Telemedicine & Advanced Technology Research Center

In Brief Summary Background Data: To inform surgeons about the practical issues to be considered for successful integration of virtual reality simulation into a surgical training program. The learning and practice of minimally invasive surgery (MIS) makes unique demands on surgical training programs. A decade ago Satava proposed virtual reality (VR) surgical simulation as a solution for this problem. Only recently have robust scientific studies supported that vision Methods: A review of the surgical education, human-factor, and psychology literature to identify important factors which will impinge on the successful integration of VR training into a surgical training program. Results: VR is more likely to be successful if it is systematically integrated into a well-thought-out education and training program which objectively assesses technical skills improvement proximate to the learning experience. Validated performance metrics should be relevant to the surgical task being trained but in general will require trainees to reach an objectively determined proficiency criterion, based on tightly defined metrics and perform at this level consistently. VR training is more likely to be successful if the training schedule takes place on an interval basis rather than massed into a short period of extensive practice. High-fidelity VR simulations will confer the greatest skills transfer to the in vivo surgical situation, but less expensive VR trainers will also lead to considerably improved skills generalizations. Conclusions: VR for improved performance of MIS is now a reality. However, VR is only a training tool that must be thoughtfully introduced into a surgical training curriculum for it to successfully improve surgical technical skills. Evidence has been published demonstrating the power of simulation for training surgical skills. Simulation training is more likely to produce better training outcomes if it is systematically integrated into the curriculum of a training program with proficiency-based progression founded on objective feedback with validated metrics proximate to performance.

The Y-balance test (YBT) is one of the few field expedient tests that have shown predictive validity for injury risk in an athletic population. However, analysis of the YBT in a heterogeneous population of active adults (e.g., military, specific occupations) involving multiple raters with limited experience in a mass screening setting is lacking. The primary purpose of this study was to determine interrater test-retest reliability of the YBT in a military setting using multiple raters. Sixty-four service members (53 males, 11 females) actively conducting military training volunteered to participate. Interrater test-retest reliability of the maximal reach had intraclass correlation coefficients (2,1) of 0.80 to 0.85 with a standard error of measurement ranging from 3.1 to 4.2 cm for the 3 reach directions (anterior, posteromedial, and posterolateral). Interrater test-retest reliability of the average reach of 3 trails had an intraclass correlation coefficients (2,3) range of 0.85 to 0.93 with an associated standard error of measurement ranging from 2.0 to 3.5cm. The YBT showed good interrater test-retest reliability with an acceptable level of measurement error among multiple raters screening active duty service members. In addition, 31.3% (n = 20 of 64) of participants exhibited an anterior reach asymmetry of >4cm, suggesting impaired balance symmetry and potentially increased risk for injury.

In drug development, early assessments of pharmacokinetic and toxic properties are important stepping stones to avoid costly and unnecessary failures. Considerable progress has recently been made in the development of computer-based (in silico) models to estimate such properties. Nonetheless, such models can be further improved in terms of their ability to make predictions more rapidly, easily, and with greater reliability. To address this issue, we have used our vNN method to develop 15 absorption, distribution, metabolism, excretion, and toxicity (ADMET) prediction models. These models quickly assess some of the most important properties of potential drug candidates, including their cytotoxicity, mutagenicity, cardiotoxicity, drug-drug interactions, microsomal stability, and likelihood of causing drug-induced liver injury. Here we summarize the ability of each of these models to predict such properties and discuss their overall performance. Using a 10-fold cross-validation procedure, we show that each model has an accuracy of greater than 75%. All of these ADMET models are publically available on our website (https://vnnadmet.bhsai.org/), which also offers the capability of using the vNN method to customize and build new models.

Peer Reviewed

The laboratory rat has been used as a surrogate to study human biology for more than a century. Here we present the first genome-scale network reconstruction of Rattus norvegicus metabolism, iRno, and a significantly improved reconstruction of human metabolism, iHsa. These curated models comprehensively capture metabolic features known to distinguish rats from humans including vitamin C and bile acid synthesis pathways. After reconciling network differences between iRno and iHsa, we integrate toxicogenomics data from rat and human hepatocytes, to generate biomarker predictions in response to 76 drugs. We validate comparative predictions for xanthine derivatives with new experimental data and literature-based evidence delineating metabolite biomarkers unique to humans. Our results provide mechanistic insights into species-specific metabolism and facilitate the selection of biomarkers consistent with rat and human biology. These models can serve as powerful computational platforms for contextualizing experimental data and making functional predictions for clinical and basic science applications.

This paper presents a continuous-flow microfluidic device for sorting stem cells and their differentiation progenies. The principle of the device is based on the accumulation of multiple dielectrophoresis (DEP) forces to deflect cells laterally in conjunction with the alternating on/off electric field to manipulate the cell trajectories. The microfluidic device containing a large array of oblique interdigitated electrodes was fabricated using a combination of standard and soft lithography techniques to generate a PDMS-gold electrode construct. Experimental testing with human mesenchymal stem cells (hMSC) and their differentiation progenies (osteoblasts) was carried out at different flow rates, and clear separation of the two populations was achieved. Most of the osteoblasts experiencing stronger DEP forces were deflected laterally and continuously, following zig-zag trajectories, and moved towards the desired collection outlet, whereas most of the hMSCs remained on the original trajectory due to weaker DEP forces. The experimental measurements were characterized and evaluated quantitatively, and consistent performance was demonstrated. Collection efficiency up to 92% and 67% for hMSCs and osteoblasts, respectively, along with purity up to 84% and 87% was obtained. The experimental results established the feasibility of our microfluidic DEP sorting device for continuous, label-free sorting of stem cells and their differentiation progenies.

Using a personal computer (PC) for simple visual reaction time testing is advantageous because of the relatively low hardware cost, user familiarity, and the relative ease of software development for specific neurobehavioral testing protocols. However, general-purpose computers are not designed with the millisecond-level accuracy of operation required for such applications. Software that does not control for the various sources of delay may return reaction time values that are substantially different from the true reaction times. We have developed and characterized a freely available system for PC-based simple visual reaction time testing that is analogous to the widely used psychomotor vigilance task (PVT). In addition, we have integrated individualized prediction algorithms for near-real-time neurobehavioral performance prediction. We characterized the precision and accuracy with which the system as a whole measures reaction times on a wide range of computer hardware configurations, comparing its performance with that of the "gold standard" PVT-192 device. We showed that the system is capable of measuring reaction times with an average delay of less than 10 ms, a margin of error that is comparable to that of the gold standard. The most critical aspect of hardware selection is the type of mouse used for response detection, with gaming mice showing a significant advantage over standard ones. The software is free to download from http://bhsai.org/downloads/pc-pvt/ .

In vitro selection of RNA aptamers that bind to a specific ligand usually begins with a random pool of RNA sequences. We propose a computational approach for designing a starting pool of RNA sequences for the selection of RNA aptamers for specific analyte binding. Our approach consists of three steps: (i) selection of RNA sequences based on their secondary structure, (ii) generating a library of three-dimensional (3D) structures of RNA molecules and (iii) high-throughput virtual screening of this library to select aptamers with binding affinity to a desired small molecule. We developed a set of criteria that allows one to select a sequence with potential binding affinity from a pool of random sequences and developed a protocol for RNA 3D structure prediction. As verification, we tested the performance of in silico selection on a set of six known aptamer-ligand complexes. The structures of the native sequences for the ligands in the testing set were among the top 5% of the selected structures. The proposed approach reduces the RNA sequences search space by four to five orders of magnitude--significantly accelerating the experimental screening and selection of high-affinity aptamers.

The science of virtual reality provides an entirely new opportunity in the area of simulation of surgical skills using computers for training, evaluation, and eventually certification. A taxonomy of the types of simulators is proposed based upon the level of complexity of the task which is being simulated. These tasks are precision placement, simple manipulation, complex manipulation, and integrated procedure. Representative simulators in each category are illustrated and discussed in the context of their contribution to the education and training of a surgeon. The importance of a curriculum is to give content to the role of simulators as another advanced tool for education. Simulators must be integrated into a comprehensive curriculum and not considered as a stand-alone system. The current accomplishments as well as challenges are discussed.

Table of Contents PREAMBLE SCOPE INTRODUCTION Internet-Based Telemental Health Models of Care Today CLINICAL GUIDELINES A. Professional and Patient Identity and Location 1. Provider and Patient Identity Verification 2. Provider and Patient Location Documentation 3. Contact Information Verification for Professional and Patient 4. Verification of Expectations Regarding Contact Between Sessions B. Patient Appropriateness for Videoconferencing-Based Telemental Health 1. Appropriateness of Videoconferencing in Settings Where Professional Staff Are Not Immediately Available C. Informed Consent D. Physical Environment E. Communication and Collaboration with the Patient's Treatment Team F. Emergency Management 1. Education and Training 2. Jurisdictional Mental Health Involuntary Hospitalization Laws 3. Patient Safety When Providing Services in a Setting with Immediately Available Professionals 4. Patient Safety When Providing Services in a Setting Without Immediately Available Professional Staff 5. Patient Support Person and Uncooperative Patients 6. Transportation 7. Local Emergency Personnel G. Medical Issues H. Referral Resources I .Community and Cultural Competency TECHNICAL GUIDELINES A. Videoconferencing Applications B. Device Characteristics C. Connectivity D. Privacy ADMINISTRATIVE GUIDELINES A. Qualification and Training of Professionals B. Documentation and Record Keeping C. Payment and Billing REFERENCES.

Soldiers are expected to maintain the highest possible level of physical readiness because they must be ready to mobilize and perform their duties anywhere in the world at any time. The objective of Army body composition standards is to motivate physical training and good nutrition habits to ensure a high state of readiness. Establishment of enforceable and rational standards to support this objective has been challenging even at extremes of body size. Morbidly obese individuals are clearly not suited to military service, but very large muscular individuals may be superbly qualified for soldier performance demands. For this reason, large individuals are measured for body fat using a waist circumference-based equation (female soldiers are also measured for hip circumference). The main challenge comes in setting appropriate fat standards to support the full range of Army requirements. Military appearance ideals dictate the most stringent body fat standards, whereas health risk thresholds anchor the most liberal standards, and physical performance associations fall on a spectrum between these 2 poles. Standards should not exclude or penalize specialized performance capabilities such as endurance running or power lifting across a spectrum of body sizes and fat. The full integration of women into the military further complicates the issue because of sexually dimorphic characteristics that make gender-appropriate standards essential and where inappropriately stringent standards can compromise both health and performance of this segment of the force. Other associations with body composition such as stress effects on intraabdominal fat distribution patterns and metabolic implications of a fat reserve for survival in extreme environments are also relevant considerations. This is a review of the science that underpins the U.S. Army body composition standards.

The combination of predictive data-driven models with frequent glucose measurements may provide for an early warning of impending glucose excursions and proactive regulatory interventions for diabetes patients. However, from a modeling perspective, before the benefits of such a strategy can be attained, we must first be able to quantitatively characterize the behavior of the model coefficients as well as the model predictions as a function of prediction horizon. We need to determine if the model coefficients reflect viable physiologic dependencies of the individual glycemic measurements and whether the model is stable with respect to small changes in noise levels, leading to accurate near-future predictions with negligible time lag. We assessed the behavior of linear autoregressive data-driven models developed under three possible modeling scenarios, using continuous glucose measurements of nine subjects collected on a minute-by-minute basis for approximately 5 days. Simulation results indicated that stable and accurate models for near-future glycemic predictions (< 60 min) with clinically acceptable time lags are attained only when the raw glucose measurements are smoothed and the model coefficients are regularized. This study provides a starting point for further needed investigations before real-time deployment can be considered.

This paper presents a microfluidic electrical impedance flow cytometer (FC) for identifying the differentiation state of single stem cells. This device is comprised of a novel dual micropore design, which not only enhances the processing throughput, but also allows the associated electrodes to be used as a reference for one another. A signal processing algorithm, based on the support vector machine (SVM) theory, and a data classification method were developed to automate the identification of sample types and cell differentiation state based on measured impedance values. The device itself was fabricated using a combination of standard and soft lithography techniques to generate a PDMS-gold electrode construct. Experimental testing with non-biological particles and mouse embryonic carcinoma cells (P19, undifferentiated and differentiated) was carried out using a range of excitation frequencies. The effects of the frequency and the interrogation parameters on sample identification performance were investigated. It was found that the real and imaginary part of the detected impedance signal were adequate for distinguishing the undifferentiated P19 cells from non-biological polystyrene beads at all tested frequencies. A higher frequency and an opacity index were required to resolve the undifferentiated and differentiated P19 cells by capturing capacitive changes in electrophysiological properties arising from differentiation. The experimental results demonstrated salient accuracy of the device and algorithm, and established its feasibility for non-invasive, label-free identification of the differentiation state of the stem cells.

This paper tests the hypothesis that a "universal," data-driven model can be developed based on glucose data from one diabetic subject, and subsequently applied to predict subcutaneous glucose concentrations of other subjects, even of those with different types of diabetes. We employed three separate studies, each utilizing a different continuous glucose monitoring (CGM) device, to verify the model's universality. Two out of the three studies involved subjects with type 1 diabetes and the other one with type 2 diabetes. We first filtered the subcutaneous glucose concentration data by imposing constraints on their rate of change. Then, using the filtered data, we developed data-driven autoregressive models of order 30, and used them to make short-term, 30-min-ahead glucose-concentration predictions. We used same-subject model predictions as a reference for comparisons against cross-subject and cross-study model predictions, which were evaluated using the root-mean-squared error (RMSE) and Clarke error grid analysis (EGA). We found that, for each studied subject, the average cross-subject and cross-study RMSEs of the predictions were small and indistinguishable from those obtained with the same-subject models. These observations were corroborated by EGA, where better than 99.0% of the paired sensor-predicted glucose concentrations lay in the clinically acceptable zone A. In addition, the predictive capability of the models was found not to be affected by diabetes type, subject age, CGM device, and interindividual differences. We conclude that it is feasible to develop universal glucose models that allow for clinical use of predictive algorithms and CGM devices for proactive therapy of diabetic patients.

We compared the clinical outcomes after store-and-forward teledermatology with those following conventional clinic-based consultation. Subjects were randomized to either usual care (a conventional clinic-based dermatology appointment) or a store-and-forward teledermatology consultation. All subjects received baseline digital imaging and re-imaging was performed four months later. A total of 776 subjects were approached for inclusion, and a total of 508 image sets were reviewed, 236 in usual care and 272 in teledermatology. The image sets from both study arms were used to make clinical outcome assessments between baseline and four months. A dermatologist who was blinded to the randomization rated the clinical outcomes using a three-point clinical course rating scale (1 = improved, 2 = no change, 3 = worse). In the usual care group, 65% were rated as 'improved', 32% were rated as 'no change' and 3% were rated as 'worse'. For teledermatology, 64% were rated as 'improved', 33% as 'no change' and 4% as 'worse'. The results of the study indicate that store-and-forward teledermatology consultations produce similar clinical outcomes when compared with conventional clinic-based consultations.

BACKGROUND: Recent developments and expected near-future improvements in continuous glucose monitoring (CGM) devices provide opportunities to couple them with mathematical forecasting models to produce predictive monitoring systems for early, proactive glycemia management of diabetes mellitus patients before glucose levels drift to undesirable levels. This article assesses the feasibility of data-driven models to serve as the forecasting engine of predictive monitoring systems. METHODS: We investigated the capabilities of data-driven autoregressive (AR) models to (1) capture the correlations in glucose time-series data, (2) make accurate predictions as a function of prediction horizon, and (3) be made portable from individual to individual without any need for model tuning. The investigation is performed by employing CGM data from nine type 1 diabetic subjects collected over a continuous 5-day period. RESULTS: With CGM data serving as the gold standard, AR model-based predictions of glucose levels assessed over nine subjects with Clarke error grid analysis indicated that, for a 30-minute prediction horizon, individually tuned models yield 97.6 to 100.0% of data in the clinically acceptable zones A and B, whereas cross-subject, portable models yield 95.8 to 99.7% of data in zones A and B. CONCLUSIONS: This study shows that, for a 30-minute prediction horizon, data-driven AR models provide sufficiently-accurate and clinically-acceptable estimates of glucose levels for timely, proactive therapy and should be considered as the modeling engine for predictive monitoring of patients with type 1 diabetes mellitus. It also suggests that AR models can be made portable from individual to individual with minor performance penalties, while greatly reducing the burden associated with model tuning and data collection for model development.

OBJECTIVE: In 1995, the Children's Medical Services (CMS) of the State of Georgia contracted with the Department of Pediatrics of the Medical College of Georgia (MCG) and the MCG Telemedicine Center to develop telemedicine programs to provide subspecialty care for children with special health care needs. This article presents project statistics and results of client evaluation of services, as well as physician faculty attitudes toward telemedicine. DESIGN: A demonstration project using telemedicine between a tertiary center and a rural clinic serving children with special health care needs was established. Data were collected and analyzed for December 12, 1995 to May 31, 1997, during which 333 CMS telemedicine consultations were performed. RESULTS: Most CMS telemedicine consultations (35%) involved pediatric allergy/immunology. Other subspecialties included pulmonology (29%), neurology (19%), and genetics (16%). Overall, patients were satisfied with the services received. Initially, physician faculty members were generally positive but conservative in their attitudes toward using telemedicine for delivering clinical consultation. After a year's exposure and/or experience with telemedicine, 28% were more positive, 66% were the same, and only 4% were more negative about telemedicine. The more physicians used telemedicine, the more positive they were about it (r =.30). CONCLUSIONS: In terms of family attitudes and individual care, telemedicine is an acceptable means of delivering specific pediatric subspecialty consultation services to children with special health care needs, living in rural areas distant to tertiary centers. Telemedicine is more likely to be successful as part of an integrated health services delivery than when it is the sole mode used for delivery of care.

In 2020, the U.S. Department of Defense officially disclosed a set of ethical principles to guide the use of Artificial Intelligence (AI) technologies on future battlefields. Despite stark differences, there are core similarities between the military and medical service. Warriors on battlefields often face life-altering circumstances that require quick decision-making. Medical providers experience similar challenges in a rapidly changing healthcare environment, such as in the emergency department or during surgery treating a life-threatening condition. Generative AI, an emerging technology designed to efficiently generate valuable information, holds great promise. As computing power becomes more accessible and the abundance of health data, such as electronic health records, electrocardiograms, and medical images, increases, it is inevitable that healthcare will be revolutionized by this technology. Recently, generative AI has garnered a lot of attention in the medical research community, leading to debates about its application in the healthcare sector, mainly due to concerns about transparency and related issues. Meanwhile, questions around the potential exacerbation of health disparities due to modeling biases have raised notable ethical concerns regarding the use of this technology in healthcare. However, the ethical principles for generative AI in healthcare have been understudied. As a result, there are no clear solutions to address ethical concerns, and decision-makers often neglect to consider the significance of ethical principles before implementing generative AI in clinical practice. In an attempt to address these issues, we explore ethical principles from the military perspective and propose the "GREAT PLEA" ethical principles, namely Governability, Reliability, Equity, Accountability, Traceability, Privacy, Lawfulness, Empathy, and Eutonomy, for generative AI in healthcare. Furthermore, we introduce a framework for adopting and expanding these ethical principles in a practical way that has been useful in the military and can be applied to healthcare for generative AI, based on contrasting their ethical concerns and risks. Ultimately, we aim to proactively address the ethical dilemmas and challenges posed by the integration of generative AI into healthcare practice.

Analysis and review of combat casualty care challenges in future large-scale and medical multi-domain operations from the perspective of past, present, and potential future conflicts.

Accurate estimation of expression levels from RNA-Seq data entails precise mapping of the sequence reads to a reference genome. Because the standard reference genome contains only one allele at any given locus, reads overlapping polymorphic loci that carry a non-reference allele are at least one mismatch away from the reference and, hence, are less likely to be mapped. This bias in read mapping leads to inaccurate estimates of allele-specific expression (ASE). To address this read-mapping bias, we propose the construction of an enhanced reference genome that includes the alternative alleles at known polymorphic loci. We show that mapping to this enhanced reference reduced the read-mapping biases, leading to more reliable estimates of ASE. Experiments on simulated data show that the proposed strategy reduced the number of loci with mapping bias by ≥ 63% when compared with a previous approach that relies on masking the polymorphic loci and by ≥ 18% when compared with the standard approach that uses an unaltered reference. When we applied our strategy to actual RNA-Seq data, we found that it mapped up to 15% more reads than the previous approaches and identified many seemingly incorrect inferences made by them.