Background Nutrient dense food that supports health is a goal of food service in long-term care (LTC). The objective of this work was to characterize the “healthfulness” of foods in Canadian LTC and inflammatory potential of the LTC diet and how this varied by key covariates. Here, we define foods to have higher “healthfulness” if the are in accordance with the evidence-based 2019 Canada’s Food Guide, or with comparatively lower inflammatory potential. Methods We conducted a secondary analysis of the Making the Most of Mealtimes dataset (32 LTC homes; four provinces). A novel computational algorithm categorized food items from 3-day weighed food records into 68 expert-informed categories and Canada’s Food Guide (CFG) food groups. The dietary inflammatory potential of these food sources was assessed using the Dietary Inflammatory Index (DII). Comparisons were made by sex, diet texture, and nutritional status. Results Consumption patterns using expert-informed categories indicated no single protein or vegetable source was among the top 5 most commonly consumed foods. In terms of CFG’s groups, protein food sources (i.e., foods with a high protein content) represented the highest proportion of daily calorie intake (33.4%; animal-based: 31.6%, plant-based: 1.8%), followed by other foods (31.3%) including juice (9.8%), grains (25.0%; refined: 15.0%, whole: 10.0%), and vegetables/fruits (10.3%; plain: 4.9%, with additions: 5.4%). The overall DII score (mean, IQR) was positive (0.93, 0.23 to 1.75) indicating foods consumed tend towards a pro-inflammatory response. DII was significantly associated with sex (female higher; p<0.0001), and diet (minced higher; p=0.036). Conclusions “Healthfulness” of Canadian LTC menus may be enhanced by lowering inflammatory potential to support chronic disease management through further shifts from refined to whole grains, incorporating more plant-based proteins, and moving towards serving plain vegetables and fruits. However, there are multiple layers of complexities to consider when optimising foods aligned with the CFG, and shifting to foods with anti-inflammatory potential for enhanced health benefits, while balancing nutrition and ensuring sufficient food and fluid intake to prevent or treat malnutrition.

Nonintrusive estimation of oxygen uptake (V_O2) is possible with wearable sensor technology , artificial intelligence. V_O2 kinetics have been accurately predicted during moderate exercise using easy-to-obtain sensor inputs. However, V_O2 prediction algorithms for higher-intensity exercise with inherent nonlinearities are still being refined. The purpose of this investigation was to test if a machine learning model can accurately predict dynamic V_O2 across exercise intensities, including slower V_O2 kinetics normally observed during heavy-compared with moderate-intensity exercise. Fifteen young healthy adults (seven females; peak V_O2: 42 +/- 5 mL center dot min-1 center dot kg-1) performed three different pseudorandom binary sequence (PRBS) exercise tests ranging in intensity from low-to-moderate, low-to-heavy , ventilatory threshold-to-heavy work rates. A temporal convolutional network was trained to predict instantaneous V_O2, with model inputs including heart rate, percent heart rate reserve, estimated minute ventilation, breathing frequency, and work rate. Frequency domain analyses between V_O2 and work rate were used to evaluate measured and predicted V_O2 kinetics. Predicted V_O2 had low bias (-0.017 L center dot min-1, 95% limits of agreement: [-0.289, 0.254]), and was very strongly correlated (rrm =0.974, P < 0.001) with the measured V_O2. The extracted indicator of kinetics, mean normalized gain (MNG), was not different between predicted and measured V_O2 responses (main effect: P = 0.374, gp2 = 0.01), and decreased with increasing exercise intensity (main effect: P < 0.001, gp2= 0.64). Predicted and measured V_O2 kinetics indicators were moderately correlated across repeated measurements (MNG: rrm= 0.680, P < 0.001). Therefore, the temporal convolutional network accurately predicted slower V_O2 kinetics with increasing exercise intensity, enabling nonintrusive monitoring of cardior-espiratory dynamics across moderate-and heavy-exercise intensities.NEW , NOTEWORTHY Machine learning analysis of wearable sensor data with a sequential model, which utilized a receptive field of approximately 3 min to make instantaneous oxygen uptake estimations, accurately predicted oxygen uptake kinetics from moderate through to higher-intensity exercise. This innovation will enable nonintrusive cardiorespiratory monitoring over a wide range of exercise intensities encountered in vigorous training and competitive sports.

Malnutrition is a multidomain problem affecting 54% of older adults in long-term care (LTC). Monitoring nutritional intake in LTC is laborious and subjective, limiting clinical inference capabilities. Recent advances in automatic image-based food estimation have not yet been evaluated in LTC settings. Here, we describe a fully automatic imaging system for quantifying food intake. We propose a novel deep convolutional encoder-decoder food network with depth-refinement (EDFN-D) using an RGB-D camera for quantifying a plate's remaining food volume relative to reference portions in whole and modified texture foods. We trained and validated the network on the pre-labelled UNIMIB2016 food dataset and tested on our two novel LTC-inspired plate datasets (689 plate images, 36 unique foods). EDFN-D performed comparably to depth-refined graph cut on IOU (0.879 vs. 0.887), with intake errors well below typical 50% (mean percent intake error: -4.2%). We identify how standard segmentation metrics are insufficient due to visual-volume discordance, and include volume disparity analysis to facilitate system trust. This system provides improved transparency, approximates human assessors with enhanced objectivity, accuracy, and precision while avoiding hefty semi-automatic method time requirements. This may help address short-comings currently limiting utility of automated early malnutrition detection in resource-constrained LTC and hospital settings.

Non-contact coded hemodynamic imaging (CHI) is a novel wide-field near-infrared spectroscopy system which monitors blood volume by quantifying attenuation of light passing through the underlying vessels. This study tested the hypothesis that CHI-based jugular venous attenuation (JVA) would be larger in men, and change in JVA would be greater in men compared to women during two fluid shift challenges. The association of JVA with ultrasound-based cross-sectional area (CSA) was also tested. Ten men and 10 women completed three levels of head-down tilt (HDT) and four levels of lower body negative pressure (LBNP). Both JVA and CSA were increased by HDT and reduced by LBNP (all p < 0.001). Main effects of sex indicated that JVA was higher in men than women during both HDT (p = 0.003) and LBNP (p = 0.011). Interaction effects of sex and condition were observed for JVA during HDT (p = 0.005) and LBNP (p < 0.001). We observed moderate repeated-measures correlations (rrm ) between JVA and CSA in women during HDT (rrm = 0.57, p = 0.011) and in both men (rrm = 0.74, p < 0.001) and women (rrm = 0.66, p < 0.001) during LBNP. While median within-person correlation coefficients indicated an even stronger association between JVA and CSA, this association became unreliable for small changes in CSA. As hypothesized, JVA was greater and changed more in men compared to women during both HDT and LBNP. CHI provides a non-contact method of tracking large changes in internal jugular vein blood volume that occur with acute fluid shifts, but data should be interpreted in a sex-dependent manner.

Half of long-term care (LTC) residents are malnourished increasing hospitalization, mortality, morbidity, with lower quality of life. Current tracking methods are subjective and time consuming. This paper presents the automated food imaging and nutrient intake tracking (AFINI-T) technology designed for LTC. We propose a novel convolutional autoencoder for food classification, trained on an augmented UNIMIB2016 dataset and tested on our simulated LTC food intake dataset (12 meal scenarios; up to 15 classes each; top-1 classification accuracy: 88.9%; mean intake error: -0.4 mL$\pm$36.7 mL). Nutrient intake estimation by volume was strongly linearly correlated with nutrient estimates from mass ($r^2$ 0.92 to 0.99) with good agreement between methods ($\sigma$= -2.7 to -0.01; zero within each of the limits of agreement). The AFINI-T approach is a deep-learning powered computational nutrient sensing system that may provide a novel means for more accurately and objectively tracking LTC resident food intake to support and prevent malnutrition tracking strategies.

Significance The internal jugular veins are critical cerebral venous drainage pathways that are affected by right heart function. Cardiovascular disease and microgravity can alter central venous pressure (CVP) and venous return, which may contribute to increased intracranial pressure and decreased cardiac output. Assessing jugular venous compliance may provide insight into cerebral drainage and right heart function, but monitoring changes in vessel volume is challenging. Aim We investigated the feasibility of quantifying jugular venous compliance from jugular venous attenuation (JVA), a non-contact optical measurement of blood volume, alongside CVP from antecubital vein cannulation. Approach CVP was progressively increased through a guided graded Valsalva maneuver, increasing mouth pressure by 2 mmHg every 2 s until a maximum expiratory pressure of 20 mmHg. JVA was extracted from a 1 cm segment between the clavicle and mid-neck. Contralateral internal jugular vein cross-sectional area (CSA) was measured with ultrasound to validate changes in vessel size. Compliance was calculated using both JVA and CSA between four-beat averages over the duration of the maneuver. Results JVA and CSA were strongly correlated (median, interquartile range) over the Valsalva maneuver across participants (r=0.986, [0.983, 0.987]). CVP more than doubled on average between baseline and peak strain (10.7 ± 4.4 vs 25.8 ± 5.4 cmH 2 O; p <.01). JVA and CSA increased non-linearly with CVP, and both JVA- and CSA-derived compliance decreased progressively from baseline to peak strain (49% and 56% median reduction, respectively), with no significant difference in compliance reduction between the two measures ( Z =–1.24, p =.21). Pressure-volume curves showed a logarithmic relationship in both CSA and JVA. Conclusions Optical jugular vein assessment may provide new ways to assess jugular distention and cardiac function.

Objective: Frequency-domain diffuse optical spectroscopic imaging (FD-DOS) is a non-invasive method for measuring absolute concentrations of tissue chromophores such as oxy- and deoxy-hemoglobin in vivo. The utility of FD-DOS for clinical applications such as monitoring chemotherapy response in breast cancer has previously been demonstrated, but challenges for further clinical translation, such as slow acquisition speed and lack of user feedback, remain. Here, we propose a new high speed FD-DOS instrument that allows users to freely acquire measurements over the tissue surface, and is capable of rapidly imaging large volumes of tissue. Methods: We utilize 3D monocular probe tracking combined with custom digital FD-DOS hardware and a high-speed data processing pipeline for the instrument. Results are displayed during scanning over the surface of the sample using a probabilistic Monte Carlo light propagation model. Results: We show this instrument can measure absorption and scattering coefficients with an error of 7% and 1% respectively, with 0.7 mm positional accuracy. We demonstrate the equivalence of our visualization methodology with a standard interpolation approach, and demonstrate two proof-of-concept in vivo results showing superficial vasculature in the human forearm and surface contrast in a healthy human breast. Conclusion: Our new FD-DOS system is able to compute chromophore concentrations in real-time (1.5 Hz) in vivo. Significance: This method has the potential to improve the quality of FD-DOS image scans while reducing measurement times for a variety of clinical applications.

Oxygen consumption (V̇O2) provides established clinical and physiological indicators of cardiorespiratory function and exercise capacity. However, V̇O2 monitoring is largely limited to specialized laboratory settings, making its widespread monitoring elusive. Here, we investigate temporal prediction of V̇O2 from wearable sensors during cycle ergometer exercise using a temporal convolutional network (TCN). Cardiorespiratory signals were acquired from a smart shirt with integrated textile sensors alongside ground-truth V̇O2 from a metabolic system on twentytwo young healthy adults. Participants performed one rampincremental and three pseudorandom binary sequence exercise protocols to assess a range of V̇O2 dynamics. A TCN model was developed using causal convolutions across an effective history length to model the time-dependent nature of V̇O2. Optimal history length was determined through minimum validation loss across hyperparameter values. The best performing model encoded 218 s history length (TCN-VO2 A), with 187 s, 97 s, and 76 s yielding less than 3% deviation from the optimal validation loss. TCN-VO2 A showed strong prediction accuracy (mean, 95% CI) across all exercise intensities (−22 ml·min−1, [−262, 218]), spanning transitions from low-moderate (−23 ml·min−1, [−250, 204]), low-heavy (14 ml·min−1, [−252, 280]), ventilatory thresholdheavy (−49 ml·min−1, [−274, 176]), and maximal (−32 ml·min−1, [−261, 197]) exercise. Second-by-second classification of physical activity across 16090 s of predicted V̇O2 was able to discern between vigorous, moderate, and light activity with high accuracy (94.1%). This system enables quantitative aerobic activity monitoring in non-laboratory settings across a range of exercise intensities using wearable sensors for monitoring exercise prescription adherence and personal fitness.

The objective of this study was to investigate the impact of internal jugular vein (IJV) distension on blood volume pulsatility. We hypothesized that pulsatility would be enhanced during cephalad fluid shifts and attenuated during caudal fluid shifts. Thirteen young healthy participants (8/5 male/female) were assessed during supine rest (repeated); 3 and 6 degrees of head-down tilt (HDT); and -20, -30, -40 mmHg of lower body negative pressure (LBNP) after at least 5 minutes in each condition. The order of the HDT and LBNP protocols was randomized. Right-side IJV blood volume was estimated using widefield near-infrared spectroscopy and quantified as jugular venous optical attenuation (JVA). Central venous pressure (CVP) was measured using a transducer attached to a catheter inserted into an antecubital vein in the right arm and corrected to the level of the heart. Participants were rightward tilted throughout to facilitate venous pressure measurements. Continuous signals were acquired for 30s in each condition and time-synced at 60Hz. Volume and pressure pulsatility were calculated as the differences between peak and nadir points along the cardiac cycle in JVA and CVP, respectively. Pressure-volume loops were created after ensemble averaging waveforms normalized to the cardiac cycle. Area of the pressure-volume loops and slope of the pressure-volume relationship during atrial filling (i.e., venous drainage) were assessed. Group mean JVA and CVP were highest during 6° HDT and lowest during -40 mmHg LBNP (see figure). Compared to baseline, JVA pulsatility was attenuated during -30 mmHg LBNP (P=0.048) and -40 mmHg LBNP (P=0.004) and was unchanged during each level of HDT (P>0.9). In contrast, CVP pulsatility was attenuated compared to baseline during -20, -30, and -40 mmHg LBNP (P<0.001) and was unchanged during each level of HDT (P>0.68). Area (P=0.088) and slope (P=0.269) of the pressure-volume loop were unchanged across all conditions. In multiple regression analysis, loop area was associated directly with pressure amplitude (P<0.001) but not volume amplitude (P=0.130); model r2=0.69. These data suggest that, while volume pulsatility did decrease when venous distension was reduced by higher levels of LBNP, this was primarily a function of changes in pressure pulsatilty. IJV compliance appeared to be unchanged across a wide range of venous distension. This work has relevance to states where venous pressure and distension are altered, including heart failure and hemorrhagic shock.

Wearable technologies and artificial intelligence have enabled continuous and non-intrusive cardiorespiratory monitoring. Machine learning techniques, such as random forest (RF) and long short-term memory (LSTM) models, have been used to predict the oxygen uptake (V̇O2) response to exercise, but their abilities to track V̇O2 during changes in work rate lack precision. Here, we propose using a sequential deep learning model based on temporal convolutional networks (TCN) to estimate V̇O2 from wearable sensor data. Twenty-two healthy adults (9 females, age: 26±5 yr, peak V̇O2: 42±6 ml·min-1·kg-1) completed a 25 W·min-1 ramp cycling test to exhaustion, and a combination of 3 different pseudorandom binary sequence (PRBS) cycling tests to simulate non-constant work rate exercise ranging from low to moderate, low to heavy, and ventilatory threshold to heavy-intensity exercise, respectively. Breath-by-breath V̇O2 was measured using a portable metabolic device, and wearable sensor data were simultaneously collected using a Hexoskin® sensor shirt. Work rate, heart rate, percent heart rate reserve, estimated minute ventilation, and breathing rate were used as model inputs to predict instantaneous V̇O2. Participant data were split into 3 groups to train (n=10), validate (n=7), and test (n=5) the newly proposed TCN model and previously reported RF and LSTM models. Repeated measures Bland-Altman analysis (bias; 95% limits of agreement) revealed that the TCN model (-1 ml·min-1; -237 to 235 ml·min-1) was more accurate at estimating the dynamic V̇O2 responses to ramp and PRBS exercise than the LSTM (-10 ml·min-1; -321 to 301 ml·min-1) and RF models (62 ml·min-1; -259 to 383 ml·min-1), despite containing fewer trainable parameters (63% and 97% reduction vs. LSTM and RF, respectively). These results suggest that the TCN model is more accurate and efficient at estimating V̇O2 than other previously used machine learning models across a wide range of exercise intensities. Our findings are an important step in the development of a framework to non-intrusively assess aerobic fitness levels and energy expenditure during real-life situations outside of the laboratory without cumbersome equipment.

Here we implemented a Monte Carlo photon migration model to simulate optical tissue interaction during external jugular vein distension, supporting the potential of non-contact hemodynamic imaging techniques.

Abstract Oxygen consumption ( $$\dot{\,{{\mbox{V}}}}{{{\mbox{O}}}}_{2}$$ V ̇ O 2 ) provides established clinical and physiological indicators of cardiorespiratory function and exercise capacity. However, $$\dot{\,{{\mbox{V}}}}{{{\mbox{O}}}}_{2}$$ V ̇ O 2 monitoring is largely limited to specialized laboratory settings, making its widespread monitoring elusive. Here we investigate temporal prediction of $$\dot{\,{{\mbox{V}}}}{{{\mbox{O}}}}_{2}$$ V ̇ O 2 from wearable sensors during cycle ergometer exercise using a temporal convolutional network (TCN). Cardiorespiratory signals were acquired from a smart shirt with integrated textile sensors alongside ground-truth $$\dot{\,{{\mbox{V}}}}{{{\mbox{O}}}}_{2}$$ V ̇ O 2 from a metabolic system on 22 young healthy adults. Participants performed one ramp-incremental and three pseudorandom binary sequence exercise protocols to assess a range of $$\dot{\,{{\mbox{V}}}}{{{\mbox{O}}}}_{2}$$ V ̇ O 2 dynamics. A TCN model was developed using causal convolutions across an effective history length to model the time-dependent nature of $$\dot{\,{{\mbox{V}}}}{{{\mbox{O}}}}_{2}$$ V ̇ O 2 . Optimal history length was determined through minimum validation loss across hyperparameter values. The best performing model encoded 218 s history length (TCN-VO2 A), with 187, 97, and 76 s yielding <3% deviation from the optimal validation loss. TCN-VO2 A showed strong prediction accuracy (mean, 95% CI) across all exercise intensities (−22 ml min− 1, [−262, 218]), spanning transitions from low–moderate (−23 ml min− 1, [−250, 204]), low–high (14 ml min− 1, [−252, 280]), ventilatory threshold–high (−49 ml min− 1, [−274, 176]), and maximal (−32 ml min− 1, [−261, 197]) exercise. Second-by-second classification of physical activity across 16,090 s of predicted $$\dot{\,{{\mbox{V}}}}{{{\mbox{O}}}}_{2}$$ V ̇ O 2 was able to discern between vigorous, moderate, and light activity with high accuracy (94.1%). This system enables quantitative aerobic activity monitoring in non-laboratory settings, when combined with tidal volume and heart rate reserve calibration, across a range of exercise intensities using wearable sensors for monitoring exercise prescription adherence and personal fitness.

An optical imaging system is proposed for quantitatively assessing jugular venous response to altered central venous pressure. The proposed system assesses sub-surface optical absorption changes from jugular venous waveforms with a spatial calibration procedure to normalize incident tissue illumination. Widefield frames of the right lateral neck were captured and calibrated using a novel flexible surface calibration method. A hemodynamic optical model was derived to quantify jugular venous optical attenuation (JVA) signals, and generate a spatial jugular venous pulsatility map. JVA was assessed in three cardiovascular protocols that altered central venous pressure: acute central hypovolemia (lower body negative pressure), venous congestion (head-down tilt), and impaired cardiac filling (Valsalva maneuver). JVA waveforms exhibited biphasic wave properties consistent with jugular venous pulse dynamics when time-aligned with an electrocardiogram. JVA correlated strongly (median, interquartile range) with invasive central venous pressure during graded central hypovolemia (r=0.85, [0.72, 0.95]), graded venous congestion (r=0.94, [0.84, 0.99]), and impaired cardiac filling (r=0.94, [0.85, 0.99]). Reduced JVA during graded acute hypovolemia was strongly correlated with reductions in stroke volume (SV) (r=0.85, [0.76, 0.92]) from baseline (SV: 79$\pm$15 mL, JVA: 0.56$\pm$0.10 a.u.) to -40 mmHg suction (SV: 59$\pm$18 mL, JVA: 0.47$\pm$0.05 a.u.; p$<$0.01). The proposed non-contact optical imaging system demonstrated jugular venous dynamics consistent with invasive central venous monitoring during three protocols that altered central venous pressure. This system provides non-invasive monitoring of pressure-induced jugular venous dynamics in clinically relevant conditions where catheterization is traditionally required, enabling monitoring in non-surgical environments.

SIGNIFICANCE:Diffuse optical spectroscopic imaging (DOSI) is a versatile technology sensitive to changes in tissue composition and hemodynamics and has been used for a wide variety of clinical applications. Specific applications have prompted the development of versions of the DOSI technology to fit specific clinical needs. This work describes the development and characterization of a multi-modal DOSI (MM-DOSI) system that can acquire metabolic, compositional, and pulsatile information at multiple penetration depths in a single hardware platform. Additionally, a 3D tracking system is integrated with MM-DOSI, which enables registration of the acquired data to the physical imaging area. AIM:We demonstrate imaging, layered compositional analysis, and metabolism tracking capabilities using a single MM-DOSI system on optical phantoms as well as in vivo human tissue. APPROACH:We characterize system performance with a silicone phantom containing an embedded object. To demonstrate multi-layer sensitivity, we imaged human calf tissue with a 4.8-mm skin-adipose thickness. Human thenar tissue was also measured using a combined broadband DOSI and continuous-wave near-infrared spectroscopy method (∼15  Hz acquisition rate). RESULTS:High-resolution optical property maps of absorption (μa) and reduced scattering (μs  '  ) were recovered on the phantom by capturing over 1000 measurement points in under 5 minutes. On human calf tissue, we show two probing depth layers have significantly different (p  <  0.001) total-hemo/myoglobin and μs  '   composition. On thenar tissue, we calculate tissue arterial oxygen saturation, venous oxygen saturation, and tissue metabolic rate of oxygen consumption during baseline and after release of an arterial occlusion. CONCLUSIONS:The MM-DOSI can switch between collection of broadband spectra, high-resolution images, or multi-depth hemodynamics without any hardware reconfiguration. We conclude that MM-DOSI enables acquisition of high resolution, multi-modal data consolidated in a single platform, which can provide a more comprehensive understanding of tissue hemodynamics and composition for a wide range of clinical applications.

Diffuse optical methods have demonstrated significant potential to monitor breast cancer treatment response. Unfortunately, many diffuse optical tools involve lengthy acquisition, slow data processing, or both, limiting their clinical utility. Here, we present a new frequency domain diffuse optical device capable of providing real-time tissue oxygenation maps. This device is made possible through the integration of three enabling technologies: real-time 3D probe tracking, rapid translation of raw data to optical properties, and new high-speed acquisition electronics. We have demonstrated the ability of this instrument to identify inclusions in phantoms as well as to make proof-of-concept measurements in breast tissue.

Background & Purpose: The femoral bifurcation houses complex flow recirculation that may be impacted by downstream peripheral resistance; however, conventional ultrasound is unable to quantify multi-directional flow to study these phenomena. The purpose of this study was to examine the multi-directional behavior of blood in the femoral bifurcation during reactive hyperemia (RH) using vector flow imaging. Methods: Nine healthy adults (25 ± 4 years; 5 men) underwent high-frame-rate ultrasound imaging of their right femoral bifurcation at rest, and during the first cardiac cycle of RH after 5-min of 200 mmHg calf occlusion. Multi-angle Doppler ultrasound was used to qualitatively describe vector blood flow at 750 fps [1]. The bifurcation region was isolated for blood vector dispersion analysis, an index of flow directionality [2]. Results: Blood flow during RH was characterized by near-constant anterograde blood flow to the superficial femoral artery (SFA), adjacent to low/oscillatory flow immediately proximal to the deep femoral artery (DFA) (Figure 1). We observed collateral blood flow redistribution in 3/9 participants, where blood from the DFA appeared to re-enter the SFA during the retrograde flow period. Bifurcation vector dispersion was reduced from baseline (0.53 ± 0.08 vs 0.67 ± 0.09; p < 0.01), but not eliminated, due to preservation of DFA flow reversal. Conclusion: We demonstrate the presence of distinct, and potentially interacting, blood flow streams within the femoral bifurcation during RH. These observations were made possible by the emerging vector flow imaging technique to study complex hemodynamics in vivo, and suggest novel mechanisms by which blood is distributed to vascular beds, in extreme flow profiles. Figure 1Left: Representative example (31 yr old man) of vector projectile imaging of the femoral bifurcation at end-systole; vectors within the red region of interest were extracted to calculate vector dispersion. Right: Representative example of blood vector dispersion at rest and during hyperemia during a single cardiac cycle; values closer to 0 indicate greater dispersion.

Postural instability is prevalent in aging and neurodegenerative disease, decreasing quality of life and independence. Quantitatively monitoring balance control is important for assessing treatment efficacy and rehabilitation progress. However, existing technologies for assessing postural sway are complex and expensive, limiting their widespread utility. Here, we propose a monocular imaging system capable of assessing sub-millimeter 3D sway dynamics during quiet standing. Two anatomical targets with known feature geometries were placed on the lumbar and shoulder. Upper and lower trunk 3D kinematic motion was automatically assessed from a set of 2D frames through geometric feature tracking and an inverse motion model. Sway was tracked in 3D and compared between control and hypoperfusion conditions in 14 healthy young adults. The proposed system demonstrated high agreement with a commercial motion capture system (error $1.5 \times 10^{-4}~\text{mm}$, [$-0.52$, $0.52$]). Between-condition differences in sway dynamics were observed in anterior-posterior sway during early and mid stance, and medial-lateral sway during mid stance commensurate with decreased cerebral perfusion, followed by recovered sway dynamics during late stance with cerebral perfusion recovery. This inexpensive single-camera system enables quantitative 3D sway monitoring for assessing neuromuscular balance control in weakly constrained environments.

Pulse arrival time (PAT) is a method used to estimate systolic blood pressure (SBP), as the amount of time it takes for a pulse to travel from the heart to a peripheral location is inversely related to blood pressure. However, the validity and accuracy of this method’s blood pressure estimation has been questioned. Therefore, the purpose of this study was to evaluate the accuracy of a PAT model’s estimation of SBP over a range of exercise intensities. Six participants (5 men, 1 woman; age: 26 ± 4 yrs) completed three cycling exercise tests (25 Watt/min ramp incremental test, and moderate and heavy pseudorandom binary sequence exercise) during three separate laboratory visits. PAT was calculated as the time difference between the R‐wave of an electrocardiogram and the pulse arrival at the forehead measured with a pulse oximeter. SBP was estimated using the following equation: SBPPAT = 64.578/(PAT) – 43.957, and was compared to calibrated brachial blood pressure measured at the finger (FBP) by photoplethysmography during the exercise trials. Participant blood pressure responses were pooled, and Bland‐Altman analysis was conducted to evaluate the accuracy of estimated SBP for each exercise test. PAT model estimation of SBP had the smallest bias during heavy (−2.8 mmHg), then moderate (8.1 mmHg), and the largest bias during ramp incremental exercise (−11 mmHg), with their limits of agreement being between −56 and 50 mmHg, −29 and 45 mmHg, and −74 and 53 mmHg, respectively. Linear regression of the ramp incremental SBP response revealed that SBP estimated from the PAT does not increase by the same magnitude as the pressure measured at the finger (SBPPAT = 0.295 · SBPFBP + 106.236, r2 = 0.350, p < 0.05). Overall, these findings support that SBP estimated by PAT is linearly related to brachial SBP measured at the finger; however, this PAT model is a poor estimator of absolute SBP during exercise ranging from moderate to maximal intensities.Support or Funding InformationSupported by NSERC.

Malnutrition impacts quality of life and places annually-recurring burden on the health care system. Half of older adults are at risk for malnutrition in long-term care (LTC). Monitoring and measuring nutritional intake is paramount yet involves time-consuming and subjective visual assessment, limiting current methods' reliability. The opportunity for automatic image-based estimation exists. Some progress outside LTC has been made (e.g., calories consumed, food classification), however, these methods have not been implemented in LTC, potentially due to a lack of ability to independently evaluate automatic segmentation methods within the intake estimation pipeline. Here, we propose and evaluate a novel fully-automatic semantic segmentation method for pixel-level classification of food on a plate using a deep convolutional neural network (DCNN). The macroarchitecture of the DCNN is a multi-scale encoder-decoder food network (EDFN) architecture comprising a residual encoder microarchitecture, a pyramid scene parsing decoder microarchitecture, and a specialized per-pixel food/no-food classification layer. The network was trained and validated on the pre-labelled UNIMIB 2016 food dataset (1027 tray images, 73 categories), and tested on our novel LTC plate dataset (390 plate images, 9 categories). Our fully-automatic segmentation method attained similar intersection over union to the semi-automatic graph cuts (91.2% vs. 93.7%). Advantages of our proposed system include: testing on a novel dataset, decoupled error analysis, no user-initiated annotations, with similar segmentation accuracy and enhanced reliability in terms of types of segmentation errors. This may address several short-comings currently limiting utility of automated food intake tracking in time-constrained LTC and hospital settings.

The cardiovascular system is a complex dynamic system whose function changes with various external and internal stimuli. Cardiovascular monitoring has widespread clinical and preclinical use for assessing cardiovascular function and monitoring disease states. Local hemodynamics provides information such as perfusion sufficiency, blood pressure, and vessel dynamics. Vascular hemodynamics (i.e., the dynamics of blood flow) can provide important insight into cardiovascular health. Hemodynamic imaging extends on single-point probe sensors by enabling hemodynamic assessment over large tissue regions in a noncontact manner, providing information about spatial perfusion and blood flow noninvasively. This section provides an overview of the primary methods for hemodynamic monitoring. Starting with ultrasound, which has been considered a gold standard modality for vascular imaging, this section explores newer biophotonic imaging technologies that overcome many of ultrasound's limitations. Two laser-based systems that have gained clinical attention are laser Doppler imaging and laser speckle imaging, both of which use coherent laser light sources to monitor blood flow through fundamental light–blood interaction. Lastly, a newer modality, photoplethysmographic imaging, is discussed, which uses diffuse noncoherent light sources for assessing vascular pulsatility across a wide field of view and overcomes some of the fundamental limitations of laser-based modalities.