To define training zones, ventilatory thresholds (VTs) are commonly established by cardiopulmonary gas-exchange analysis during incremental exercise tests. Portable near-infrared spectroscopy (NIRS) devices have emerged as a potential tool for detecting these thresholds by monitoring muscle oxygenation. This study evaluated the accuracy of NIRS measurements to determine VTs or critical power (CP) based on muscle oxygen saturation and assesses the device’s consistency across 2 constant-load tests. Data from 2 cross-sectional studies involving trained recreational endurance athletes (26 from study 1) and CrossFit athletes (59 from study 2) were examined. Incremental ramp tests on a cycle ergometer were performed and followed by either a constant-load test (study 1) or a CP test (study 2). When comparing power output or heart rate between NIRS-derived breakpoints and VTs, weak to moderate agreement was found. Mean differences in power output and heart rate ranged from 16.8 to 22.4 W and 3.8 to 6.0 beats·min−1 at the first threshold and 27.4 to 31.2 W and 7.1 to 7.8 beats·min−1 at the second threshold. Comparing with CP, mean differences ranged from −0.4 to 0.4 W and −0.6 to 0.9 beats·min−1. Test–retest reliability showed moderate agreement, with a mean bias of 1.2 percentage points between constant-load tests. Thus, NIRS may not be accurate for determining VTs or CP during exercise due to limited agreement in power output or hear rate, notable variability on individual level, and moderate reproducibility.

Background Phosphocreatine dynamics provide the gold standard evaluation of in-vivo mitochondrial function and is tightly coupled with oxygen availability. Low mitochondrial oxidative capacity has been associated with health issues and low exercise performance. Methods To evaluate the relationship between near-infrared spectroscopy-based muscle oxygen dynamics and magnetic resonance spectroscopy-based energy-rich phosphates, a systematic review of the literature related to muscle oxygen dynamics and energy-rich phosphates was conducted. PRISMA guidelines were followed to perform a comprehensive and systematic search of four databases on 02-11-2021 (PubMed, MEDLINE, Scopus and Web of Science). Beforehand pre-registration with the Open Science Framework was performed. Studies had to include healthy humans aged 18–55, measures related to NIRS-based muscle oxygen measures in combination with energy-rich phosphates. Exclusion criteria were clinical populations, laboratory animals, acutely injured subjects, data that only assessed oxygen dynamics or energy-rich phosphates, or grey literature. The Effective Public Health Practice Project Quality Assessment Tool was used to assess methodological quality, and data extraction was presented in a table. Results Out of 1483 records, 28 were eligible. All included studies were rated moderate. The studies suggest muscle oxygen dynamics could indicate energy-rich phosphates under appropriate protocol settings. Conclusion Arterial occlusion and exercise intensity might be important factors to control if NIRS application should be used to examine energetics. However, more research needs to be conducted without arterial occlusion and with high-intensity exercises to support the applicability of NIRS and provide an agreement level in the concurrent course of muscle oxygen kinetics and muscle energetics. Trial registration https://osf.io/py32n/ . Key points 1. NIRS derived measures of muscle oxygenation agree with gold-standard measures of high energy phosphates when assessed in an appropriate protocol setting. 2. At rest when applying the AO protocol, in the absence of muscle activity, an initial disjunction between the NIRS signal and high energy phosphates can been seen, suggesting a cascading relationship. 3. During exercise and recovery a disruption of oxygen delivery is required to provide the appropriate setting for evaluation through either an AO protocol or high intensity contractions.

ObjectivesTo compare performance progression and variety in race distances of comparable lengths (timewise) between pool swimming and track running. Quality of within-sport variety was determined as the performance differences between individual athletes' main and secondary race distances across (top-) elite and (highly-) trained swimmers and runners.MethodsA total of 3,827,947 race times were used to calculate performance points (race times relative to the world record) for freestyle swimmers (n = 12,588 males and n = 7,561 females) and track runners (n = 9,230 males and n = 5,841 females). Athletes were ranked based on their personal best at peak performance age, then annual best times were retrospectively traced throughout adolescence.ResultsPerformance of world-class swimmers differentiates at an earlier age from their lower ranked peers (15–16 vs. 17–20 year age categories, P < 0.05), but also plateaus earlier towards senior age compared to runners (19–20 vs. 23 + year age category, P < 0.05), respectively. Performance development of swimmers shows a logarithmic pattern, while runners develop linearly. While swimmers compete in more secondary race distances (larger within-sport variety), runners specialize in either sprint, middle- or long-distance early in their career and compete in only 2, 4 or 3 other race distances, respectively. In both sports, sprinters specialize the most (P < 0.05). Distance-variety of middle-distance swimmers covers more longer rather than sprint race distances. Therefore, at peak performance age, (top-) elite female 200 m swimmers show significantly slower sprint performances, i.e., 50 m (P < 0.001) and 100 m (P < 0.001), but not long-distance performances, i.e., 800 m (P = 0.99) and 1,500 m (P = 0.99). In contrast, (top-) elite female 800 m middle-distance runners show significantly slower performances in all their secondary race distances (P < 0.001). (Top-) elite female athletes specialize more than (highly-) trained athletes in both sports (P < 0.05).ConclusionsThe comparison to track running and lower ranked swimmers, the early performance plateau towards senior age, and the maintenance of a large within-sport distance variety indicates that (top-) elite sprint swimmers benefit from greater within-sport specialization.

The critical force (CF) concept, differentiating steady and non-steady state conditions, extends the critical power paradigm for sport climbing. This study aimed to validate CF for finger flexors derived from the 4 min all-out test as a boundary for the highest sustainable work intensity in sport climbers. Twelve participants underwent multiple laboratory visits. Initially, they performed the 4 min intermittent contraction all-out test for CF determination. Subsequent verification visits involved finger-flexor contractions at various intensities, including CF, CF −2 kg, CF −4 kg, and CF −6 kg, lasting for 720 s or until failure, while monitoring muscle-oxygen dynamics of forearm muscles. CF, determined from the mean force of last three contractions, was measured at 20.1 ± 5.7 kg, while the end-force at 16.8 ± 5.2 kg. In the verification trials, the mean time to failure at CF was 440 ± 140 s, with only one participant completing the 720 s task. When the load was continuously lowered (−2 kg, −4 kg, and −6 kg), a greater number of participants (38

Monitoring muscle metabolic activity via blood lactate is a useful tool for understanding the physiological response to a given exercise intensity. Recent indications suggest that skeletal muscle oxygen saturation (SmO2), an index of the balance between local O2 supply and demand, may describe and predict endurance performance outcomes. We tested the hypothesis that SmO2 rate is tightly related to blood lactate concentration across exercise intensities, and that deflections in SmO2 rate would coincide with established blood lactate thresholds (i.e., lactate thresholds 1 and 2). Ten elite male soccer players completed an incremental running protocol to exhaustion using 3-min work to 30 s rest intervals. Blood lactate samples were collected during rest and SmO2 was collected continuously via near-infrared spectroscopy from the right and left vastus lateralis, left biceps femoris and the left gastrocnemius. Muscle O2 saturation rate (

BACKGROUND: Forearm muscle strength and endurance are essential determinants of sports climbing success. This study aimed to investi-gate whether delayed rates of muscle oxygen saturation and total hemoglobin correlate to sustained contraction performance of youth climbers.METHODS: Twelve recreational and competitive youth sport climbers (six females, six males) participated in the study. Variables included fin-ger flexors muscle maximal voluntary contraction, sustained contraction test (SCT), muscle oxygen dynamics (SmO2), and blood volume (tHb) parameters. Pearson's correlation coefficients were calculated to determine the correlation between physiological and performance variables.RESULTS: SCT had a significant positive relationship to SmO2 delayed rate (r=0.728, P=0.007), and a significant negative relationship to tHb delayed rate (r=-0.690, P=0.013). SmO2 delayed rate and tHb delayed rate also had a significant negative correlation (r=-0.760, P=0.004).CONCLUSIONS: According to the results of this study, it can be suggested that delayed rates of SmO2 and tHb could be used in determining and predicting sustainable finger flexors performance in youth climbers. However, future studies investigating delayed rates of SmO2 and tHb in climbers of different ability levels are warranted to investigate this issue in more detail.

PURPOSE:Following short-term all-out exercise, the maximal rate of glycolysis is frequently assessed on the basis of the maximal rate of lactate accumulation in the blood. Since the end of the interval without significant accumulation (talac) is 1 of 2 denominators in the calculation employed, accurate determination of this parameter is crucial. Although the very existence and definition of talac, as well as the validity of its determination as time-to-peak power (tPpeak), remain controversial, this parameter plays a key role in anaerobic diagnostics. Here, we describe a novel approach to determination of talac and compare it to the current standard.METHODS:Twelve elite track cyclists performed 3 maximal sprints (3, 8, and 12 s) and a high-rate, low-resistance pedaling test on an ergometer with monitoring of crank force and pedaling rate. Before and after each sprint, capillary blood samples were taken for determination of lactate accumulation. Fatigue-free force-velocity and power-velocity profiles were generated. talac was determined as tPpeak and as the time point of the first systematic deviation from the force-velocity profile (tFf).RESULTS:Accumulation of lactate after the 3-second sprint was significant (0.58 [0.19] mmol L-1; P < .001, d = 1.982). tFf was <3 seconds and tPpeak was ≥3 seconds during all sprints (P < .001, d = - 2.111). Peak power output was lower than maximal power output (P < .001, d = -0.937). Blood lactate accumulation increased linearly with increasing duration of exercise (R2 ≥ .99) and intercepted the x-axis at ∼tFf.CONCLUSION:Definition of talac as tPpeak can lead to incorrect conclusions. We propose determination of talac based on tFf, the end of the fatigue-free state that may reflect the beginning of blood lactate accumulation.

We investigated the relationship of the time-dependent behaviour of muscle oxygen saturation SmO2(t), phosphagen energy supply WPCr(t) and blood lactate accumulation ΔBLC(t) during a 60-s all-out cycling sprint and tested SmO2(t) for correlations with the end of the fatigue-free state tFf, maximal pedalling rate PRmax and maximal blood lactate accumulation rate v̇Lamax. Nine male elite track cyclists performed four maximal sprints (3, 8, 12, 60 s) on a cycle ergometer. Crank force and cadence were monitored continuously to determine PRmax and tFf based on force-velocity profiles. SmO2 of the vastus lateralis muscle and respiratory gases were measured until the 30th minute after exercise. WPCr was calculated based on the fast component of the post-exercise oxygen uptake for each sprint. Before and for 30 minutes after each sprint, capillary blood samples were taken to determine the associated ΔBLC. Temporal changes of SmO2, WPCr and ΔBLC were analysed via non-linear regression analysis. v̇Lamax was calculated based on ΔBLC(t) as the highest blood lactate accumulation rate. All models showed excellent quality (R2 > 0.95). The time constant of SmO2(t) τSmO2 = 2.93 ± 0.65 s was correlated with the time constant of WPCr(t) τPCr = 3.23 ± 0.67 s (r = 0.790, p < 0.012), v̇Lamax = 0.95 ± 0.18 mmol · l-1 · s-1 (r = 0.768, p < 0.017) and PRmax = 299.51 ± 14.70 rpm (r = -0.670, p < 0.049). tFf was correlated with τSmO2 (r = 0.885, p < 0.001). Our results show a time-dependent reflection of SmO2 kinetics and phosphagen energy contribution during a 60-s maximal cycling sprint. A high v̇Lamax results in a reduction, a high PRmax in an increase of the desaturation rate. The half-life of SmO2 desaturation indicates the end of the fatigue-free state.

Pulmonary gas exchange analysis was compared to changes in muscle oxygen saturation as measured by near-infrared spectroscopy. First, ventilatory thresholds determined by common gas exchange analysis and breakpoints in muscle oxygen saturation were assessed for agreement during exercise with increasing intensity. Secondly, the relationship between muscle oxygen saturation as a surrogate for local oxygen extraction and peak oxygen uptake was assessed. In order to lend robustness to future NIRS testing on a broader scale, considering its potential for simple and cost-effective application, the question of a running versus a cycling modality was integrated into the design. Ten participants, of whom five were recreationally trained cyclists and five recreationally trained runners, were tested; each during a cycling test and a running test with increasing intensity to voluntary exhaustion. Muscle oxygen saturation and pulmonary gas exchange measurements were conducted. Bland-Altman analysis showed a moderate degree of agreement between both muscle oxygen saturation breakpoint 1 and muscle oxygen saturation breakpoint 2 and corresponding ventilatory threshold 1 and ventilatory threshold 2, for both cycling and running disciplines; generally speaking, muscle oxygen saturation breakpoints underestimated ventilatory thresholds. Additionally, a strong relationship could be seen between peak oxygen uptake and the minimally attained muscle oxygen saturation during cycling exercise. Muscle oxygen saturation measured using NIRS was determined to be a suitable method to assess ventilatory thresholds by finding breakpoints in muscle oxygen saturation, and muscle oxygen saturation minimum was linked to peak oxygen uptake.

The ability to supply and utilize oxygen is a critical component of exercise performance. Near-infrared spectroscopy (NIRS) to measure muscle oxygenation (SmO2) has shown to be useful in both assessing maximal oxygen uptake, as well as the utilization and recovery of high-energy phosphates. Therefore, SmO2 dynamics should play an important role in performance during repetitive sprinting tasks and could provide valuable insights into sport-specific performance. PURPOSE: Assess the relationship between performance and SmO2 during a Repetitive Ice Shuttle Sprint test (RISS). METHODS: Twenty elite-level hockey players completed a RISS. The RISS consisted of 3 shifts of 4 x 30 m maximum intensity skating sprints with 3 min of passive recovery between shifts. Sprint time was documented and SmO2 data collected from both right and left leg quadriceps. SmO2 data was analyzed for maximal deoxygenation (SmO2min) and reoxygenation dynamics during the interspersed sprint and recovery periods and compared to sprint times. For shift comparisons, a repeated-measures ANOVA was used, and a Pearson Product-Moment correlation was used to assess the relationship between SmO2 and shift times. RESULTS: Both sprint time and SmO2min show a significant change between shifts respectively; F(2,38) = 5.018, p = .012, F(2,38) = 5.705, p = .007. All 3 shifts correlate with medium to large effects sizes, between time and SmO2min, respectively; r(19) = .411, p = .036, r(19) = .338, p = .072, r(19) = .516, p = .01. All 3 shifts show small to medium effects sizes for intra-shift recovery and time relationships, respectively; r(19) = .424, p = .031, r(19) = .214, p = .183, r(19) = .316, p = .087. Finally, oxidative capacity using NIRS can be assessed through both reoxygenation dynamics and SmO2min, which show strong correlations for all shifts; r(19) = .820, p < .001, r(19) = .784, p < .001, r(19) = .834, p < .001. CONCLUSIONS: SmO2min and reoxygenation dynamics appear to be related to repeated sprint performance during high-intensity skating. This understanding could benefit both athletic diagnostics and training guidance by better understanding local bioenergetic systems and their limitations.

High-intensity training (HIT) is known to have deteriorating effects on performance which manifest in various physiological changes such as lowered force production and oxidative capacity. However, the effect of HIT in climbing on finger flexor performance has not been investigated yet. Twenty-one climbers partook in an intervention study with three assessment time points: pre-HIT, post-HIT, and 24-h post-HIT. The HIT involved four five-minute exhaustive climbing tasks. Eight climbers were assigned to a control group. Assessments consisted of three finger flexor tests: maximum voluntary contraction (MVC), sustained contraction (SCT), and intermittent contraction tests (ICT). During the SCT muscle oxygenation (SmO 2 ) metrics were collected via NIRS sensors on the forearm. The HIT had significant deteriorating effects on all force production metrics (MVC − 18%, SCT − 55%, ICT − 59%). Post-24 h showed significant recovery, which was less pronounced for the endurance tests (MVC − 3%, SCT − 16%, ICT − 22%). SmO 2 metrics provided similar results for the SCT with medium to large effect sizes. Minimally attainable SmO 2 and resting SmO 2 both showed moderate negative correlations with pre-HIT force production respectively; r = − 0.41, P = 0.102; r = − 0.361, P = 0.154. A strong association was found between a loss of force production and change in minimally attainable SmO 2 ( r = − 0.734, P = 0.016). This study presents novel findings on the deteriorating effects of HIT on finger flexor performance and their oxidative capacity. Specifically, the divergent results between strength and endurance tests should be of interest to coaches and athletes when assessing athlete readiness.

Sport climbing requires high-intensity finger flexor contractions, along with a substantial whole-body systemic oxygen uptake ( $$\mathop V\limits^{.}$$ O2) contribution. Although fatigue is often localised to the finger flexors, the role of systemic ̇ $$\mathop V\limits^{.}$$ O2 and local aerobic mechanisms in climbing performance remains unclear. As such, the primary purpose of this study was to determine systemic and local muscle oxygen responses during both isolated finger flexion and incremental exhaustive whole-body climbing tests. The secondary aim was to determine the relationship of isolated and whole-body climbing endurance tests to climbing ability. Twenty-two male sport climbers completed a series of isometric sustained and intermittent forearm flexor contractions, and an exhaustive climbing test with progressive steepening of the wall angle on a motorised climbing ergometer. Systemic $$\mathop V\limits^{.}$$ O2 and flexor digitorum profundus oxygen saturation (StO2) were recorded using portable metabolic analyser and near-infra red spectroscopy, respectively. Muscle oxygenation breakpoint (MOB) was identifiable during an incremental exhaustive climbing test with progressive increases in angle (82 ± 8% and 88 ± 8% $$\mathop V\limits^{.}$$ O2 and heart rate climbing peak). The peak angle from whole-body treadwall test and impulse from isolated hangboard endurance tests were interrelated (R2 = 0.58–0.64). Peak climbing angle together with mean $$\mathop V\limits^{.}$$ O2 and StO2 from submaximal climbing explained 83% of variance in self-reported climbing ability. Both systemic and muscle oxygen kinetics determine climbing-specific endurance. Exhaustive climbing and isolated finger flexion endurance tests are interrelated and suitable to assess climbing-specific endurance. An exhaustive climbing test with progressive wall angle allows determination of the MOB.

The power-duration relationship is well documented for athletic performance and is formulated out mathematically in the critical power (CP) model. The CP model, when applied properly, has great predictive power, e.g. pedaling at a specific power output on an ergometer the model precisely calculates the time over which an athlete can sustain this power. However, CP presents physiological inconsistencies and process-oriented problems. The rapid development of near-infrared spectroscopy (NIRS) to measure muscle oxygenation (SmO2) dynamics provides a physiological exploration of the CP model on a conceptual and empirical level. Conceptually, the CP model provides two components: first CP is defined as the highest metabolic rate that can be achieved through oxidative means. And second, work capacity above CP named W'. SmO2 presents a steady-state in oxygen supply and demand and thereby represents CP specifically at a local level of analysis. Empirically, exploratory data quickly illustrates the relationship between performance and SmO2, as shown during 3-min all-out cycling tests to assess CP. During these tests, performance and SmO2 essentially mirror each other, and both CP and W' generate solid correlation with what would be deemed their SmO2 counterparts: first, the steady-state of SmO2 correlates with CP. And second, the tissue oxygen reserve represented in SmO2, when calculated as an integral corresponds to W'. While the empirical data presented is preliminary, the proposition of a concurring physiological model to the current CP model is a plausible inference. Here we propose that SmO2 steady-state representing CP as critical oxygenation or CO. And the tissue oxygen reserve above CO would then be identified as O'. This new CO model could fill in the physiological gap between the highly predictive CP model and at times its inability to track human physiology consistently. For simplicity's sake, this would include acute changes in physiology as a result of changing climate or elevation with travel, which can affect performance. These types of acute fluctuations, but not limited to, would be manageable when applying a CO model in conjunction with the CP model. Further, modeling is needed to investigate the true potential of NIRS to model CP, with a focus on repeatability, recovery, and systemic vs local workloads.

PURPOSE: The critical power (CP) model identifies a performance-based fatigue threshold which can be effectively used to predict time to task failure (TTF). The model’s hyperbolic form identifies a clear asymptote (CP) which determines an over-under threshold with which an athlete can perform a task. Tasks over CP result in a depletion of work potential (W’) which ultimately results in failure when W’ is depleted. CP and W’ are performance characteristics and the physiological mechanism behind this phenomenon remain difficult to pinpoint. Nonetheless, performance above CP is characterised by unsustainable metabolic process which are a potential explanation for failure. A physiological measure that could represent CP and W’ would enhance the model’s effectiveness for athletics. Muscle oxygenation (SmO2) as measured by Near-infrared spectroscopy (NIRS) offers itself as a potential physiological surrogate for CP and W’, through a time-SmO2 integral (O’) identifying unsustainable metabolic process. METHODS: Eighteen participants (age: 21±1.7; weight: 68±11.1 kg) performed three trials of single-leg knee extensions, at 5%, 10% and 20% 1-RM, to exhaustion in order to evaluate individual power-duration curves. In order to eliminate the effect of blood flow knee extension trials were performed in occluded conditions (pressure cuff >300mmHg). NIRS and EMG sensors were placed on the vastus lateralis, vastus medialis and rectus femoris. O’ of the NIRS curve was calculated for each trial. RESULTS: The CP model predicts that W’ is constant across tasks to failure. The same prediction was made for O’. The results suggest that this assumption is correct and O’ remains constant for TTF; 5% 1-RM: M = -43.85, SD = 17.93, [95% CI:-35.3, -52.5]; 10% 1-RM: M = -44.75, SD 17.76, [95% CI: -36.2, -53.3]; 20% 1-RM M = -44.79, SD 16.16 [95% CI: -37.0, -52.5]. CONCLUSIONS: SmO2 represents a dynamic balance between O2 supply and O2 demand in real-time. An imbalance between O2 supply and O2 demand effects local muscle metabolism resulting in accumulating fatigue which if not restored results in specific TTF. The consistency between O’ and TTF over the power-duration curve suggest potential for a physiological approach to a classical performance threshold. This knowledge could be invaluable to TTF prediction at an individual physiological level.

The aim of this study was to measure muscle oxygen saturation (SmO2) dynamics during a climbing specific task until failure in varying conditions. Our prediction was that SmO2 should be a good marker to predict task failure. Eleven elite level climbers performed a finger-hang test on a 23 mm wooden rung under four different weighted conditions, 1. body weight (BW), 2. body weight +20% (BW +20), 3. body weight −20% (BW −20) and 4. body weight −40% (BW −40), maintaining half crimp grip until voluntary exhaustion. During each trial SmO2 and time to task failure (TTF) were measured. TTF was then compared to the minimally attainable value of SmO2 (SmO2min) and time to SmO2min (TTmin). There is a considerable degree of agreement between attainable SmO2min at high intensity conditions (MBW = 21.6% ± 6.4; MBW+20 = 24.0% ± 7.0; MBW−20 = 23.0% ± 7.3). Bland-Altman plot with an a priori set equivalency interval of ±5% indicate that these conditions are statistically not different (MBW-BW + 20 = −2.4%, 95% CI [1.4, −6.2]; MBW−Bw−20 = −1.3, 95% CI [2.5, −5.1]). The fourth and lowest intensity condition (MBW −40 = 32.4% ± 8.8) was statistically different and not equivalent (MBW-BW −40 = −8.8%, 95% CI [−5.0, −12.6]). The same agreement was found between TTF and TTmin for the high intensity conditions plotted via Bland-Altman. While the rate with which oxygen was extracted and utilised changed with the conditions, the attainable SmO2min remained constant at high intensity conditions and was related to TTF.

Reported Relationships: A. Feldmann: Salary; Part-Time employment in development. Contracted Research (Only include research funds received directly from industry; institution grants are NOT reportable.) If you are a PI, you must report a financial relationship even if those funds came to your institution; private industry contribution. Ownership Interest (stocks, stock options, or other ownership interest excluding diversified mutal funds). Please specify in the field specific ownership interests, (i.e. own the company, own stocks, future stock options); own stocks.

Near-infrared spectroscopy (NIRS) to monitor muscle oxygen saturation (SmO2) is rapidly expanding into applied sports settings. However, the technology is limited due to its inability to convey quantifiable values. A test battery to assess reliability and validity of a 0% to 100% scale modeled by a commercially available NIRS device was established. This test battery applies a commonly used technique, the arterial occlusion method (AOM) to assess repeatability, reproducibility, and face validity. A total of 22 participants completed the test battery to scrutinize the 0% to 100% scale provided by the device. All participants underwent repeated AOM tests in passive and active conditions. The SmO2 minimum and SmO2 maximum values were obtained from the AOM and were used in the subsequent analysis. Repeatability and reproducibility were tested for equivalency and Bland-Altman plots were generated. Face validity was assessed by testing SmO2 values against an a priori defined threshold for mixed venous blood during AOM response. The device exhibits an appropriately functional 0% to 100% scale that is reliable in terms of repeatability and reproducibility. Under the conditions applied in the test battery design, the device is considered valid for application in sports. (C) The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License.