplifier (SOGE-SA) to tolerate a small read margin. A 28nm 64Kb TWT SRAM-CIM macro was fabricated using a foundry-provided compact 6T-SRAM cell for SRAM-CIM devices supporting both inference and training operations for the first time. This macro also demonstrates the fastest t AC (3.8 – 21ns) and highest EF (7 – 61.1TOPS/w) for MAC operations using 2 – 8b inputs, 4 – 8b weights and 12 − 20b outputs.

Background: Medial open wedge high tibial osteotomy (MOWHTO) is a well-established treatment for osteoarthritis of the medial tibiofemoral compartment. Surgical site infection (SSI) after MOWHTO is a devastating complication that may require further surgery. In this study, we aimed to identify the risk factors for infection after MOWHTO over 1 to 4 years of follow-up. Methods: Fifty-nine patients who underwent MOWHTO combined with knee arthroscopic surgery were included in this prospective study. Artificial bone grafts were used in all cases. Possible risk factors, including sex, age, body mass index (BMI), underlying disease, hospitalization length, correction angle, and surgery time, were recorded. Both univariate and multivariate analysis were used. Results: A total of 59 patients who underwent 61 operations were included. Eleven patients (18.0%) were reported to have SSI. Univariate analysis showed that smoking and diabetes mellitus were positively associated with SSI. Multivariate analysis showed that smoking and age were positively associated with SSI. Three patients (4.9%) were reported to suffer from deep SSI, requiring surgical debridement, all of whom were male smokers. Conclusion: Smoking, diabetes mellitus, and old age were identified to be possible risk factors of SSI after MOWHTO. These findings are common risk factors of SSI after orthopedic surgery according to the literature. Patient selection should be performed cautiously, and postoperative prognosis for MOWHTO should be carefully explained to patients who smoke.

Treatments with pegylated interferon/ribavirin (PEG-IFN/RBV) has been standard-of-care in patients with chronic hepatitis C virus (HCV) (CHC) infection and reimbursed in Taiwan. However, the actual cost-effectiveness remains unclear. We aimed to evaluate a real-world cost-effectiveness for CHC patients treated with PEG-IFN/RBV by using a clinical cohort with linkage to the National Health Insurance Research Database of Taiwan. The total and itemized medical-care expenses of outpatient visits of 117 treatment-naïve CHC patients with linkage to the two million sampling of the National Health Insurance Research Database were collected. Four components of costs were assessed, including antiviral agents, nonantiviral agents, laboratory testing and consultation costs. The cost per sustained virological response (SVR) achieved was calculated to evaluate the cost-effectiveness. The average cost per treatment in 117 naïve Taiwanese CHC patients was $4620. With an overall SVR rate of 78.6%, the average cost per SVR was $5878. The average medical-care cost per treatment for 52 Genotype 1 (G1) patients was $5133, including $4420 for antivirals, $380 for nonantivirals, $302 for laboratory, and $78 for consultation, compared to $4209, $3635, $317, $233, and $56 for 65 Genotype 2 (G2) patients. With an SVR rate at 67.3% for G1 and 87.7% for G2 patients, the cost per SVR achieved was significantly higher in G1 patients than those in G2 patients ($7627 vs. $4799, p = 0.001). In the current study, we provided the real-world cost-effectiveness of PEG-IFN/RBV for treatment-naïve CHC patients. The genotype-specific cost-effectiveness could enhance decision-making for policy-makers in the coming era of directly acting antiviral therapy.

Background and aimsChronic hepatitis B patients in Taiwan with no or limited liver injury are not reimbursed for antiviral treatment by the Taiwan National Health Insurance (NHI). Innovative fibrosis marker, age-adjusted Fibrosis-4 Index (FIB4-AA), was implemented to evaluate the tendency of liver fibrosis in these patients.MethodsThe FIB-4 indices of 256 antiviral treatment-naïve chronic hepatitis B patients at Kaohsiung Medical University Hospital from 2003 to 2019 were reviewed. The difference in initial FIB-4 and last FIB4-AA was treated as a categorical variable, representing the tendency of liver fibrosis in each individual aside from ageing. Logistic regression was implemented to evaluate the three parameters most dependent on increment of FIB4-AA: e seroconversion, body mass index (BMI) and initial FIB-4 index.ResultsThe yearly FIB-4 growth rate of an individual without chronic hepatitis was lower than that of the study group (0.0237 vs 0.0273 for males, 0.02 vs 0.0288 for females). Patients undergoing or completing e seroconversion were less prone to increment of FIB4-AA (p=0.036, OR 0.524). Logistic regression revealed that BMI ≥25 kg/m2 significantly less increment of FIB4-AA (p=0.001, OR 0.383, 95% CI 0.212 to 0.690), while patients with initial FIB-4 ConclusionChronic hepatitis B patients not meeting the reimbursement criteria of the Taiwan NHI are prone to increment of liver fibrosis marker. Overweight is associated with less increment of fibrosis marker, while initial FIB-4

Hepatitis B virus surface antigen (HBsAg) levels reflect disease status and can predict the clinical response to antiviral treatment; however, the emergence of HBsAg mutant strains has become a challenge. The Abbott HBsAg quantification assay provides enhanced detection of HBsAg and HBsAg mutants. We aimed to evaluate the performance of the Abbott HBsAg quantification assay with automatic sample dilutions (shortened as automatic Architect assay), compared with the Abbott HBsAg quantification assay with manual sample dilutions (shortened as manual Architect assay) and the Roche HBsAg quantification assay with automatic sample dilutions (shortened as Elecsys). A total of 130 sera samples obtained from 87 hepatitis B virus (HBV)-infected patients were collected to assess the correlation between the automatic and manual Architect assays. Among the 87 patients, 41 provided 42 sera samples to confirm the linearity and reproducibility of the automatic Architect assay, and find out the correlation among the Elecsys and two Architect assays. The coefficients of variation (0.44–9.53%) and R2 = 0.996–1, which were both determined using values obtained from the automatic Architect assay, showed good reproducibility and linearity. Results of the two Architect assays demonstrated a feasible correlation (n = 130 samples; R = 0.898, p 0.93 in all cases). In conclusion, the correlation between the automatic and manual dilution Architect assays was feasible, particularly in the HBeAg-negative and low DNA groups. With lower labor costs and less human error than the manual version, the Abbott automatic dilution Architect assay provided a good clinical performance with regard to the HBsAg levels.

BACKGROUND Weekly training volumes for triathlete are typically higher and may cause fatigue and musculoskeletal injury risk. High-intensity interval training (HIT) is a potent time-efficient strategy to induce adaptations normally associated with traditional endurance training. Therefore, the aim of this study was to investigate the effect of two weeks of in-season HIT on exercise capacity and hormonal responses in young triathletes. METHODS Twelve adolescent triathletes performed 18 sessions of HIT over 2 weeks including swim, cycle, and run events. The 6-day training blocks were separated by 1 day of recovery. Pre- and post-training, peak oxygen uptake (VO2peak) and exercise performance were assessed, and blood samples were obtained to detect changes in hormone and metabolite levels. RESULTS VO2peak was significantly higher (P=0.02) post-training (56.4±8.1 mL·min-1·kg-1) versus pre-training (55.1±7.5 mL·min-1·kg-1). Mean power and total work during 6×10 s repeated-sprint tests significantly increased (P=0.03) after HIT. Additionally, 750 m swim time (pre- vs. post: 689.7±102.5 s vs. 662.0±75 s, P=0.01) and 20 km cycling time (pre- vs. post: 1856.6±274.8 s vs. 1705.4±266.8 s, P=0.02) were significantly lower post-training compared to pre-training, but there was no significant difference in 5 km run time after HIT (pre- vs. post: 1315.8±81.3 s vs. 1292.0±112.9 s, P=0.31). In contrast to pre-training, ammonia concentration was significantly increased (P

A continuous nearest neighbor (CNN) query retrieves the nearest neighbor of every point on a line segment and indicates its valid segments. Zheng et al. have proposed a Hilbert-curve index for the CNN query. This method contains two phases, searching candidates in the approximate search range, and filtering the candidates to get the final answer. However, it may determine a wide search range in the first phase based on this method, resulting in the decrease of the accuracy and the increase of the processing time. Therefore, in this paper, to avoid this disadvantage, we propose a forward moving method to efficiently support the CNN queries. The proposed method locally expands the search range along the query line segment to find the neighbors. Experimental results show that our method outperforms Zheng et al.’s method in terms of the accuracy and the processing time.

A new tunable multiwavelength fiber laser is proposed. Such a laser contains a homogeneous and inhomogeneous broadening media, i.e., a Fabry–Perot laser diode and an erbium-doped fiber amplifier, in the laser cavity. The Fabry–Perot laser diode is used to obtain tunable multiwavelength lasing. By adjusting the injection current of the Fabry–Perot laser diode, emission at a single wavelength, dual wavelengths, triple wavelengths or quadruple wavelengths are obtained. The lasing wavelength is tuned by bending a section of fiber in the laser cavity. The tuning ranges for single wavelength lasing and dual wavelength lasing are 20 nm and 10 nm, respectively, while those for triple wavelength lasing and quadruple wavelength lasing are 7 nm and 3 nm, respectively.

A new method is proposed and demonstrated to fabricate planar thin-film transistors and trigated nanowire (NW) devices simultaneously on the same panel. By using an oxide-nitride-oxide stack as the gate dielectric, the NW devices could also serve as nonvolatile Si-oxide-nitride-oxide-Si (SONOS) memory devices. Our results indicate that the combination of trigate and NW channels help to improve the device performance in terms of steppers subthreshold swing and reduced threshold voltage. Improvement in programming and erasing efficiency of the nonvolatile SONOS memory devices is also demonstrated with the trigated NW structure.

Several types of poly-Si nanowire (NW) thin-film transistors (TFTs) with multiple-gated (MG) configuration were demonstrated and characterized. These devices were fabricated with simple methods without resorting to costly lithographic tools and processes. The fabricated trigated devices show a low subthreshold swing (SS) of around 100 mV/dec and on/off current ratio higher than 108. These results clearly indicate the effectiveness of MG scheme in enhancing the device performance. Furthermore, a multiple-channel scheme was demonstrated to further increase the drive current without compromising device performance. Finally, the impact of MG on the variation of NWTFT characteristics is investigated with a clever method that allows the fabrication of test structures with identical NW channel but different gate configurations. The results clearly show that the variation could be reduced by increasing the portion of NW channel surface that is modulated by the gate.

Background: For decades, peginterferon and ribavirin (PegIFN/RBV) have been the standard-of-care for chronic hepatitis C virus (CHC) infection. However, the actual cost-effectiveness of this therapy remains unclear. We purposed to explore the real-world cost effectiveness for subgroups of treatment-naïve CHC patients with PegIFN/RBV therapy in a large real-world cohort using a whole population database. Methods: A total of 1809 treatment-naïve chronic hepatitis C virus (HCV) patients (829 HCV genotype 1 [G1] and 980 HCV G2) treated with PegIFN/RBV therapies were linked to the National Health Insurance Research Database, covering the entire population of Taiwan from 1998 to 2013 to collect the total medical-care expenses of outpatient (antiviral agents, nonantiviral agents, laboratory, and consultation costs) and inpatient (medication, logistic, laboratory, and intervention costs) visits. The costs per treatment and the cost per sustained virological response (SVR) achieved were calculated. Results: The average medical-care cost was USD $4823 (±$2984) per treatment and $6105 (±$3778) per SVR achieved. With SVR rates of 68.6% and 87.8%, the cost/SVR was significantly higher in G1 than those in G2 patients, respectively ($8285 vs $4663, P

Si nanowire (NW) SONOS devices have recently been demonstrated as a good candidate for high-density non-volatile memory application [1][2]. Owing to the high surface-to-volume ratio of the NW channel, the programming and erasing (P/E) operation of the device could be performed at a lower voltage and much faster speed over the planar counterpart [2]. However, the fabrication of NW devices typically requires advanced lithographic tools and/or complicated process flow. These are not compatible with the manufacturing of flat-panel products where the device feature size is generally several microns or larger. In this work, we propose a simple and cost-effective approach to integrate planar poly-Si thin-film transistors (TFTs) and tri-gated poly-Si NW SONOS devices without resorting to advanced lithographic tools. Greatly enhanced P/E speed with the use of NW structure is clearly demonstrated.

Recently, various features for content-based image retrieval (CBIR) have been proposed, such as texture, color, shape, and spatial features. In this paper we propose a new feature, called orientation-distance histogram for CBIR. Firstly, we transform the RGB color model of a given image to the HSI color model and detect edge points by using the H-vector information. Secondly, we evaluate the orientation-distance histogram from the edge points to form a feature vector. After normalization of feature, our proposed method can cope with most problems of variations in image. Finally, we show some results of query for real life images with the precision and recall rates to measure the performance. The experimental results show that the proposed retrieval method is efficient and effective