Investigating Personalization Techniques for Improved Cybersickness Prediction in Virtual Reality Environments.

In recent cybersickness research, there has been a growing interest in predicting cybersickness using real-time physiological data such as heart rate, galvanic skin response, eye tracking, postural sway, and electroencephalogram. However, the impact of individual factors such as age and gender, which are pivotal in determining cybersickness susceptibility, remains unknown in predictive models. Our research seeks to address this gap, underscoring the necessity for a more personalized approach to cybersickness prediction to ensure a better, more inclusive virtual reality experience. We hypothesize that a personalized cybersickness prediction model would outperform non-personalized models in predicting cybersickness. Evaluating this, we explored four personalization techniques: 1) data grouping, 2) transfer learning, 3) early shaping, and 4) sample weighing using an open-source cybersickness dataset. Our empirical results indicate that personalized models significantly improve prediction accuracy. For instance, with early shaping, the Deep Temporal Convolutional Neural Network (DeepTCN) model achieved a 69.7% reduction in RMSE compared to its non-personalized version. Our study provides evidence of personalization techniques' benefits in improving cybersickness prediction. These findings have implications for developing personalized cybersickness prediction models tailored to individual differences, which can be used to develop personalized cybersickness reduction techniques in the future.

Fully Automated Artificial Intelligence Assessment of Aortic Stenosis by Echocardiography.

BACKGROUND: Aortic stenosis (AS) is a common form of valvular heart disease, present in over 12% of the population age 75 years and above. Transthoracic echocardiography (TTE) is the first line of imaging in the adjudication of AS severity but is time-consuming and requires expert sonographic and interpretation capabilities to yield accurate results. Artificial intelligence (AI) technology has emerged as a useful tool to address these limitations but has not yet been applied in a fully hands-off manner to evaluate AS. Here, we correlate artificial neural network measurements of key hemodynamic AS parameters to experienced human reader assessment. METHODS: Two-dimensional and Doppler echocardiographic images from patients with normal aortic valves and all degrees of AS were analyzed by an artificial neural network (Us2.ai) with no human input to measure key variables in AS assessment. Trained echocardiographers blinded to AI data performed manual measurements of these variables, and correlation analyses were performed. RESULTS: Our cohort included 256 patients with an average age of 67.6 ± 9.5 years. Across all AS severities, AI closely matched human measurement of aortic valve peak velocity (r = 0.97, P < .001), mean pressure gradient (r = 0.94, P < .001), aortic valve area by continuity equation (r = 0.88, P < .001), stroke volume index (r = 0.79, P < .001), left ventricular outflow tract velocity-time integral (r = 0.89, P < .001), aortic valve velocity-time integral (r = 0.96, P < .001), and left ventricular outflow tract diameter (r = 0.76, P < .001). CONCLUSIONS: Artificial neural networks have the capacity to closely mimic human measurement of all relevant parameters in the adjudication of AS severity. Application of this AI technology may minimize interscan variability, improve interpretation and diagnosis of AS, and allow for precise and reproducible identification and management of patients with AS.

Can Hierarchical Transformers Learn Facial Geometry?

Human faces are a core part of our identity and expression, and thus, understanding facial geometry is key to capturing this information. Automated systems that seek to make use of this information must have a way of modeling facial features in a way that makes them accessible. Hierarchical, multi-level architectures have the capability of capturing the different resolutions of representation involved. In this work, we propose using a hierarchical transformer architecture as a means of capturing a robust representation of facial geometry. We further demonstrate the versatility of our approach by using this transformer as a backbone to support three facial representation problems: face anti-spoofing, facial expression representation, and deepfake detection. The combination of effective fine-grained details alongside global attention representations makes this architecture an excellent candidate for these facial representation problems. We conduct numerous experiments first showcasing the ability of our approach to address common issues in facial modeling (pose, occlusions, and background variation) and capture facial symmetry, then demonstrating its effectiveness on three supplemental tasks.

Personalized 3D exergames for in-home rehabilitation after stroke: a pilot study.

PURPOSE: To describe a novel 3-dimensional (3D) exergames system and the results of a clinical feasibility study of stroke survivors needing in-home rehabilitation. MATERIALS AND METHODS: The customisable Personalized In-home eXErgames for Rehabilitation (PIXER) system captures the user's image, generates a live model, and incorporates it into a virtual exergame. PIXER provides a recording system for home exercise programs (HEPs) by adapting virtual objects, customizes the exergame and creates a digital diary. Ten persons with stroke, performed HEPs with PIXER for 1 month, and without PIXER for 2 additional months. In-game performance data, measures of physical functioning (PF) including Stroke Impact Scale (SIS), Timed Up & Go (TUG) and Goal Attainment (GA) Scale obtained at baseline, 1- and 3 months were evaluated. RESULTS: Seventy percent of participants completed the 1-month timepoint, 50% completed all timepoints. In-game data: Number of repetitions completed; Anomalies reported; and Percentage of bubbles hit showed positive trends. Compared to baseline, all SIS physical functioning (PF) scores were higher at 1 month, TUG scores showed no overall improvement and GA scale scores were 77% at 3 months. CONCLUSION: It is feasible for community-dwelling patients to perform HEP after stroke using PIXER, a novel, exergames system, and potentially improve their function.IMPLICATIONS FOR REHABILITATIONHome Exercises performed using a novel, 3-dimensional, customizable Personalized In-home eXErgames for Rehabilitation (PIXER) system is feasible for community-dwelling patients after stroke.In-game performance data obtained in this clinical pilot study showed positive trends of improvement in several study participants.PIXER has potential to improve functional outcomes for community-dwelling adults with stroke.

Application of a Novel Two-Dimensional Echocardiographic Calcium Quantification Method to Assess All Severities of Aortic Stenosis.

BACKGROUND: Aortic valve (AV) calcification (AVC) is a strong predictor of aortic stenosis (AS) severity. The two-dimensional AVC (2D-AVC) ratio, a gain-independent ratio composed of the average pixel density of the AV and the aortic annulus, has previously shown strong correlations with two-dimensional (2D) echocardiographic hemodynamic parameters for severe AS and AVC by cardiac computed tomography. We hypothesize that the 2D-AVC ratio correlates with hemodynamic parameters in all severities of AS. METHODS: A total of 285 patients with a normal AV (n = 49), aortic sclerosis (n = 75), or mild (n = 38), moderate (n = 72), or severe (n = 51) AS undergoing 2D echocardiography were retrospectively evaluated, and the 2D-AVC ratios were correlated to mean AV gradient, peak AV velocity, AV area, and dimensionless index. The 2D-AVC ratios of various AS severities were compared against each other via area under the curve (AUC) analysis. RESULTS: The 2D-AVC ratio is strongly correlated with mean AV gradient (r = 0.79, P < .0001) and peak AV velocity (r = 0.78, P < .0001). There was moderate correlation with the AV area (r = -0.58, P < .0001) and dimensionless index (r = -0.67, P < .0001) across all AS severities. The 2D-AVC ratio also distinguished nonmoderate AS (mild AS + normal AV) from moderate or greater (moderate + severe) AS (AUC = 0.93) and moderate versus severe AS (AUC = 0.88). CONCLUSION: The 2D-AVC ratio exhibits moderate to strong correlation with 2D echocardiographic hemodynamic parameters across all severities of AS.

Elevated AT1R Antibody and Morbidity in Patients Bridged to Heart Transplant Using Continuous Flow Left Ventricular Assist Devices.

BACKGROUND: We studied longitudinal levels of angiotensin-II type 1 receptor antibody (AT1R-Ab) and their effects on adverse events (death, treated rejection and cardiac allograft vasculopathy) in patients who were bridged to heart transplant using a continuous flow left ventricular assist device (LVAD). METHODS AND RESULTS: Sera of 77 patients bridged to heart transplant (from 2009 to 2017) were tested for AT1R-Ab and CRP before and after LVAD. Elevated AT1R-Ab was defined as >10.0 U/mL. The median follow-up after transplant was 3.6 years (interquartile range, 2.2-5.6 years). After LVAD, AT1R-Ab levels increased from baseline and remained elevated until transplant. Freedom from adverse events at 5 years was lower in those with elevated AT1R-Ab levels at time of transplant. In an adjusted, multivariable Cox analysis, an AT1R-Ab level of >10 U/mL was associated with developing the primary end point (adjusted hazard ratio 3.4, 95% confidence interval 1.2-9.2, P = .017). Although C-reactive protein levels were high before and after LVAD placement, C-reactive protein did not correlate with AT1R-Ab. CONCLUSIONS: In LVAD patients bridged to heart transplant, an increased AT1R-Ab level at time of transplant was associated with poor outcomes after heart transplant. Post-LVAD AT1R-Ab elevations were not correlated with serum markers of systemic inflammation. Larger studies are needed to examine the pathologic role of AT1R-Ab in heart transplant.

Predictors of Renal Failure in Patients Treated With the Total Artificial Heart.

BACKGROUND: The incidence of hemodialysis (HD)-dependent renal failure after total artificial heart (TAH) implantation is high. We sought to determine the preoperative predictors of HD after TAH implantation. METHODS AND RESULTS: We studied 87 patients after TAH implantation at our institution between April 2006 and March 2017. Baseline clinical data were obtained from the medical records, and patients were followed until death or heart transplantation. We performed logistic regression analysis to identify predictors of HD after TAH implantation. Of the patients, 24 (28%) required postimplantation HD. Those requiring HD were more likely to have histories of coronary artery disease (58% vs 29%; P = 0.01), required preoperative membrane oxygenation (33% vs 4.8%; P = 0.001) and had lower baseline estimated glomerular filtration rates (54 ± 29 vs 67 ± 24 mL/min/1.73m2; P = 0.04). Patients requiring HD were at a higher risk of death on device at 1 year (33% vs 5%, P = 0.001; log rank test: P =0.001, hazard ratio 6.6 [95% CI:1.8-23], P = 0.003). CONCLUSIONS: The incidence of postimplantation HD is high and is associated with increased likelihood of mortality. Lower baseline estimated glomerular filtration rates, histories of coronary artery disease and preoperative membrane oxygenation support are predictors of postimplantation requirement of HD. These data may help to identify patients at risk for adverse outcomes after TAH implantation.

Left Ventricular Assist Device Outflow Graft Compression: Incidence, Clinical Associations and Potential Etiologies.

BACKGROUND: Left ventricular assist devices (LVADs) have revolutionized the treatment of advanced heart failure, but proliferation of device therapy has unmasked potential complications. Reports have emerged of outflow graft narrowing due to extrinsic compression. METHODS AND RESULTS: The records of patients with LVADs that had been implanted at our institution were reviewed. Those who had postimplantation computed tomography angiographies sufficient to analyze the outflow graft lumen were identified, and the studies were analyzed to characterize the outflow graft lumen. We identified 241 patients; 110 (46%) had suitable computed tomography angiographies. Of those, 15 (14%) had evidence of outflow graft lumen narrowing, all in HeartMate devices and all within the portion covered by the bend relief. Of the 15, 3 underwent invasive examination, all without intraluminal thrombus but, rather, with biodebris between the bend relief and the outflow graft. Patients with HeartWare devices had a wide range of biodebris accumulation surrounding the outflow graft but no cases of lumen narrowing. On multivariable analysis, 1) time from device implant to scan, 2) nonischemic cardiomyopathy and 3) age at implant were significantly associated with higher risk of graft narrowing. CONCLUSION: Outflow graft narrowing can be seen in a number of patients with HeartMate LVADs within the portion covered by the bend relief. In the limited number of patients who underwent invasive evaluation, the narrowing was found to arise from extrinsic compression rather than intraluminal thrombus. The clinical significance of this requires further investigation.

Coevolution of RtcB and Archease created a multiple-turnover RNA ligase.

RtcB is a noncanonical RNA ligase that joins either 2',3'-cyclic phosphate or 3'-phosphate termini to 5'-hydroxyl termini. The genes encoding RtcB and Archease constitute a tRNA splicing operon in many organisms. Archease is a cofactor of RtcB that accelerates RNA ligation and alters the NTP specificity of the ligase from Pyrococcus horikoshii. Yet, not all organisms that encode RtcB also encode Archease. Here we sought to understand the differences between Archease-dependent and Archease-independent RtcBs so as to illuminate the evolution of Archease and its function. We report on the Archease-dependent RtcB from Thermus thermophilus and the Archease-independent RtcB from Thermobifida fusca. We find that RtcB from T. thermophilus can catalyze multiple turnovers only in the presence of Archease. Remarkably, Archease from P. horikoshii can activate T. thermophilus RtcB, despite low sequence identity between the Archeases from these two organisms. In contrast, RtcB from T. fusca is a single-turnover enzyme that is unable to be converted into a multiple-turnover ligase by Archease from either P. horikoshii or T. thermophilus. Thus, our data indicate that Archease likely evolved to support multiple-turnover activity of RtcB and that coevolution of the two proteins is necessary for a functional interaction.

Structure of RNA 3'-phosphate cyclase bound to substrate RNA.

RNA 3'-phosphate cyclase (RtcA) catalyzes the ATP-dependent cyclization of a 3'-phosphate to form a 2',3'-cyclic phosphate at RNA termini. Cyclization proceeds through RtcA-AMP and RNA(3')pp(5')A covalent intermediates, which are analogous to intermediates formed during catalysis by the tRNA ligase RtcB. Here we present a crystal structure of Pyrococcus horikoshii RtcA in complex with a 3'-phosphate terminated RNA and adenosine in the AMP-binding pocket. Our data reveal that RtcA recognizes substrate RNA by ensuring that the terminal 3'-phosphate makes a large contribution to RNA binding. Furthermore, the RNA 3'-phosphate is poised for in-line attack on the P-N bond that links the phosphorous atom of AMP to N(ε) of His307. Thus, we provide the first insights into RNA 3'-phosphate termini recognition and the mechanism of 3'-phosphate activation by an Rtc enzyme.

A tRNA splicing operon: Archease endows RtcB with dual GTP/ATP cofactor specificity and accelerates RNA ligation.

Archease is a 16-kDa protein that is conserved in all three domains of life. In diverse bacteria and archaea, the genes encoding Archease and the tRNA ligase RtcB are localized into an operon. Here we provide a rationale for this operon organization by showing that Archease and RtcB from Pyrococcus horikoshii function in tandem, with Archease altering the catalytic properties of the RNA ligase. RtcB catalyzes the GTP and Mn(II)-dependent joining of either 2',3'-cyclic phosphate or 3'-phosphate termini to 5'-hydroxyl termini. We find that catalytic concentrations of Archease are sufficient to activate RtcB, and that Archease accelerates both the RNA 3'-P guanylylation and ligation steps. In addition, we show that Archease can alter the NTP specificity of RtcB such that ATP, dGTP or ITP is used efficiently. Moreover, RtcB variants that have inactivating substitutions in the guanine-binding pocket can be rescued by the addition of Archease. We also present a 1.4 Å-resolution crystal structure of P. horikoshii Archease that reveals a metal-binding site consisting of conserved carboxylates located at the protein tip. Substitution of the Archease metal-binding residues drastically reduced Archease-dependent activation of RtcB. Thus, evolution has sought to co-express archease and rtcB by creating a tRNA splicing operon.

Structures of the noncanonical RNA ligase RtcB reveal the mechanism of histidine guanylylation.

RtcB is an atypical RNA ligase that joins either 2',3'-cyclic phosphate or 3'-phosphate termini to 5'-hydroxyl termini. In contrast to typical RNA ligases, which rely on ATP and Mg(II), catalysis by RtcB is dependent on GTP and Mn(II) with ligation proceeding through a covalent RtcB-histidine-GMP intermediate. Here, we present three structures of Pyrococcus horikoshii RtcB complexes that capture snapshots along the entire guanylylation pathway. These structures show that prior to binding GTP, a single manganese ion (Mn1) is bound to RtcB. To capture the step immediately preceding RtcB guanylylation, we determined a structure of RtcB in complex with Mn(II) and the unreactive GTP analogue guanosine 5'-(α-thio)triphosphate (GTPαS). This structure shows that Mn1 is poised to stabilize the pentavalent transition state of guanylylation while a second manganese ion (Mn2) is coordinated to a nonbridging oxygen of the γ-phosphoryl group. The pyrophosphate leaving group of GTPαS is oriented apically to His404 with the ε-nitrogen poised for in-line attack on the α-phosphorus atom. The structure of RtcB in complex with GTPαS also reveals the network of hydrogen bonds that recognize GTP and illuminates the significant conformational changes that accompany the binding of this cofactor. Finally, a structure of the enzymic histidine-GMP intermediate depicts the end of the guanylylation pathway. The ensuing molecular description of the RtcB guanylylation pathway shows that RtcB and classical ATP- and Mg(II)-dependent nucleic acid ligases have converged upon a similar two-metal mechanism for formation of the nucleotidylated enzyme intermediate.

tRNA ligase catalyzes the GTP-dependent ligation of RNA with 3'-phosphate and 5'-hydroxyl termini.

The RNA ligase RtcB is conserved in all domains of life and is essential for tRNA maturation in archaea and metazoa. Here we show that bacterial and archaeal RtcB catalyze the GTP-dependent ligation of RNA with 3'-phosphate and 5'-hydroxyl termini. Reactions with analogues of RNA and GTP suggest a mechanism in which RtcB heals the 3'-phosphate terminus by forming a 2',3'-cyclic phosphate before joining it to the 5'-hydroxyl group of a second RNA strand. Thus, RtcB can ligate RNA cleaved by RNA endonucleases, which generate 2',3'-cyclic phosphate and then 3'-phosphate termini on one strand, and a 5'-hydroxyl terminus on another strand.

Recruitment of genes and enzymes conferring resistance to the nonnatural toxin bromoacetate.

Microbial niches contain toxic chemicals capable of forcing organisms into periods of intense natural selection to afford survival. Elucidating the mechanisms by which microbes evade environmental threats has direct relevance for understanding and combating the rise of antibiotic resistance. In this study we used a toxic small-molecule, bromoacetate, to model the selective pressures imposed by antibiotics and anthropogenic toxins. We report the results of genetic selection experiments that identify nine genes from Escherichia coli whose overexpression affords survival in the presence of a normally lethal concentration of bromoacetate. Eight of these genes encode putative transporters or transmembrane proteins, while one encodes the essential peptidoglycan biosynthetic enzyme, UDP-N-acetylglucosamine enolpyruvoyl transferase (MurA). Biochemical studies demonstrate that the primary physiological target of bromoacetate is MurA, which becomes irreversibly inactivated via alkylation of a critical active-site cysteine. We also screened a comprehensive library of E. coli single-gene deletion mutants and identified 63 strains displaying increased susceptibility to bromoacetate. One hypersensitive bacterium lacks yliJ, a gene encoding a predicted glutathione transferase. Herein, YliJ is shown to catalyze the glutathione-dependent dehalogenation of bromoacetate with a k(cat)/K(m) value of 5.4 × 10(3) M(-1) s(-1). YliJ displays exceptional substrate specificity and produces a rate enhancement exceeding 5 orders of magnitude, remarkable characteristics for reactivity with a nonnatural molecule. This study illustrates the wealth of intrinsic survival mechanisms that can be exploited by bacteria when they are challenged with toxins.

L-glyceraldehyde 3-phosphate reductase from Escherichia coli is a heme binding protein.

Recently, we reported that YghZ from Escherichia coli functions as an efficient L-glyceraldehyde 3-phosphate reductase (Gpr). Here we show that Gpr co-purifies with a b-type heme cofactor. Gpr associates with heme in a 1:1 stoichiometry to form a complex that is characterized by a K(d) value of 5.8+/-0.2 microM in the absence of NADPH and a K(d) value of 11+/-1.3 microM in the presence of saturating NADPH. The absorbance spectrum of reconstituted Gpr indicates that heme is bound in a hexacoordinate low-spin state under both oxidizing and reducing conditions. The physiological function of heme association with Gpr is unclear, as the L-glyceraldehyde 3-phosphate reductase activity of Gpr does not require the presence of the cofactor. Bioinformatics analysis reveals that Gpr clusters with a family of putative monooxygenases in several organisms, suggesting that Gpr may act as a heme-dependent monooxygenase. The discovery that Gpr associates with heme is interesting because Gpr shares 35% amino acid identity with the mammalian voltage-gated K+ channel beta-subunit, an NADPH-dependent oxidoreductase that endows certain voltage-gated K+ channels with hemoprotein-like, O2-sensing properties. To date the molecular origin of O2 sensing by voltage-gated K+ channels is unknown and the results presented herein suggest a role for heme in this process.

A metabolic bypass of the triosephosphate isomerase reaction.

Triosephosphate isomerase (TIM) catalyzes the interconversion of d-glyceraldehyde 3-phosphate and dihydroxyacetone phosphate, an essential step in glycolytic and gluconeogenic metabolism. To uncover promiscuous isomerases embedded within the Escherichia coli genome, we searched for genes capable of restoring growth of a TIM-deficient bacterium under gluconeogenic conditions. Rather than discovering an isomerase, we selected yghZ, a gene encoding a member of the aldo-keto reductase superfamily. Here we show that YghZ catalyzes the stereospecific, NADPH-dependent reduction of l-glyceraldehyde 3-phosphate, the enantiomer of the TIM substrate. This transformation provides an alternate pathway to the formation of dihydroxyacetone phosphate.

Coordinated peak expression of MMP-26 and TIMP-4 in preinvasive human prostate tumor.

The identification of novel biomarkers for early prostate cancer diagnosis is highly important because early detection and treatment are critical for the medical management of patients. Disruption in the continuity of both the basal cell layer and basement membrane is essential for the progression of high-grade prostatic intraepithelial neoplasia (HGPIN) to invasive adenocarcinoma in human prostate. The molecules involved in the conversion to an invasive phenotype are the subject of intense scrutiny. We have previously reported that matrix metalloproteinase-26 (MMP-26) promotes the invasion of human prostate cancer cells via the cleavage of basement membrane proteins and by activating the zymogen form of MMP-9. Furthermore, we have found that tissue inhibitor of metalloproteinases-4 (TIMP-4) is the most potent endogenous inhibitor of MMP-26. Here we demonstrate higher (p<0.0001) MMP-26 and TIMP-4 expression in HGPIN and cancer, compared to non-neoplastic acini. Their expression levels are highest in HGPIN, but decline in invasive cancer (p<0.001 for each) in the same tissues. Immunohistochemical staining of serial prostate cancer tissue sections suggests colocalization of MMP-26 and TIMP-4. The present study indicates that MMP-26 and TIMP-4 may play an integral role during the conversion of HGPIN to invasive cancer and may also serve as markers for early prostate cancer diagnosis.