Search

There are currently few large-scale studies describing the clinical characteristics of these histological variants of prostate cancer (PCa) and the prognostic results of these patients. Thus, we used the Surveillance, Epidemiology, and End Results (SEER) database to achieve a more comprehensive understanding of the current prevalence and prognosis of these histological variants. A total of 780 (3.3%) patients with histological variants and 235834 patients with typical adenocarcinoma were included in the present study. The variants group presented with a more advanced stage than the typical adenocarcinoma group. The results also showed that patients with neuroendocrine (NE) PCa were more likely to have visceral metastases. In prognostic comparison, we found that variants, except for mucinous, had a worse prognosis than typical adenocarcinoma. Moreover, among these variants, we also discovered that NEPCa patients with low prostate-specific antigen level had the worst survival. In contrast, mucinous PCa had similar overall survival to typical adenocarcinoma. PATIENT SUMMARY: We reported follow-up results for prostate cancer (PCa) patients from the Surveillance, Epidemiology, and End Results database (2004-2016). We focused on the prognosis of PCa with different histological variants and evaluated the risk factors that were significant in predicting poor prognosis. We suggested that neuroendocrine PCa patients should receive active treatment as soon as possible to improve their prognosis. We still need further studies to identify the role of mucinous PCa in patients' prognosis compared with typical adenocarcinoma.

To study the effect of inflammatory markers in blood such as interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α) on the Gleason score (GS) changes in patients with prostate cancer (PCa) after radical prostatectomy (RP), we conducted this study.From November 2012 to September 2021, a total of 237 patients underwent RP at our institution. Blood samples from all patients were collected within 1 week before surgery. Preoperative clinical characteristics include age, serum IL-6 and TNF-α, neutrophil-to-lymphocyte ratio, C-reactive protein, the platelet-to-lymphocyte ratio, lymphocyte-to-monocyte ratio, systemic immune-inflammation index, the prostate imaging reporting and data system (PI-RADS) score, prostate-specific antigen, and biopsy GS were assessed. Univariate and multivariate logistic regression analyzes were used to determine the risk factors of GS changes after RP. The efficiency of this prediction model was identified with the area under the curve of the receiver operating characteristic curve.Seventy-three patients (30.8%) had GS upgraded in the overall cohort, and 55 patients (23.2%) had GS downgraded. In comparing PCa patients with and without GS upgraded, multivariate logistic regression analysis showed that serum TNF-α (odds ratio [OR]: 2.518, p = 0.019) and IL-6 (OR: 0.478, p = 0.023) were independent factors predicting the occurrence of GS upgrade. We also compared the characteristics of patients with GS upgraded and GS downgraded; multivariate logistic regression analysis also demonstrated significant differences in serum IL-6 and TNF-α between these two groups (all p 0.05). In addition, we found that low prostate volume and biopsy GS ≥ 7 were significantly associated with higher PI-RADS sores in multivariate analysis.The high expression of serum TNF-α level is positively correlated with GS upgraded in PCa patients. High expression of serum IL-6 level is negatively correlated with GS upgraded in PCa patients and positively related with GS downgraded.

Prostate biopsy is the gold standard for the diagnosis of prostate cancer. However, not all patients are suitable for prostate biopsy. For example, some patients have anal stenosis, some patients are too old to withstand the pain caused by puncture, patients who are unwilling to undergo prostate biopsy. We found that there was currently no literature report on a specific solution to this problem. This is the first report of a laparoscopic radical prostatectomy (LRP) in a pituitary dwarfism who didn’t have a prostate biopsy before LRP due to anal stenosis. And this report added a new method to diagnose prostate cancer. We present a case of a 61-year-old pituitary dwarfism who had a prostate specific antigen (PSA) of 32.13 ng/mL by physical examination and didn’t perform prostate biopsy due to anal stenosis. Preoperative prostate MRI suggests a low-signal mass on the left side of the prostate and (68)Ga PSMA-11 PET/CT demonstrated that Abnormally high PSMA and CHO uptake on the left side of the prostate. Therefore, combined with the patient’s PSA, MRI and (68)Ga PSMA-11 PET/CT, our clinical diagnosis was prostate cancer. Surgery was difficult due to narrow pelvic space, but achievable through LRP. Histological analysis revealed multifocal prostate cancer, with negative surgical margins and no extraprostatic extension. Postoperative patient had no serious complications and was discharged. Based on this case, For the first time, we proposed to make full use of the results of clinical tests and imaging examinations for the diagnosis and treatment of diseases without prostate biopsy.

Objective We sought the optimal preprocessing method to improve the long-term transfer efficiency of adeno-associated viral (AAV) vectors of rat bladder urothelium. Methods Rat bladders were pretreated by eight different chemical reagents including PBS, trypsin, poly-L-lysine (PLL), protamine sulfate (PS), sodium dodecyl sulfate (SDS), n-dodecyl-β-d-maltoside (DDM), HCl or NaOH followed by transurethral instillation with AAV2/9-CMV-zsgreen. The transfection efficiency was evaluated by the percentage of urothelium with green fluorescence after 4 weeks. Results We employed eight different chemical components and eleven pretreatment methods. The rat bladder was harvested and evaluated without any mortality at 4 weeks post-infection. Compared with other treatment groups, the rate of urothelium transfection is comparatively high in the groups of rats treated with HCl or NaOH. In addition, rats treated with 0.1N NaOH 4 min showed significantly highest transfection efficiency (about 3%). The histopathologic results showed regular morphology of the bladder treated with HCl or NAOH and no significant damage was observed. Conclusion Pretreatment with 0.1N NaOH 4 min could improve the transfection efficiency of AAV in bladder urothelium and maintain the transfection efficiency for a long time without pathological damage. Importance We found that pretreatment of the bladder with HCl and NaOH can improve the transfection efficiency of AAV for a long time compared to control group. In addition, rats treated with 0.1N NaOH 4 min could have the most significant effect among these strategies, which was discovered for the first time. This could be valuable for future bladder gene therapy.

Background To investigate the effect of serum aspartate transaminase/alanine transaminase (AST/ALT) on the risk of prostate cancer. Methods A total of 404 patients undergoing prostate biopsy from April 2016 to July 2019 were enrolled. One hundred and ninety‐four patients with prostatic cancer (PCa) were diagnosed by pathology. Two hundred and ten patients were diagnosed with benign prostatic hyperplasia (BPH). Multivariate logistic regression was used to analyze the effect of AST/ALT ratio and other factors on the incidence of PCa. Result AST/ALT ratio was significantly higher in PCa than in BPH patients (OR 2.313, 95%CI 1.337‐4.003, P = .002). ROC curve indicated that the best cutoff was 1.155 in predicting the incidence risk of PCa. The age of PCa patients is higher than BPH patients (OR 1.054, 95%CI 1.027‐1.082, P, To investigate the effect of serum aspartate transaminase/alanine transaminase (AST/ALT) on the risk of prostate cancer. AST/ALT ratio was an independent risk factor in predicting the incidence of PCa.

The relationship between obstructive sleep apnea (OSA) and the risk of nocturia remains unclear. Therefore, we sought to identify whether or not OSA affects the incidence of nocturia. A thorough literature search was executed in September 1st 2018 from PubMed, Web of Science database, and Embase. We used DerSimonian and Laird random-effects to calculate the pooled relative ratio (RR). Total of 13 studies met inclusion criteria and in total comprised, 406 patients and 9518 controls. There was a significant association between OSA and the risk of nocturia (RR = 1.41, 95% CI 1.26–1.59). Through subgroup analysis by different severity of OSA, we found patients who had severe OSA were at high risk of nocturia. Through another subgroup analysis, we found a statistically significant association between OSA and risk of nocturia in the men (RR = 1.487, 95% CI 1.087–2.034, P = 0.013). However, there was no significant relationship between OSA and nocturia in the women (RR = 1.537, 95% CI 0.831–2.842, P > 0.05). Subgroup analysis of different diagnostic methods indicated that OSA was significantly associated with the risk of nocturia regardless what method was used to diagnose OSA (P

Background Currently, clinical studies on the prognosis of prostate cancer (PC) taking aspirin were developing, but the precise mechanism of aspirin on tumor cells was still unclear. In addition, the conclusion that aspirin can improve the prognosis of PC patients continues to be controversial. Therefore, we collected comprehensive literatures and performed our study to explore the prognostic effect of aspirin on PC. Methods A comprehensive literature search was performed in April 2019 based on PUBMED. EMBASE. Hazard Ratio (HR) as well as its 95% confidence interval (CIs) for prostate cancer specific mortality (PCSM) was extracted from eligible studies. Result A total of 10 eligible articles were used in our study. The pooled results showed that PC patients who used aspirin or taking aspirin did not have lower PCSM than those who had not used (HR =0.89, 95% CI: 0.73–1.08, P>0.05). In subgroup analysis, we found that taking aspirin before diagnosis of prostate cancer and taking aspirin after diagnosis of prostate cancer did not have significant association with PCSM. (pre-diagnostic use, HR = 0.88, 95% CI: 0.72–1.06; post-diagnosis use, HR = 0.88, 95% CI: 0.67–1.17). In addition, we found no significant association between aspirin use or its duration and the risk of PCSM. Another important result demonstrated that aspirin use was not associated with risk of PSCM in either high risk (T ≥ 3 and/or Gleason score ≥ 8) or low risk PC patients(low-risk PC, HR = 1.05, 95% CI: 0.81–1.35; high-risk PC, HR = 0.97, 95% CI: 0.75–1.24). Conclusion Our results demonstrated that there was no significant association between aspirin use and the risk of PCSM. At the same time, the dosage and duration of aspirin use had no statistical influence on the risk of PCSM in high/low risk PC. Further studies are needed to confirm the findings.

Background: The aim of this article was to review the association between androgen deprivation therapy (ADT) and the risk of stroke in patients with prostate cancer (PC). Thus, we performed this study to understand the impact of ADT on the incidence of stroke in PC patients. Methods: A comprehensive literature search was performed in June 2019 based on PubMed, EMBASE, and Web of science databases. Pooled rate ratio (RR), hazard ratio (HR), and their 95% confidence intervals (95% CIs) were calculated with a DerSimonian and Laird random effects. Result: A total of 239,099 patients from 10 studies were included in this analysis. There was no significant association in pooled RR analysis. Pooled HR analysis showed that ADT treatment increased the risk of stroke (HR = 1.129, 95% CI: 1.019–1.251, p = 0.02). In a subgroup analysis of RR results, we found that different ADT treatments had no significant effect on increasing the risk of stroke. And in the subgroup analysis of HR results, only PC patients treated with gonadotropin-releasing hormone (GnRH) agonists, orchiectomy, or oral antiandrogen had significantly higher risk of stroke. In addition, we observed the result from another comparison that PC patients treated with GnRH agonists combined with oral antiandrogens might have a lower risk of stroke compared with using GnRH agonists alone. Conclusion: Our results showed that GnRH agonists, orchiectomy, or oral antiandrogen might play an important role in the incidence of stroke. We still need further studies to clarify the role of ADT in the increased risk of stroke.

To compare the accuracy of several volume and diameters modified prostate health index (mPHI) models with PHI density (PHID), PHI, and other prostate-specific antigen (PSA) derivatives in detecting PSA grey zone prostate cancer (PCa).Between August 2020 and September 2021, a consecutive cohort of 214 suspected PCa patients with elevated total PSA values ranged from 4.0 to 10.0 ng/mL were prospectively recruited and received PHI detections and transrectal ultrasonography (TRUS) measurements, followed by systematic prostate biopsies confirmation.Among the 214 patients enrolled in the project, a total of 80 were diagnosed with PCa. In univariate analysis for the training cohort, the area under curve (AUC) of mPHI-2 [Formula: see text] was 0.8310, which outperformed PHID in identifying PSA grey zone PCa (P ≤ 0.0001) and showed the best net benefit in decision curve analysis (DCA). By a threshold of 0.2835, the sensitivity and specificity in the prediction of PCa were 78.9% and 90.3%, while the positive predictive value (PPV) and negative predictive value (NPV) were 78.3% and 78.6%, respectively.According to our present single-center experience, the mPHI-2 risk predictor outperformed PHID or other classical parameters alone in the PCa detection with a grey zone PSA level in Asian males.

Although speaker verification has conventionally been an audio-only task, some practical applications provide both audio and visual streams of input. In these cases, the visual stream provides complementary information and can often be leveraged in conjunction with the acoustics of speech to improve verification performance. In this study, we explore audio-visual approaches to speaker verification, starting with standard fusion techniques to learn joint audio-visual (AV) embeddings, and then propose a novel approach to handle cross-modal verification at test time. Specifically, we investigate unimodal and concatenation based AV fusion and report the lowest AV equal error rate (EER) of 0.7% on the VoxCeleb1 dataset using our best system. As these methods lack the ability to do cross-modal verification, we introduce a multi-view model which uses a shared classifier to map audio and video into the same space. This new approach achieves 28% EER on VoxCeleb1 in the challenging testing condition of cross-modal verification.

With the maturity of cryotherapy for prostate cancer, the complications after operation are also decreasing, which can improve the prognosis of patients. However, erectile dysfunction (ED) is still one of the main complications after cryotherapy. Therefore, we performed a meta-analysis to evaluate the incidence of erectile dysfunction in patients after cryotherapy. A comprehensive literature search was performed in August 2018. PUBMED and EMBASE databases were searched to collect studies reporting the incidence rate of ED after cryotherapy from 2002 to 2018. Two reviewers independently screened the literatures, extracted data and assessed the risk of bias of included studies. Pooled ratio and its 95% confidence intervals (95% CIs) were performed by Stata 12.1. Of the 157 articles identified on August 1st 2018, 23 studies which reported ED after cold ablative therapy were identified, however, only 12 used validated outcome measures and met inclusion criteria. A total of 12 studies were included in this meta-analysis. Overall, the results of this meta-analysis showed that the pooled incidence rate of ED was 0.27 (95% CI 0.26–0.28) which means that the incidence rate of ED after cryotherapy for prostate cancer was not high, but we still found that there are great heterogeneity between the 12 articles. By subgroup analysis, we found a statistically significant incidence rate of ED in primarily localized PCa which was 0.49 (95% CI 0.30–0.68), which is clearly lower than the incidence of recurrent prostate cancer after failed primary radiotherapy 0.61 (95% CI 0.43–0.79). ED is one of the major complications after cryotherapy for PCa. Furthermore, subgroup analysis revealed a higher incidence rate in PCa undergoing radiotherapy. Significantly, with the development of cryotherapy technology, the incidence of ED after cryotherapy for prostate cancer is decreasing. While we still need further researches to advance knowledge in this field.

BACKGROUND: Positive surgical margins (PSM) is one of the most important factors affecting the prognosis of prostate cancer (PCa) patients after radical prostatectomy (RP). Although some studies have found the preoperative systematic inflammation-based scores the neutrophil-to-lymphocyte ratio (NLR), the platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR) can predict the incidence and prognosis of PCa, few studies have explored the predictive value of preoperative systematic inflammation-based scores on the PSMs for PCa patients after RP. METHODS: From June 2014 to September 2020 a total of 497 patients underwent RP at our institution. Blood samples from all patients were collected within one week before surgery. Preoperative clinical characteristics including age, body mass index (BMI), prostate-specific antigen (PSA), and biopsy Gleason sum (BGS) were assessed. Postoperatively pathological specimens were assessed for pathological Gleason sum (PGS), pathological stage, and margin status. RESULTS: In the multivariable analysis including preoperative variables, PSA and LMR were the independent predictive factors for PSM (OR: 2.817; 95% CI, 1.836–4.320, P

Background: Anogenital distance（AGD） can be used as a biomarker to indirectly indicate androgen levels during the sexual maturation of mammals. In order to gain a deeper understanding of the association between the AGD and the risk of prostate cancer(PCa), we performed this studyMaterials: A total of 107 patients undergoing perineal prostate biopsy from November 2019 to August 2020 were enrolled. All patients diagnosed 57 PCa patients and 50 benign prostatic hyperplasia(BPH) patients through perineal prostate biopsy in our hospital . The anus to the posterior base of the scrotum (AGDAS) and the cephalad insertion of the penis (AGDAP) of all patients were measured before prostate biopsy.Result: The mean age of patients with PCa was 69.46 ± 7.52 and the mean age of patients with BPH was 67.38± 8.26. We did not find a significant association between age and BMI in the risk of PCa(all p＞0.05). Besides, there was no significant association between AGD and the risk of PCa(p＞0.05). we only discovered that PSA could play an important role in the development of PCa.Conclusion: AGD may not play an important role in predicting the incidence risk of PCa. we still need a large-scale prospective study to explore the ture association between AGD and the incidence of PCa.

Purpose: At present, there is no clear relationship between prostatitis and prostate cancer(PCa). Therefore, in order to further understand the inflammation of prostate tissue and the occurrence of PCa, we conducted this study.Method: A total of 686 patients were enrolled in the study. All patients underwent prostate biopsy in our hospital. A retrospective analysis of the biopsy results was performed to assess the assocition between chronic inflammation and the risk of PCa.Result: Of the 686 patients, 354 were diagnosed with PCa, and 332 were benign. A total of 403 patients had prostate inflammation. PCa patients had lower prostate volume and transition zone volume than benign group(p＜0.001). Compared with benign group, PCa patients had lower PSA and PSA density (PSAD)(p＜0.001). We also found that the probability of inflammation in PCa patients is lower than that in the benign group (p＜0.001). In multivariate analyses, chronic inflammation was negatively correlated with the incidence of PCa（OR=0.80; p=0.015). Conclusion: Chronic inflammation in biopsy tissue might serve as a predicted factor for low incidence risk of PCa.

OBJECTIVES: The purpose of this study was to explore the predictive value of preoperative prognostic nutritional index(PNI) and systemic immune‐inflammation index(SII) for local tumor stage in bladder cancer(BC) after radical cystectomy(RC).METHODS: We researched our database between April 2011 and October 2019. There were 195 BC patients who underwent RC. The PNI and SII were calculated using preoperative blood sample results. The predictive value of SII and PNI was analysed with univariate and multivariate Cox regression models. Receiver operating characteristic (ROC) was used to determine the optimum PNI. Signifcant P was P

Purpose: To determine the potential role of several biochemical and clinical markers in predicting adverse pathology (AP) and ISUP GG upgrading at radical prostatectomy (RP) with low-grade (ISUP Gleason Group (ISUP GG) 1 and 2) prostate cancer (PCa). Methods: We retrospectively reviewed the patients who underwent radical prostatectomy following criteria: clinical stage T2a or less,and were identified low-grade PCa (ISUP GG 1−2, prostate-specifific antigen (PSA)

As speech-enabled devices such as smartphones and smart speakers become increasingly ubiquitous, there is growing interest in building automatic speech recognition (ASR) systems that can run directly on-device; end-to-end (E2E) speech recognition models such as recurrent neural network transducers and their variants have recently emerged as prime candidates for this task. Apart from being accurate and compact, such systems need to decode speech with low user-perceived latency (UPL), producing words as soon as they are spoken. This work examines the impact of various techniques - model architectures, training criteria, decoding hyperparameters, and endpointer parameters - on UPL. Our analyses suggest that measures of model size (parameters, input chunk sizes), or measures of computation (e.g., FLOPS, RTF) that reflect the model's ability to process input frames are not always strongly correlated with observed UPL. Thus, conventional algorithmic latency measurements might be inadequate in accurately capturing latency observed when models are deployed on embedded devices. Instead, we find that factors affecting token emission latency, and endpointing behavior have a larger impact on UPL. We achieve the best trade-off between latency and word error rate when performing ASR jointly with endpointing, while utilizing the recently proposed alignment regularization mechanism., Comment: Proc. of Interspeech 2021

Purpose: To determine the potential role of several biochemical and clinical markers in predicting adverse pathology (AP) and ISUP GG upgrading at radical prostatectomy (RP) with low-grade (ISUP Gleason Group (ISUP GG) 1 and 2) prostate cancer (PCa). Methods: We retrospectively reviewed the patients who underwent radical prostatectomy following criteria: clinical stage T2a or less,and were identified low-grade PCa (ISUP GG 1−2, prostate-specifific antigen (PSA)

Recent studies have found that metastatic castrated-resistant prostate cancer (mCRPC) with positive androgen receptor splice variant 7 (AR-V7) may have poor prognosis during endocrine or chemotherapy treatment, but the specific mechanism was still unclear. We had finished literature search in March 2019 from PubMed, Web of Science database, and Embase. The final results were presented in this research. The pooled results showed that AR-V7 status predicted pooled PSA-PFS (HR = 4.31, 95% CI: 2.57-7.24, p < .001), rPFS (HR = 2.39, 95% CI: 1.28-4.48, p = .006) and OS (HR = 4.27, 95% CI: 3.22-5.66, p < .001) in mCRPC patients after endocrine or chemotherapy treatment. Subgroup analysis of different treatments revealed that mCRPC patients treated with chemotherapy had significant association between positive AR-V7 and OS (HR = 2.82, 95% CI: 1.72-4.62, p < .001), and also during endocrine therapy (HR = 4.78, 95% CI: 3.33-6.86, p < .001). Our study demonstrated that AR-V7-positive mCRPC patients may have worse prognosis. AR-V7 may be an independent prognostic factor for endocrine therapy or chemotherapy in patients with mCRPC.

Until now, there is no clear conclusion on the relationship between the surgical margin status after radical prostatectomy (RP) and prostate cancer-specific mortality (PCSM). Therefore, we conducted this systematic review and meta-analysis based on all eligible case–control studies. A systematic and comprehensive literature search was performed based on PUBMED and EMBASE to identify all of the potentially relevant publications which were published before September 2019. Hazard ratio (HR) for PCSM was independently extracted by two reviewers from all eligible studies. Pooled HR estimates with their corresponding 95% confidence intervals (95% CIs) were computed by Stata12.0. Total 15 eligible studies were included in this meta-analysis. The pooled results showed that patients with positive surgical margin (PSM) after RP may have higher PCSM than those who had a negative surgical margin (HR 1.44, P = 0.043). In the subgroup analysis, we found that no matter whether the pathological stage of the patients is T2 or T3, PSM is indicative of a high PCSM and that the PCSM of T3 is higher than T2 (Pathological stage T3, HR 1.77, P = 0.032; Pathological stage T2, HR 1.56, P = 0.003). In addition, by performing the subgroup analysis of Gleason score, we concluded that both Gleason score 8–10 and Gleason score ≤ 7 would increase the risk of PCSM, and the former was more significant than the latter (Gleason score 8–10, HR 1.88, P

The specific association between PTEN deletion or ERG rearrangement and the recurrence of prostate cancer (PC) treated with radical prostatectomy (RP) or brachytherapy is still unelaborated. Therefore, we performed a comprehensive meta‑analysis to understand the impact of these factors on cancer recurrence.A comprehensive literature search was performed in November 2018 based on PUBMED, EMBASE and Web of science database. Hazard ratio (HR) for biochemical recurrence free (BRF) which was defined as a PSA greater than or equal to 0.4 ng/mL after RP or another therapy for any detectable PSA and recurrence-free survival (RFS) which defined the time from the beginning of treatment to the earliest occurrence of local recurrence, distant metastasis or death. Which were extracted from eligible studies. IA total of 6744 patients from 17 studies were included in this analysis Overall, The pooled results showed that PTEN loss predict pooled BRF (HR 1.79, 95% CI 1.49-2.16, P0.001) and RFS (HR 1.71, 95% CI 1.50-1.95, P0.001) in patients after radical prostatectomy or brachytherapy for prostate cancer. Subgroup analysis revealed that PTEN deletion significantly predicted poor BRF or RFS in heterozygous studies group (HR 1.70, 95% CI 1.31-2.21, P0.001). The PTEN deletion also significantly predicted poor BRF or PFS in homozygous studies (HR 2.54, 95% CI 1.89-3.17, P0.001). And we had found that there was no significant association between ERG rearrangement and cancer recurrence regardless of PTEN loss or not. In addition, we concluded that Gleason score6 significantly predicted the poor BRF or RFS in studies, especially in Gleason score = 4 + 3 (HR 3.16, 95% CI 2.08-4.80, P0.01).This study presented that PTEN deletion significantly reduce time of BRF or RFS, especially based on homozygous deletion. And we also found ERG rearrangement in tumor cell could not significantly predict BRF or RFS.

Attention has shown to be a pivotal development in deep learning and has been used for a multitude of multimodal learning tasks such as visual question answering and image captioning. In this work, we pinpoint the potential limitations to the design of a traditional attention model. We identify that 1) current attention mechanisms discard the latent information from intermediate reasoning, losing the positional information already captured by the attention heatmaps and 2) stacked attention, a common way to improve spatial reasoning, may have suboptimal performance because of the vanishing gradient problem. We introduce a novel attention architecture to address these problems, in which all spatial configuration information contained in the intermediate reasoning process is retained in a pathway of convolutional layers. We show that this new attention leads to substantial improvements in multiple multimodal reasoning tasks, including achieving single model performance without using external knowledge comparable to the state-of-the-art on the VQA dataset, as well as clear gains for the image captioning task.