DOSED: A deep learning approach to detect multiple sleep micro-events in EEG signal.

BACKGROUND: Electroencephalography (EEG) monitors brain activity during sleep and is used to identify sleep disorders. In sleep medicine, clinicians interpret raw EEG signals in so-called sleep stages, which are assigned by experts to every 30s window of signal. For diagnosis, they also rely on shorter prototypical micro-architecture events which exhibit variable durations and shapes, such as spindles, K-complexes or arousals. Annotating such events is traditionally performed by a trained sleep expert, making the process time consuming, tedious and subject to inter-scorer variability. To automate this procedure, various methods have been developed, yet these are event-specific and rely on the extraction of hand-crafted features. NEW METHOD: We propose a novel deep learning architecture called Dreem One Shot Event Detector (DOSED). DOSED jointly predicts locations, durations and types of events in EEG time series. The proposed approach, applied here on sleep related micro-architecture events, is inspired by object detectors developed for computer vision such as YOLO and SSD. It relies on a convolutional neural network that builds a feature representation from raw EEG signals, as well as two modules performing localization and classification respectively. RESULTS AND COMPARISON WITH OTHER METHODS: The proposed approach is tested on 4 datasets and 3 types of events (spindles, K-complexes, arousals) and compared to the current state-of-the-art detection algorithms. CONCLUSIONS: Results demonstrate the versatility of this new approach and improved performance compared to the current state-of-the-art detection methods.

Effect of liraglutide 3.0 mg in individuals with obesity and moderate or severe obstructive sleep apnea: the SCALE Sleep Apnea randomized clinical trial.

BACKGROUND: Obesity is strongly associated with prevalence of obstructive sleep apnea (OSA), and weight loss has been shown to reduce disease severity. OBJECTIVE: To investigate whether liraglutide 3.0 mg reduces OSA severity compared with placebo using the primary end point of change in apnea-hypopnea index (AHI) after 32 weeks. Liraglutide's weight loss efficacy was also examined. SUBJECTS/METHODS: In this randomized, double-blind trial, non-diabetic participants with obesity who had moderate (AHI 15-29.9 events h(-1)) or severe (AHI ⩾30 events h(-1)) OSA and were unwilling/unable to use continuous positive airway pressure therapy were randomized for 32 weeks to liraglutide 3.0 mg (n=180) or placebo (n=179), both as adjunct to diet (500 kcal day(-1) deficit) and exercise. Baseline characteristics were similar between groups (mean age 48.5 years, males 71.9%, AHI 49.2 events h(-1), severe OSA 67.1%, body weight 117.6 kg, body mass index 39.1 kg m(-2), prediabetes 63.2%, HbA1c 5.7%). RESULTS: After 32 weeks, the mean reduction in AHI was greater with liraglutide than with placebo (-12.2 vs -6.1 events h(-1), estimated treatment difference: -6.1 events h(-1) (95% confidence interval (CI), -11.0 to -1.2), P=0.0150). Liraglutide produced greater mean percentage weight loss compared with placebo (-5.7% vs -1.6%, estimated treatment difference: -4.2% (95% CI, -5.2 to -3.1%), P<0.0001). A statistically significant association between the degree of weight loss and improvement in OSA end points (P<0.01, all) was demonstrated post hoc. Greater reductions in glycated hemoglobin (HbA1c) and systolic blood pressure (SBP) were seen with liraglutide versus placebo (both P<0.001). The safety profile of liraglutide 3.0 mg was similar to that seen with doses ⩽1.8 mg. CONCLUSIONS: As an adjunct to diet and exercise, liraglutide 3.0 mg was generally well tolerated and produced significantly greater reductions than placebo in AHI, body weight, SBP and HbA1c in participants with obesity and moderate/severe OSA. The results confirm that weight loss improves OSA-related parameters.

The structured diagnostic interview for sleep patterns and disorders: rationale and initial evaluation.

OBJECTIVES: We aimed to describe and report the initial validity of a newly developed structured interview for sleep disorders (Diagnostic Interview for Sleep Patterns and Disorders [DISP]) administered by trained lay interviewers. METHODS: A total of 225 patients with various sleep disorders were recruited from two nationally recognized sleep centers in the United States. The International Classification of Sleep Disorders, second edition (ICSD-2) criteria, were used to classify sleep disorders (e.g., delayed sleep phase disorder, hypersomnia, narcolepsy with cataplexy [NC], restless legs syndrome [RLS], periodic limb movement disorder [PLMD], insomnia, rapid eye movement sleep behavior disorder [RBD], and obstructive sleep apnea [OSA]). Interview diagnoses were compared with final diagnoses by sleep specialists (reference diagnosis based on clinical history, examination, and polysomnography [PSG] when indicated). RESULTS: DISP diagnoses had fair to substantial concordance with clinician diagnoses for various sleep disorders, with area under the receiver operator characteristic curves (AUC) ranging from 0.65 to 0.84. Participants classified by the clinician as having a sleep disorder were moderately well-detected (sensitivity ranging from 0.50 for RBD disorder to 0.87 for insomnia). Substantial specificity (>0.8) also was seen for five of the eight sleep disorders (i.e., delayed sleep phase, hypersomnia, NC, PLMD, and RBD). Interviews were more likely than clinicians to detect disorders secondary to the primary sleep problem. CONCLUSIONS: The DISP provides an important tool for the detection of a wide range of sleep disorders in clinical settings and is particularly valuable in the detection of secondary disorders that were not the primary referral diagnosis.

HLA-DQ association and allele competition in Chinese narcolepsy.

In Japanese, Koreans and Caucasians, narcolepsy/hypocretin deficiency is tightly associated with the DRB1*15:01-DQA1*01:02-DQB1*06:02 haplotype. Studies in African-Americans suggest a primary effect of DQB1*06:02, but this observation has been difficult to confirm in other populations because of high linkage disequilibrium between DRB1*15:01/3 and DQB1*06:02 in most populations. In this study, we studied human leucocyte antigen (HLA) class II in 202 Chinese narcolepsy patients (11% from South China) and found all patients to be DQB1*06:02 positive. Comparing cases with 103 unselected controls, and 110 and 79 controls selected for the presence of DQB1*06:02 and DRB1*15:01, we found that the presence of DQB1*06:02 and not DRB1*15:01 was associated with narcolepsy. In particular, Southern Chinese haplotypes such as the DRB1*15:01-DQA1*01:02-DQB1*06:01 and DRB1*15:01-DQA1*01:02-DQB1*05 were not associated with narcolepsy. As reported in Japanese, Koreans, African-Americans and Caucasians, additional protective effects of DQA1*01 (non-DQA1*01:02) and susceptibility effects of DQB1*03:01 were observed. These results illustrate the extraordinary conservation of HLA class II effects in narcolepsy across populations and show that DRB1*15:01 has no effect on narcolepsy susceptibility in the absence of DQB1*06:02. The results are also in line with a previously proposed 'HLA-DQ allelic competition model' that involves competition between non-DQA1*01:02, non-DQB1*06:02 'competent' (able to dimerize together) DQ1 alleles and the major DQα*01:02/ DQß*06:02 narcolepsy heterodimer to reduce susceptibility.

Experiments and 3D simulations of flow structures in junctions and their influence on location of flowmeters.

Open-channel junctions are common occurrences in sewer networks and flow rate measurement often occurs near these singularities. Local flow structures are 3D, impact on the representativeness of the local flow measurements and thus lead to deviations in the flow rate estimation. The present study aims (i) to measure and simulate the flow pattern in a junction flow, (ii) to analyse the impact of the junction on the velocity distribution according to the distance from the junction and thus (iii) to evaluate the typical error derived from the computation of the flow rate close to the junction.

Effects of sleep on the cardiovascular and thermoregulatory systems: a possible role for hypocretins.

Sleep influences the cardiovascular, endocrine, and thermoregulatory systems. Each of these systems may be affected by the activity of hypocretin (orexin)-producing neurons, which are involved in the etiology of narcolepsy. We examined sleep in male rats, either hypocretin neuron-ablated orexin/ataxin-3 transgenic (narcoleptic) rats or their wild-type littermates. We simultaneously monitored electroencephalographic and electromyographic activity, core body temperature, tail temperature, blood pressure, electrocardiographic activity, and locomotion. We analyzed the daily patterns of these variables, parsing sleep and circadian components and changes between states of sleep. We also analyzed the baroreceptor reflex. Our results show that while core temperature and heart rate are affected by both sleep and time of day, blood pressure is mostly affected by sleep. As expected, we found that both blood pressure and heart rate were acutely affected by sleep state transitions in both genotypes. Interestingly, hypocretin neuron-ablated rats have significantly lower systolic and diastolic blood pressure during all sleep stages (non-rapid eye movement, rapid eye movement) and while awake (quiet, active). Thus, while hypocretins are critical for the normal temporal structure of sleep and wakefulness, they also appear to be important in regulating baseline blood pressure and possibly in modulating the effects of sleep on blood pressure.

Ventilatory chemoresponsiveness, narcolepsy-cataplexy and human leukocyte antigen DQB1*0602 status.

We hypothesised that hypocretin (orexin) plays a role in the determination of ventilatory chemosensitivity. 130 patients with narcolepsy-cataplexy (mean ± SD age 20 ± 10 yrs, 69% male) and 117 controls (22 ± 6.9 yrs, 62% male) were recruited and tested for human leukocyte antigen (HLA)-DQB1*0602 status, hyperoxia hypercapnic (change in minute ventilation (δV'(E))/carbon dioxide tension (δP(CO(2))) L·min(-1)·mmHg(-1)) and hypoxic (δV'(E) /change in arterial oxygen saturation measured by probe oximetry (δS(p,O(2))) L·min(-1) per %S(p,O(2))) responsiveness, and by spirometry. Hypocretin deficiency was determined either by measures of cerebrospinal fluid hypocretin-1 (37 patients) or by positive HLA-DQB1*0602 status. All patients and 49% of controls underwent polysomnography and multiple sleep latency testing. Despite similar spirometric values, patients had a higher apnoea/hypopnoea index (AHI) (2.8 ± 5.4 versus 0.8 ± 1.6 h(-1); p = 0.03) and lower minimal oxygen saturation during sleep (87% ± 7 versus 91 ± 4%; p = 0.0002), independent of age, sex and body mass index. Patients had depressed hypoxic responsiveness (0.13 ± 0.09 versus 0.19 ± 0.13 L·min(-1) per %S(p,O(2)); p<0.0001), independent of AHI, but hypercapnic responsiveness did not differ. Examined by HLA status, positive (26 out of 117) controls had lower hypoxic but similar hypercapnic responsiveness than those marker-negative (0.13 ± 0.08 versus 0.20 ± 0.14 L·min(-1) per %S(p,O(2)); p<0.0001). Thus, a lower hypoxic responsiveness in the narcolepsy-cataplexy group is a result of DQB1*0602 status rather than the clinical features of disease.

Absence of ubiquitinated inclusions in hypocretin neurons of patients with narcolepsy.

OBJECTIVES: The cause of hypocretin cell loss in human narcolepsy-cataplexy is unknown but has been suggested to be neurodegenerative in nature. To test this hypothesis, we evaluated the remaining hypocretin cells in human narcolepsy brains for the presence of aggregated protein inclusions, gliosis, and inflammation. METHODS: Brains were examined by routine histologic methods for potential comorbid neurodegenerative diseases and through immunohistochemical screening for protein inclusions in the hypothalamus. Hypothalamic sections of 4 subjects with narcolepsy and 5 nonneurologic controls were examined immunohistochemically with antibodies against ubiquitin (a marker of aggregated protein), allograft inflammatory factor 1 (AIF1, a microglial activation marker), glial fibrillary acidic protein (GFAP, a reactive astrocytic marker), and hypocretin. Hypothalami of subjects with narcolepsy were additionally examined for the presence of known components of protein aggregates (tau, alpha-synuclein, amyloid beta, and TDP-43). RESULTS: Hypocretin cells were markedly decreased in all 4 subjects with narcolepsy. Ubiquitinated inclusions were not observed in the total of 96 remaining hypocretin cells in these subjects. Further, we noted that even in patients with dementia neuropathology, the lateral hypothalamic hypocretin area was spared from ubiquitinated inclusions. AIF1 and GFAP staining in the perifornical area was unremarkable. CONCLUSIONS: Our findings suggest that hypocretin cell loss does not involve ubiquitinated inclusions, the hallmark of most neurodegenerative diseases. The lack of increased markers of inflammation also argues against a progressive and continuous neurodegenerative process.

The hypocretin neurotransmission system in myotonic dystrophy type 1.

BACKGROUND: Patients with myotonic dystrophy type 1 (DM1) frequently have symptoms of excessive daytime sleepiness (EDS). Some patients with DM1 show sleep-onset REM, similar to that observed in narcolepsy. Narcolepsy is characterized by impaired hypocretin (Hcrt) neurotransmission. OBJECTIVE: To test for dysregulation of Hcrt neurotransmission in a prospective cohort of patients with DM1. METHODS: Hcrt levels in CSF were measured by radioimmunoassay. Sleep physiology was assessed by overnight polysomnography (PSG) and a multiple sleep latency test (MSLT). Splicing of Hcrt receptor 1 and 2 (HcrtR1 and HcrtR2) mRNA was examined in postmortem samples of temporal cortex. RESULTS: Seventeen of 38 patients with DM1 reported symptoms of EDS. Among patients with DM1 with EDS who underwent PSG/MSLT, 7 of 13 showed reduced sleep latency, sleep-onset REM, or both. However, CSF Hcrt levels in DM1 (mean 277 pg/mL, n = 38) were not different from controls (mean 277 pg/mL, n = 33). Also, splicing of HcrtR1 and HcrtR2 mRNA in patients with DM1 was similar to controls. CONCLUSIONS: Excessive daytime sleepiness and dysregulation of REM sleep occur frequently in patients with myotonic dystrophy type 1 (DM1). However, the pathophysiologic basis is distinct from narcolepsy, as patients with DM1 do not have a consistent defect of Hcrt release or receptor splicing.

The development of hypocretin (orexin) deficiency in hypocretin/ataxin-3 transgenic rats.

Narcolepsy is linked to a widespread loss of neurons containing the neuropeptide hypocretin (HCRT), also named orexin. A transgenic (TG) rat model has been developed to mimic the neuronal loss found in narcoleptic humans. In these rats, HCRT neurons gradually die as a result of the expression of a poly-glutamine repeat under the control of the HCRT promoter. To better characterize the changes in HCRT-1 levels in response to the gradual HCRT neuronal loss cerebrospinal fluid (CSF) HCRT-1 levels were measured in various age groups (2-82 weeks) of wild-type (WT) and TG Sprague-Dawley rats. TG rats showed a sharp decline in CSF HCRT-1 level at week 4 with levels remaining consistently low (26%+/-9%, mean+/-S.D.) thereafter compared with WT rats. In TG rats, HCRT-1 levels were dramatically lower in target regions such as the cortex and brainstem (100-fold), indicating decreased HCRT-1 levels at terminals. In TG rats, CSF HCRT-1 levels significantly increased in response to 6 h of prolonged waking, indicating that the remaining HCRT neurons can be stimulated to release more neuropeptide. Rapid eye movement (REM) sleep in TG rats (n=5) was consistent with a HCRT deficiency. In TG rats HCRT immunoreactive (HCRT-ir) neurons were present in the lateral hypothalamus (LH), even in old rats (24 months) but some HCRT-ir somata were in various stages of disintegration. The low output of these neurons is consistent with a widespread dysfunction of these neurons, and establishes this model as a tool to investigate the consequences of partial hypocretin deficiency.

Comparison of numerical and experimental simulations of a flood in a dense urban area.

Although various numerical methods were used to simulate real floods occurring in cities, the validation of the models was never accurate because of the lack of data about location and event description and about observation for validation. In order to check the capacities of our 2-dimensional shallow water equations model to simulate an urban flood, we then decided to simulate numerically an experimental event with well known characteristics and accurate flow measurements. The physical model presented in (Ishigaki et al, 2003) represents the flooding of the city center of Kyoto in Japan due to an overflow from the Kamo river. The 2-dimensional numerical simulation of this event was then set up and the experimental and computed data were compared. It appears that the event was calculated quite fairly in terms of flow depth and flow rates in the streets and in terms of timing. However, some discrepancies appear between the measurements and the numerical results, mostly due to some topographical local uncertainties and to the capacities of the equations to model the complex flows in the crossroads.

One-dimensional GIS-based model compared with a two-dimensional model in urban floods simulation.

A GIS-based one-dimensional flood simulation model is presented and applied to the centre of the city of Nîmes (Gard, France), for mapping flow depths or velocities in the streets network. The geometry of the one-dimensional elements is derived from the Digital Elevation Model (DEM). The flow is routed from one element to the next using the kinematic wave approximation. At the crossroads, the flows in the downstream branches are computed using a conceptual scheme. This scheme was previously designed to fit Y-shaped pipes junctions, and has been modified here to fit X-shaped crossroads. The results were compared with the results of a two-dimensional hydrodynamic model based on the full shallow water equations. The comparison shows that good agreements can be found in the steepest streets of the study zone, but differences may be important in the other streets. Some reasons that can explain the differences between the two models are given and some research possibilities are proposed.

Narcolepsy-cataplexy associated with precocious puberty.

In children, narcolepsy may be the symptom of a brain lesion or genetic disease. The authors report two cases with severe narcolepsy-cataplexy emerging in childhood in close temporal association with obesity and precocious puberty.

Concomitant loss of dynorphin, NARP, and orexin in narcolepsy.

BACKGROUND: Narcolepsy with cataplexy is associated with a loss of orexin/hypocretin. It is speculated that an autoimmune process kills the orexin-producing neurons, but these cells may survive yet fail to produce orexin. OBJECTIVE: To examine whether other markers of the orexin neurons are lost in narcolepsy with cataplexy. METHODS: We used immunohistochemistry and in situ hybridization to examine the expression of orexin, neuronal activity-regulated pentraxin (NARP), and prodynorphin in hypothalami from five control and two narcoleptic individuals. RESULTS: In the control hypothalami, at least 80% of the orexin-producing neurons also contained prodynorphin mRNA and NARP. In the patients with narcolepsy, the number of cells producing these markers was reduced to about 5 to 10% of normal. CONCLUSIONS: Narcolepsy with cataplexy is likely caused by a loss of the orexin-producing neurons. In addition, loss of dynorphin and neuronal activity-regulated pentraxin may contribute to the symptoms of narcolepsy.

Kleine-Levin syndrome: a systematic review of 186 cases in the literature.

Kleine-Levin syndrome (KLS) is a rare disorder with symptoms that include periodic hypersomnia, cognitive and behavioural disturbances. Large series of patients are lacking. In order to report on various KLS symptoms, identify risk factors and analyse treatment response, we performed a systematic review of 195 articles, written in English and non-English languages, which are available on Medline dating from 1962 to 2004. Doubtful or duplicate cases, case series without individual details and reviews (n = 56 articles) were excluded. In addition, the details of 186 patients from 139 articles were compiled. Primary KLS cases (n = 168) were found mostly in men (68%) and occurred sporadically worldwide. The median age of onset was 15 years (range 4-82 years, 81% during the second decade) and the syndrome lasted 8 years, with seven episodes of 10 days, recurring every 3.5 months (median values) with the disease lasting longer in women and in patients with less frequent episodes during the first year. It was precipitated most frequently by infections (38.2%), head trauma (9%), or alcohol consumption (5.4%). Common symptoms were hypersomnia (100%), cognitive changes (96%, including a specific feeling of derealization), eating disturbances (80%), hypersexuality (43%), compulsions (29%), and depressed mood (48%). In 75 treated patients (213 trials), somnolence decreased using stimulants (mainly amphetamines) in 40% of cases, while neuroleptics and antidepressants were of poor benefit. Only lithium (but not carbamazepine or other antiepileptics) had a higher reported response rate (41%) for stopping relapses when compared to medical abstention (19%). Secondary KLS (n = 18) patients were older and had more frequent and longer episodes, but had clinical symptoms and treatment responses similar to primary cases. In conclusion, KLS is a unique disease which may be more severe in female and secondary cases.

Hypocretin-1 (orexin A) deficiency in acute traumatic brain injury.

Hypocretin-1 is involved in the regulation of the sleep-wake cycle. The authors prospectively assessed CSF hypocretin-1 levels in 44 consecutive patients with acute traumatic brain injury (TBI). Compared with controls, hypocretin-1 levels were abnormally lower in 95% of patients with moderate to severe TBI and in 97% of patients with posttraumatic brain CT changes. Hypocretin-1 deficiency after TBI may reflect hypothalamic damage and be linked with the frequent development of posttraumatic sleep-wake disorders.

Hypocretin deficiency in Prader-Willi syndrome.

Four patients with clinically and genetically confirmed Prader-Willi syndrome (PWS) underwent nocturnal polysomnograpy (PSG), multiple sleep latency test (MSLT), human leukocyte antigens (HLA) typing and estimation of cerebrospinal fluid (CSF) hypocretin-1 (Hcrt-1) level to investigate if a role of hypothalamic dysfunction and sleep disturbance might be functionally connected through the hypocretin (orexin) system. In all four patients physical examination confirmed extreme obesity (increasing with age) with dysmorphogenetic features. Excessive daytime sleepiness (EDS) was manifested in only two subjects without any imperative feature. None of the patients under study suffered from cataplexy. Nocturnal PSG revealed fragmented sleep with low efficiency, the hypopnea and apnea indexes increasing from borderline up to very high values in direct proportion to the patients' age. MSLT latency was shortened in two patients with clinically expressed EDS, only one sleep onset rapid eye movements (REM) period (SOREM) was found. HLA typing showed DQB1*0602 positivity in two patients; the further two were negative. Mean value of CSF Hcrt-1 in the patients group was down to 164 +/- 46.8 pg/ml (in comparison with 265.8 +/- 48.8 pg/ml in 10 young healthy subjects, P=0.02). The deficiency of CSF Hcrt-1 level correlated in PWS patients with their EDS severity.

Narcolepsy research: past, present, and future perspectives.

In 1877, Westphal described a patient with hypersomnia and episodic muscle weakness. He did not feel that these weakness attacks could simply be explained by "epileptoid" phenomenon. The next year, Fischer described a similar case. By 1880, Gélineau decided that patients with these symptoms represented a distinct clinical entity and he called it "narcolepsy". In 1902, Loewenfeld noted the importance of cataplexy in this disorder, and in 1934 Daniels published an important review on the topic which helped to galvanize interest in further study. In 1957, Yoss and Daly discussed the "clinical tetrad" which included hypersomnia, cataplexy, hypnagogic hallucinations, and sleep paralysis. In 1960, Vogel noted that patients with narcolepsy had early onset of REM sleep on their electroencephalograms. At the First International Symposium on Narcolepsy in 1975, the symptom of disturbed nocturnal sleep was added to the clinical diagnostic criteria for narcolepsy. For many years the etiology and mechanisms of this disease were poorly understood. It was not until the early 1970s when the exciting animal and human research first started to unravel the mysteries of the genetics and physiology of narcolepsy. This research will be discussed below.

APOE epsilon4 is associated with obstructive sleep apnea/hypopnea: the Sleep Heart Health Study.

BACKGROUND: Obstructive sleep apnea/hypopnea (OSAH) has a strong heritable component, although its genetic basis remains largely unknown. One epidemiologic study found a significant association between the APOE epsilon4 allele and OSAH in middle-aged adults, a finding that was not replicated in a cohort of elderly adults. The objective of this study was to further examine the association of the APOE epsilon4 allele with OSAH in a community-dwelling cohort, exploring age dependency of the association. METHODS: A genetic association study was performed, nested within a prospective cohort study of the cardiovascular consequences of OSAH. Unattended, in-home nocturnal polysomnography was used to measure apnea-hypopnea index (AHI) in 1,775 participants age 40 to 100 years. OSAH was defined as an AHI > or = 15. The relation of APOE genotype to prevalent OSAH was analyzed using generalized estimating equations to account for non-independent observations of individuals from the same sibship. RESULTS: At least one APOE epsilon4 allele was present in 25% of subjects, with 1.3% epsilon4/epsilon4 homozygotes. The prevalence of OSAH was 19%. After adjustment for age, sex, and BMI, the presence of any APOE epsilon4 allele was associated with increased odds of OSAH (OR 1.41, 95% CI 1.06 to 1.87, p = 0.02). The effect was approximately twice as great in subjects <75 (OR 1.61, CI 1.02 to 2.54) as in those > or =75 years old (OR 1.32, CI 0.91 to 1.90). Exploratory analyses revealed that the strongest effect of APOE epsilon4 was in subjects age <65 (OR 3.08, CI 1.43 to 6.64), and was stronger in those with hypertension or cardiovascular disease than in those without. CONCLUSION: The APOE epsilon4 allele is associated with increased risk of OSAH, particularly in individuals under age 65. The mechanisms underlying this association are uncertain. Age-dependency of the APOE-OSAH association may explain previous conflicting results.