Though people rarely speak in complete sentences, punctuation confers many benefits to the readers of transcribed speech. Unfortunately, most ASR systems do not produce punctuated output. To address this, we propose a solution for automatic punctuation that is both cost efficient and easy to train. Our solution benefits from the recent trend in fine-tuning transformer-based language models. We also modify the typical framing of this task by predicting punctuation for sequences rather than individual tokens, which makes for more efficient training and inference. Finally, we find that aggregating predictions across multiple context windows improves accuracy even further. Our best model achieves a new state of the art on benchmark data (TED Talks) with a combined F1 of 83.9, representing a 48.7% relative improvement (15.3 absolute) over the previous state of the art.
Pain catastrophizing was a significant predictor of pain severity in both seropositive and seronegative pSS patients. This study suggests that behavioural interventions designed to reduce pain catastrophizing and negative appraisal of illness could be of benefit in pSS patients. Research is needed to test the effect of psycho-educational therapies on key patient-reported outcomes, particularly pain, depression, and fatigue, in pSS.
Objective: The objective of this study is to investigate the relationship of oxidative stress to fatigue in systemic lupus erythematosus (SLE). Methods: Patients with a confirmed diagnosis of SLE by ACR criteria and healthy controls completed validated questionnaires to assess depression and fatigue. Fatigue was measured with the Fatigue Severity Scale (FSS) and the Profile of Fatigue (Prof-F). Visual analogue scales (VAS) were also used to assess fatigue and pain. Depression was measured with the Center for Epidemiologic Studies Depression Scale (CES-D). Plasma F2-isoprostane was measured with gas chromatography/mass spectroscopy to assess oxidative stress. Evaluation included medical record review, physical exam and calculation of body mass index (BMI), disease activity (SLEDAI) and damage (SLICC) in the SLE patients. Results: Seventy-one SLE patients with low disease activity (mean SLEDAI = 1.62 standard error (SE) 0.37, range 0–8) were compared to 51 controls. Fatigue-limiting physical activity (defined as FSS ≥ 4) was present in 56% of patients and 12% of controls. F2-isoprostane was higher in SLE patients with fatigue compared to not-fatigued SLE subjects ( p = .0076) who were otherwise similar in ethnicity, disease activity and cardiovascular risk factors. Plasma F2-isoprostane was strongly correlated with FSS and Profile of Somatic Fatigue (Prof-S) ( p < .0001), VAS fatigue ( p = .005), CES-D ( p = .008) and with BMI ( p = .0001.) In a multivariate model, F2-isoprostane was a significant predictor of FSS after adjustment for age, BMI, pain and depression ( p = .0002). Conclusion: Fatigue in SLE patients with low disease activity is associated with increased F2-isoprostane. F2-isoprostane could provide a useful biomarker to explore mitochondrial function and the regulation of oxidative pathways in patients with SLE in whom fatigue is a debilitating symptom.
Background Anxiety and depression are frequently associated with Sjogren’s syndrome and pain is a common complaint. Patients who experience these problems find their symptoms frightening and possibly more disabling than the sicca symptoms for which PSS is better known. Unfortunately the precise etiology of pain symptoms and the influence of psychological symptoms on pain perception is unknown. Objectives We investigated 1) the differential effect of depressive symptoms, anxiety, and catastrophizing on pain intensity and health status and 2) compared patients with neuropathic and non-neuropathic pain symptoms. Methods PSS patients who met AECG criteria (N=95) completed a survey to assess pain intensity: the Brief Pain Inventory1: BPI-S (0-10 scale), neuropathic pain symptoms: the Neuropathic Pain Questionnaire2 (NPP) and affective components of pain: the Pain Anxiety Symptom Scale3 (PASS) and the Pain Catastrophizing Scale4 (PCS). Health status was assessed with the SF-12, perceived stress with a visual analog scale. Linear regression was used to investigate the relationship between pain severity, pain catastrophizing and pain phenotype. Results PSS subjects were classified into those with neuropathic pain (NP, N=25) and those without neuropathic symptoms (Non-NP, N=59; 11 missing NP status). Daily pain for greater than 3 months was reported by 80% of NP and 49% of Non-NP patients. Based on weekly pain ratings, patients were also divided into: mild pain (N=49) and moderate-severe pain (N=42; 4 missing BPI score) groups. Perceived stress (p=0.015) was associated with more severe pain as were PASS (p=0.015) and PCS (p≤0.001). Pain severity correlated with physical domain of the SF-12(r=-.55), with HAD anxiety (r=0.38), HADs depression (r=0.48) and w/ PASS (r=0.48) and PCS (r=.61). The relationship between pain severity and pain catastrophizing was significant in PSS patients with a history of neuropathy(p=0.027) and in those with a diagnosis of FM (p=0.008). PCS also correlated with musculoskeletal pain severity (r=0.54) and with neuropathic pain symptoms (r=0.58). Conclusions Pain anxiety and pain catastrophizing are associated with increasing pain severity in PSS patients with different pain phenotypes. This study is consistent with psychological studies that have shown that “pain captures attention” and that patients with pain are hypervigilant for somatic sensations at the expense of other activities. Further study is needed to develop to develop discriminatory classifications of “central pain”, “nociceptive pain” and “neuropathic pain” which could provide insight into mechanistic pathways and a framework for more rationale therapy. References Cleeland CS, Syrjala KL. How to Assess Cancer Pain. In: Turk DC, Melzack R editors. Handbook of Pain Assessment. New York: the Guilford Press, 1992. p 362-287 Krause SJ, Backonja MM. Development of a neuropathic pain questionnaire. Clin J Pain 2003; 19: 306-414. McCracken LM, Zayfert C, Gross RT. The Pain Anxiety Symptoms Scale: development and validation of a scale t...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.