To send content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about sending content to .
To send content items to your Kindle, first ensure firstname.lastname@example.org
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about sending to your Kindle.
Note you can select to send to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be sent to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
The rocky shores of the north-east Atlantic have been long studied. Our focus is from Gibraltar to Norway plus the Azores and Iceland. Phylogeographic processes shape biogeographic patterns of biodiversity. Long-term and broadscale studies have shown the responses of biota to past climate fluctuations and more recent anthropogenic climate change. Inter- and intra-specific species interactions along sharp local environmental gradients shape distributions and community structure and hence ecosystem functioning. Shifts in domination by fucoids in shelter to barnacles/mussels in exposure are mediated by grazing by patellid limpets. Further south fucoids become increasingly rare, with species disappearing or restricted to estuarine refuges, caused by greater desiccation and grazing pressure. Mesoscale processes influence bottom-up nutrient forcing and larval supply, hence affecting species abundance and distribution, and can be proximate factors setting range edges (e.g., the English Channel, the Iberian Peninsula). Impacts of invasive non-native species are reviewed. Knowledge gaps such as the work on rockpools and host–parasite dynamics are also outlined.
Background: Cervical sponylotic myelopathy (CSM) may present with neck and arm pain. This study investiagtes the change in neck/arm pain post-operatively in CSM. Methods: This ambispective study llocated 402 patients through the Canadian Spine Outcomes and Research Network. Outcome measures were the visual analogue scales for neck and arm pain (VAS-NP and VAS-AP) and the neck disability index (NDI). The thresholds for minimum clinically important differences (MCIDs) for VAS-NP and VAS-AP were determined to be 2.6 and 4.1. Results: VAS-NP improved from mean of 5.6±2.9 to 3.8±2.7 at 12 months (P<0.001). VAS-AP improved from 5.8±2.9 to 3.5±3.0 at 12 months (P<0.001). The MCIDs for VAS-NP and VAS-AP were also reached at 12 months. Based on the NDI, patients were grouped into those with mild pain/no pain (33%) versus moderate/severe pain (67%). At 3 months, a significantly high proportion of patients with moderate/severe pain (45.8%) demonstrated an improvement into mild/no pain, whereas 27.2% with mild/no pain demonstrated worsening into moderate/severe pain (P <0.001). At 12 months, 17.4% with mild/no pain experienced worsening of their NDI (P<0.001). Conclusions: This study suggests that neck and arm pain responds to surgical decompression in patients with CSM and reaches the MCIDs for VAS-AP and VAS-NP at 12 months.
Knowledge of the effects of burial depth and burial duration on seed viability and, consequently, seedbank persistence of Palmer amaranth (Amaranthus palmeri S. Watson) and waterhemp [Amaranthus tuberculatus (Moq.) J. D. Sauer] ecotypes can be used for the development of efficient weed management programs. This is of particular interest, given the great fecundity of both species and, consequently, their high seedbank replenishment potential. Seeds of both species collected from five different locations across the United States were investigated in seven states (sites) with different soil and climatic conditions. Seeds were placed at two depths (0 and 15 cm) for 3 yr. Each year, seeds were retrieved, and seed damage (shrunken, malformed, or broken) plus losses (deteriorated and futile germination) and viability were evaluated. Greater seed damage plus loss averaged across seed origin, burial depth, and year was recorded for lots tested at Illinois (51.3% and 51.8%) followed by Tennessee (40.5% and 45.1%) and Missouri (39.2% and 42%) for A. palmeri and A. tuberculatus, respectively. The site differences for seed persistence were probably due to higher volumetric water content at these sites. Rates of seed demise were directly proportional to burial depth (α=0.001), whereas the percentage of viable seeds recovered after 36 mo on the soil surface ranged from 4.1% to 4.3% compared with 5% to 5.3% at the 15-cm depth for A. palmeri and A. tuberculatus, respectively. Seed viability loss was greater in the seeds placed on the soil surface compared with the buried seeds. The greatest influences on seed viability were burial conditions and time and site-specific soil conditions, more so than geographical location. Thus, management of these weed species should focus on reducing seed shattering, enhancing seed removal from the soil surface, or adjusting tillage systems.
The introduction of 2,4-D–resistant soybean will provide an additional POST herbicide site of action for control of herbicide-resistant broadleaf weeds. The introduction of this technology also brings concern of off-site movement of 2,4-D onto susceptible crops such as sensitive soybean and tomato. The 2,4-D formulation approved for use in 2,4-D–resistant soybean restricts application of the herbicide to nozzles that produce very coarse to ultra-coarse droplet spectrums. The use of larger droplet spectrums for broadcast applications can reduce herbicide deposition onto target weeds and thus influence herbicide efficacy. Field experiments were conducted to evaluate the influence of nozzle design on herbicide deposition onto target plants and the resulting efficacy of a POST application of 280 g ha−1 glyphosate plus 280 g ha−1 2,4-D. The TTI11004 nozzle produced an ultra-coarse droplet spectrum and reduced coverage and deposition density on spray cards as compared with the XR11004 and TT11004 nozzles that produced medium droplet spectrums. The AIXR11004 nozzle also reduced deposition density on spray cards but did not reduce coverage. Herbicide solution deposition onto glyphosate-resistant Palmer amaranth, tall waterhemp, giant ragweed, and horseweed ranged from 0.28 to 0.72 µl cm−2 and was not influenced by nozzle design. Herbicide efficacy was reduced by the TTI11004 nozzle on Palmer amaranth and horseweed compared with the AIXR11004, TT11004, and XR11004 nozzles when applications were made to either high densities of plants or plants exceeding the labeled height. The use of the AIXR11004 and TTI11004 nozzles that are listed as approved nozzles for glyphosate plus 2,4-D applications on 2,4-D–resistant soybean did not reduce herbicide deposition onto four of the most troublesome broadleaves and did not reduce herbicide efficacy when applied in conjunction with lower weed densities and smaller weeds.
Dicamba-resistant soybean technology provides an additional site of action for POST control of herbicide-resistant broadleaf weeds in soybean but also raises concern of off-site movement and damage to sensitive crops in adjacent fields. Dicamba formulations approved for use on dicamba-resistant soybean require applicators to use nozzles producing large droplets to reduce the risk of spray-particle drift. The use of nozzles with relatively larger droplet spectra can reduce herbicide deposition on target weeds, especially if a filtering effect from the crop canopy occurs. Experiments were conducted to evaluate the influence of broadcast nozzle design on the deposition and efficacy of 280 g ha−1 glyphosate plus 140 g ha−1 dicamba applied POST to four herbicide-resistant weed species. The TTI11004 nozzle, the original nozzle labeled for dicamba applications on dicamba-resistant soybean, reduced deposition coverage and density on spray cards compared with the TT11004 and XR11004 nozzle. The AIXR11004 nozzle produces a very coarse droplet spectrum and did not reduce coverage on spray cards, though it did reduce deposition density. Herbicide solution deposition onto Palmer amaranth, tall waterhemp, giant ragweed, and horseweed ranged from 0.41 to 0.52, 0.55 to 0.87, 0.49 to 0.58, and 0.38 to 0.41 µl cm−2, respectively. Nozzle design and droplet spectrum did not influence the deposition of herbicide solution onto the target weed, as all nozzles were equivalent for all species and site-years. Herbicide efficacy was not influenced by nozzle design, as weed control and plant height reduction were similar for all species. The results of this experiment show that the use of the TTI11004 nozzle for dicamba applications to dicamba-resistant soybean will provide acceptable herbicide deposition and efficacy when applied under the label requirements of weed height and carrier volume.
Giant ragweed is a highly competitive weed that continually threatens crop production systems due to evolved resistance to acetolactate synthase–inhibiting herbicides (ALS-R) and glyphosate (GR). Two biotypes of GR giant ragweed exist and are differentiated by their response to glyphosate, termed here as rapid response (RR) and non–rapid response (NRR). A comparison of data from surveys of Indiana crop fields done in 2006 and 2014 showed that GR giant ragweed has spread from 15% to 39% of Indiana counties and the NRR biotype is the most prevalent. A TaqMan® single-nucleotide polymorphism genotyping assay was developed to identify ALS-R populations and revealed 47% of GR populations to be ALS-R as well. The magnitude of glyphosate resistance for NRR populations was 4.6 and 5.9 based on GR50 and LD50 estimates, respectively. For RR populations, these values were 7.8 to 9.2 for GR50 estimates and 19.3 to 22.3 for LD50 estimates. A novel use of the Imaging-PAM fluorometer was developed to discriminate RR plants by assessing photosystem II quantum yield across the entire leaf surface. H2O2 generation in leaves of glyphosate-treated plants was also measured by 3,3′-diaminobenzidine staining and quantified using imagery analysis software. Results show photo-oxidative stress of mature leaves is far greater and occurs more rapidly following glyphosate treatment in RR plants compared with NRR and glyphosate-susceptible plants and is positively associated with glyphosate dose. These results suggest that under continued glyphosate selection pressure, the RR biotype may surpass the NRR biotype as the predominant form of GR giant ragweed in Indiana due to a higher level of glyphosate resistance. Moreover, the differential photo-oxidative stress patterns in response to glyphosate provide evidence of different mechanisms of resistance present in RR and NRR biotypes.
Although high dose n-3 PUFA supplementation reduces exercise- and hyperpnoea-induced bronchoconstriction (EIB/HIB), there are concurrent issues with cost, compliance and gastrointestinal discomfort. It is thus pertinent to establish the efficacy of lower n-3 PUFA doses. Eight male adults with asthma and HIB and eight controls without asthma were randomly supplemented with two n-3 PUFA doses (6·2 g/d (3·7 g EPA and 2·5 g DHA) and 3·1 g/d (1·8 g EPA and 1·3 g DHA)) and a placebo, each for 21 d followed by 14 d washout. A eucapnic voluntary hyperpnoea (EVH) challenge was performed before and after treatments. Outcome measures remained unchanged in the control group. In the HIB group, the peak fall in forced expiratory volume in 1 s (FEV1) after EVH at day 0 (−1005 (sd 520) ml, −30 (sd 18) %) was unchanged after placebo. The peak fall in FEV1 was similarly reduced from day 0 to day 21 of 6·2 g/d n-3 PUFA (−1000 (sd 460) ml, −29 (sd 17) % v. −690 (sd 460) ml, −20 (sd 15) %) and 3·1 g/d n-3 PUFA (−970 (sd 480) ml, −28 (sd 18) % v. −700 (sd 420) ml, −21 (sd 15) %) (P<0·001). Baseline fraction of exhaled nitric oxide was reduced by 24 % (P=0·020) and 31 % (P=0·018) after 6·2 and 3·1 g/d n-3 PUFA, respectively. Peak increases in 9α, 11β PGF2 after EVH were reduced by 65 % (P=0·009) and 56 % (P=0·041) after 6·2 and 3·1 g/d n-3 PUFA, respectively. In conclusion, 3·1 g/d n-3 PUFA supplementation attenuated HIB and markers of airway inflammation to a similar extent as a higher dose. Lower doses of n-3 PUFA thus represent a potentially beneficial adjunct treatment for adults with asthma and EIB.
Introduction: Data regarding adverse events (AEs) (unintended harm to the patient from health care provided) among children seen in the emergency department (ED) are scarce despite the high risk setting and population. The objective of our study was to estimate the risk and type of AEs, and their preventability and severity, among children treated in pediatric EDs. Methods: Our prospective cohort study enrolled children <18 years of age presenting for care during 21 randomized 8 hr-shifts at 9 pediatric EDs from Nov 2014 to October 2015. Exclusion criteria included unavailability for follow-up or insurmountable language barrier. RAs collected demographic, medical history, ED course, and systems level data. At day 7, 14, and 21 a RA administered a structured telephone interview to all patients to identify flagged outcomes (e.g. repeat ED visits, worsening/new symptoms, etc). A validated trigger tool was used to screen admitted patients’ health records. For any patients with a flagged outcome or trigger, 3 ED physicians independently determined if an AE occurred. Primary outcome was the proportion of patients with an AE related to ED care within 3 weeks of their ED visit. Results: We enrolled 6377 (72.0%) of 8855 eligible patients; 545 (8.5%) were lost to follow-up. Median age was 4.4 years (range 3 months to 17.9 yrs). Eight hundred and seventy seven (13.8%) were triaged as CTAS 1 or 2, 2638 (41.4%) as CTAS 3, and 2839 (44.7%) as CTAS 4 or 5. Top entrance complaints were fever (11.2%) and cough (8.8%). Flagged outcomes/triggers were identified for 2047 (32.1%) patients. While 252 (4.0%) patients suffered at least one AE within 3 weeks of ED visit, 163 (2.6%) suffered an AE related to ED care. In total, patients suffered 286 AEs, most (67.9%) being preventable. The most common AE types were management issues (32.5%) and procedural complications (21.9%). The need for a medical intervention (33.9%) and another ED visit (33.9%) were the most frequent clinical consequences. In univariate analysis, older age, chronic conditions, hospital admission, initial location in high acuity area of the ED, having >1 ED MD or a consultant involved in care, (all p<0.001) and longer length of stay (p<0.01) were associated with AEs. Conclusion: While our multicentre study found a lower risk of AEs among pediatric ED patients than reported among pediatric inpatients and adult ED patients, a high proportion of these AEs were preventable.
We consider the slow motion generated when a body is set into motion relative to an incompressible, inviscid, non-diffusive rotating stratified fluid, showing that there is generated in general a topographic Rossby wave which leads to non-decaying fluctuations in the lift on the obstacle and a fluctuating non-zero drag. The problem is relevant to the flow patterns and forces excited when slow oceanic flows cross bottom topography, and suggests a mechanism for slow fluctuations observed in laboratory experiments.
Epidemiological studies suggest a higher prevalence of congenital malformations in children conceived through assisted reproductive technologies. There are a few studies that address CHD specifically and most have examined data from registries. We examined the relationship between CHD and assisted conception using data collected in a specialist paediatric cardiac service in the United Kingdom.
Between April, 2010 and July, 2011, the parents of children attending paediatric cardiology clinics at the Royal Brompton Hospital, London, were invited to complete a questionnaire that enquired about the nature of their child’s conception, the route for their original referral, and a number of potential confounding exposures. “Cases” were defined as children diagnosed with one or more carefully defined CHDs and “controls” as those with normal hearts.
Of 894 new attendees with complete data, half of them were cases (n=410, 45.9%). The overall prevalence of assisted conception was 5.4% (n=44). Logistic regression analysis demonstrated a non-significant increase in the crude odds for the use of assisted reproduction (odds ratio 1.21, 95% confidence interval 0.66–2.22) in this group. After adjustment for gestation, parity, year of birth, and maternal age, the odds ratio reduced (odds ratio 0.95, 95% confidence interval 0.48–1.88). Increased rates of assisted conception were observed in a number of CHD subgroups, although no significant differences were found.
These findings do not suggest an overall association between CHD and assisted reproduction in this population.
Debris and pyroclastic flows often have bouldery flow fronts, which act as a natural dam resisting further advance. Counter intuitively, these resistive fronts can lead to enhanced run-out, because they can be shouldered aside to form static levees that self-channelise the flow. At the heart of this behaviour is the inherent process of size segregation, with different sized particles readily separating into distinct vertical layers through a combination of kinetic sieving and squeeze expulsion. The result is an upward coarsening of the size distribution with the largest grains collecting at the top of the flow, where the flow velocity is greatest, allowing them to be preferentially transported to the front. Here, the large grains may be overrun, resegregated towards the surface and recirculated before being shouldered aside into lateral levees. A key element of this recirculation mechanism is the formation of a breaking size-segregation wave, which allows large particles that have been overrun to rise up into the faster moving parts of the flow as small particles are sheared over the top. Observations from experiments and discrete particle simulations in a moving-bed flume indicate that, whilst most large particles recirculate quickly at the front, a few recirculate very slowly through regions of many small particles at the rear. This behaviour is modelled in this paper using asymmetric segregation flux functions. Exact non-diffuse solutions are derived for the steady wave structure using the method of characteristics with a cubic segregation flux. Three different structures emerge, dependent on the degree of asymmetry and the non-convexity of the segregation flux function. In particular, a novel ‘lens-tail’ solution is found for segregation fluxes that have a large amount of non-convexity, with an additional expansion fan and compression wave forming a ‘tail’ upstream of the ‘lens’ region. Analysis of exact solutions for the particle motion shows that the large particle motion through the ‘lens-tail’ is fundamentally different to the classical ‘lens’ solutions. A few large particles starting near the bottom of the breaking wave pass through the ‘tail’, where they travel in a region of many small particles with a very small vertical velocity, and take significantly longer to recirculate.
Palmer amaranth and waterhemp have become increasingly troublesome weeds throughout the United States. Both species are highly adaptable and emerge continuously throughout the summer months, presenting the need for a residual PRE application in soybean. To improve season-long control of Amaranthus spp., 19 PRE treatments were evaluated on glyphosate-resistant Palmer amaranth in 2013 and 2014 at locations in Arkansas, Indiana, Nebraska, Illinois, and Tennessee; and on glyphosate-resistant waterhemp at locations in Illinois, Missouri, and Nebraska. The two Amaranthus species were analyzed separately; data for each species were pooled across site-years, and site-year was included as a random variable in the analyses. The dissipation of weed control throughout the course of the experiments was compared among treatments with the use of regression analysis where percent weed control was described as a function of time (the number of weeks after treatment [WAT]). At the mean (i.e., average) WAT (4.3 and 3.2 WAT for Palmer amaranth and waterhemp, respectively) isoxaflutole + S-metolachlor + metribuzin had the highest predicted control of Palmer amaranth (98%) and waterhemp (99%). Isoxaflutole + S-metolachlor + metribuzin, S-metolachlor + mesotrione, and flumioxazin + pyroxasulfone had a predicted control ≥ 97% and similar model parameter estimates, indicating control declined at similar rates for these treatments. Dicamba and 2,4-D provided some, short-lived residual control of Amaranthus spp. When dicamba was added to metribuzin or S-metolachlor, control increased compared to dicamba alone. Flumioxazin + pyroxasulfone, a currently labeled PRE, performed similarly to treatments containing isoxaflutole or mesotrione. Additional sites of action will provide soybean growers more opportunities to control these weeds and reduce the potential for herbicide resistance.
Herbicide-resistant Amaranthus spp. continue to cause management difficulties in soybean. New soybean technologies under development, including resistance to various combinations of glyphosate, glufosinate, dicamba, 2,4-D, isoxaflutole, and mesotrione, will make possible the use of additional herbicide sites of action in soybean than is currently available. When this research was conducted, these soybean traits were still regulated and testing herbicide programs with the appropriate soybean genetics in a single experiment was not feasible. Therefore, the effectiveness of various herbicide programs (PRE herbicides followed by POST herbicides) was evaluated in bare-ground experiments on glyphosate-resistant Palmer amaranth and glyphosate-resistant waterhemp (both tall and common) at locations in Arkansas, Illinois, Indiana, Missouri, Nebraska, and Tennessee. Twenty-five herbicide programs were evaluated; 5 of which were PRE herbicides only, 10 were PRE herbicides followed by POST herbicides 3 to 4 wks after (WA) the PRE application (EPOST), and 10 were PRE herbicides followed by POST herbicides 6 to 7 WA the PRE application (LPOST). Programs with EPOST herbicides provided 94% or greater control of Palmer amaranth and waterhemp at 3 to 4 WA the EPOST. Overall, programs with LPOST herbicides resulted in a period of weed emergence in which weeds would typically compete with a crop. Weeds were not completely controlled with the LPOST herbicides because weed sizes were larger (≥ 15 cm) compared with their sizes at the EPOST application (≤ 7 cm). Most programs with LPOST herbicides provided 80 to 95% control at 3 to 4 WA applied LPOST. Based on an orthogonal contrast, using a synthetic-auxin herbicide LPOST improves control of Palmer amaranth and waterhemp over programs not containing a synthetic-auxin LPOST. These results show herbicides that can be used in soybean and that contain auxinic- or HPPD-resistant traits will provide growers with an opportunity for better control of glyphosate-resistant Palmer amaranth and waterhemp over a wide range of geographies and environments.
Yellow and purple nutsedge are common in the southeastern United States, and both perennial species are difficult to control in organic crop-production systems. Tubers are generally confined to the upper portions of the soil profile and are vulnerable to desiccation when brought to the soil surface. A peanut digger is a common implement found in the coastal plain region of the southeastern United States and has shown promise controlling perennial nutsedges in fallow sites. The peanut digger undercuts perennial nutsedges, deposits weeds on the soil surface, and exposes weeds to desiccation. However, rainfall after tillage with the peanut digger allows displaced nutsedges to survive. As part of a senior-level class project, undergraduate mechanical engineering students from Auburn University designed and constructed a cart attached to a peanut digger that collected nutsedges. Key features included a custom hitch that allowed the correct plane of movement and a hydraulic conveyor system that discarded the perennial nutsedges off-site, away from the field. The prototype was tested in a fallow location in the summer of 2014 with a yellow nutsedge infestation averaging 148 plants m−2. One week after the initial field test, tillage using the peanut digger with specialized cart reduced yellow nutsedge densities in the tilled area by > 99%.
Fall-applied residual and spring preplant burn-down herbicide applications are typically used to control winter annual weeds and may also provide early-season residual control of summer annual weed species such as giant ragweed. Field experiments were conducted from 2006 to 2008 in southern Illinois to (1) assess the emergence pattern of giant ragweed, (2) evaluate the efficacy of several herbicides commonly used for soil-residual control of giant ragweed, and (3) investigate the optimal application timing of soil-residual herbicides for control of giant ragweed. Six herbicide treatments were applied at four application timings: early fall, late fall, early spring, and late spring. Giant ragweed first emerged in mid- and late-March in 2007 and 2008, respectively. The duration of emergence varied by year, with 95% of emergence complete in late May of 2008, but not until early July in 2007. Giant ragweed emergence occurred more quickly in plots that received a fall application of glyphosate + 2,4-D compared with the nontreated. Fall-applied residual herbicides did not reduce giant ragweed emergence in 2007 when compared with the nontreated, with the exception of chlorimuron + tribenuron applied in late fall. Giant ragweed control from early- and late-spring herbicide applications was variable by year. In 2007, saflufenacil (50 and 100 g ai ha−1) and simazine applied in early spring reduced giant ragweed densities by 95% or greater through mid-May; however, in 2008, early-spring applications failed to reduce giant ragweed emergence in mid-April. The only treatments that reduced giant ragweed densities by > 80% through early July were late-spring applications of chlorimuron + tribenuron or saflufenacil at 100 g ha−1. Thus, the emergence patterns of giant ragweed in southern Illinois dictates that best management with herbicides would include late-spring applications of soil-residual herbicides just before crop planting and most likely requires subsequent control with foliar or soil-residual herbicides after crop emergence.
Accurate and complete reporting of study methods, results and interpretation are essential components for any scientific process, allowing end-users to evaluate the internal and external validity of a study. When animals are used in research, excellence in reporting is expected as a matter of continued ethical acceptability of animal use in the sciences. Our primary objective was to assess completeness of reporting for a series of studies relevant to mitigation of pain in neonatal piglets undergoing routine management procedures. Our second objective was to illustrate how authors can report the items in the Reporting guidElines For randomized controLled trials for livEstoCk and food safety (REFLECT) statement using examples from the animal welfare science literature. A total of 52 studies from 40 articles were evaluated using a modified REFLECT statement. No single study reported all REFLECT checklist items. Seven studies reported specific objectives with testable hypotheses. Six studies identified primary or secondary outcomes. Randomization and blinding were considered to be partially reported in 21 and 18 studies, respectively. No studies reported the rationale for sample sizes. Several studies failed to report key design features such as units for measurement, means, standard deviations, standard errors for continuous outcomes or comparative characteristics for categorical outcomes expressed as either rates or proportions. In the discipline of animal welfare science, authors, reviewers and editors are encouraged to use available reporting guidelines to ensure that scientific methods and results are adequately described and free of misrepresentations and inaccuracies. Complete and accurate reporting increases the ability to apply the results of studies to the decision-making process and prevent wastage of financial and animal resources.
The time series of vital signs, such as heart rate (HR) and blood pressure (BP), can exhibit complex dynamic behaviors as a result of internally and externally induced changes in the state of the underlying control systems (Peng et al. 1995; Ivanov et al. 1999; Costa et al. 2002). For instance, time series of BP can exhibit oscillations on the order of seconds (e.g., due to the variations in sympathovagal balance), to minutes (e.g., as a consequence of fever, blood loss, or behavioral factors), to hours (e.g., due to humoral variations, sleep-wake cycle, or circadian effects) (Mancia 2012; Parati et al. 2013). A question of interest is whether “similar” dynamical patterns can be automatically identified across a heterogeneous patient cohort, and be used for prognosis of patients' health and progress.
In this work, we present a Bayesian nonparametric switching Markov processes framework with conditionally linear dynamics to learn phenotypic dynamic behaviors from vital sign time series of a patient cohort, and use the learned dynamics to characterize the changing physiological states of patients for critical-care bed-side monitoring (Lehman et al. 2012, 2013, 2014a; Nemati 2012). We assume that although the underlying dynamical system may be nonlinear and nonstationary and the stochastic noise components can be non-Gaussian, the dynamics can be approximated by a collection of linear dynamical systems (Nemati 2012; Nemati et al. 2012). Each such linear “dynamic” (or mode) is a time-dependent rule that describes how the future state of the system evolves from its current state, centered around a given system equilibrium point. Therefore, an ideal algorithm would be able to identify time series segments that follow a “similar” dynamic, and would switch to a different mode upon a change in the state of the underlying system.
We explore several variants of the Bayesian nonparametric approach to discovery of shared dynamics among patients via switching Markov processes: hierarchical Dirichlet process (HDP) autoregressive hidden Markov model (HDP-AR-HMM) (Teh et al. 2006; Fox et al. 2008), an explicit-duration HDP-based hidden semi-Markov model (HDP-AR-HSMM) (Johnson & Willsky 2013a), and the beta process autoregressive HMM (BP-AR-HMM) (Fox 2009; Fox et al. 2009, 2014).