Low-Latency Wireless Network Extension for Industrial Internet of Things

The timely delivery of critical messages in real-time environments is an increasing requirement for industrial Internet of Things (IIoT) networks. Similar to wired time-sensitive networking (TSN) techniques, which bifurcate traffic flows based on priority, the proposed wireless method aims to ensure that critical traffic arrives rapidly across multiple hops to enable numerous IIoT use cases. IIoT architectures are migrating toward wirelessly connected edges, creating a desire to extend TSN-like functionality to a wireless format. Existing protocols possess inherent challenges to achieving this prioritized low-latency communication, ranging from rigidly scheduled time division transmissions, scalability/jitter of carrier-sense multiple access (CSMA) protocols, and encryption-induced latency. This paper presents a hardware-validated low-latency technique built upon receiver-assigned code division multiple access (RA-CDMA) techniques to implement a secure wireless TSN-like extension suitable for the IIoT. Results from our hardware prototype, constructed on the IntelFPGA Arria 10 platform, show that (sub-)millisecond single-hop latencies can be achieved for each of the available message types, ranging from 12 bits up to 224 bits of payload. By achieving one-way transmission of under 1 ms, a reliable wireless TSN extension with comparable timelines to 802.1Q and/or 5G is achievable and proven in concept through our hardware prototype.

Transfer Learning for Radio Frequency Machine Learning: A Taxonomy and Survey

Transfer learning is a pervasive technology in computer vision and natural language processing fields, yielding exponential performance improvements by leveraging prior knowledge gained from data with different distributions. However, while recent works seek to mature machine learning and deep learning techniques in applications related to wireless communications, a field loosely termed radio frequency machine learning, few have demonstrated the use of transfer learning techniques for yielding performance gains, improved generalization, or to address concerns of training data costs. With modifications to existing transfer learning taxonomies constructed to support transfer learning in other modalities, this paper presents a tailored taxonomy for radio frequency applications, yielding a consistent framework that can be used to compare and contrast existing and future works. This work offers such a taxonomy, discusses the small body of existing works in transfer learning for radio frequency machine learning, and outlines directions where future research is needed to mature the field.

Traditional injury scoring underestimates the relative consequences of orthopedic injury

OBJECTIVE

To demonstrate that patients with multiple injuries who have orthopedic injuries (ORTHO) face greater challenges regarding functional outcome than those without, to identify domains of postinjury dysfunction, and to illustrate the increasing discordance of functional recovery over time for ORTHO patients in relation to nonORTHO patients.

METHODS

A convenience sample of adult blunt force trauma patients admitted to a Level I trauma center was evaluated at admission, and at 6 and 12 months after injury. Data were collected from the trauma registry (Trauma One), chart review, and interviews. Mailed surveys were completed 6 and 12 months after injury. The Short Form 36 (SF36) general health survey and the Sickness Impact Profile work scale (SIPw) were administered at both time points. Data are presented as mean +/- SEM or percent (%). To compare means, t tests were conducted, and Injury Severity Score (ISS) was controlled by linear regression before the evaluation of the role of ORTHO injury pattern on outcome measures. Significance is noted at the 95% confidence level (p < 0.05).

RESULTS

The 165 patients studied averaged 37.2 +/- 1.1 years in age and were 67% men. The mean ISS was 14.4 +/- 0.6 and 61% had ORTHO injury. ORTHO patients were no different from nonORTHO in any measure of baseline status including the SIPw score and all domains of the SF36, except that the ISS was greater in the ORTHO group (15.6 +/- 0.96 vs. 12.7 +/- 0.73, p = 0.017). Baseline SF36 values were similar to national norms. Follow-up was 75% at 6 months, and 51% at 12 months. Those lost to follow-up differed only in that they were more likely to be men. Sixty-four percent had returned to work 12 months after injury. After controlling for ISS with linear regression, the ORTHO patients had worse scores on all physical measures of the SF36 (bodily pain, physical function, and role-physical). By 12 months after injury, the relative dysfunction of the ORTHO patients had expanded to include the SIPw score (p = 0.016) and six of eight SF36 domains (bodily pain, physical function, role-physical, mental health, role-emotional, and social function, all p < 0.05).

CONCLUSION

Injury severity affects both mortality and the potentially more consequential issues of long-term morbidity. Patients with ORTHO injury have relatively worse functional recovery, and this worsens with time. As trauma centers approach the limits of achievable survival, new advances in trauma care can be directed more toward the quality of recovery for our patients. This will be contingent on further development of screening, scoring, and treatment systems designed to address issues of functional outcome across injury boundaries for those who survive.

Prophylactic vena caval filters in trauma: the rest of the story

OBJECTIVE

The purpose of this study was to describe outcomes for patients with trauma who had vena caval filters placed in the absence of venous thromboembolic disease (group P) and compare them with outcomes for patients with trauma who had filters placed after either deep venous thrombosis or pulmonary embolism (group T).

DESIGN

The study is a case series of consecutive patients who received vena caval filters after traumatic injury. Data were collected prospectively at the time of filter placement from reports of diagnostic studies obtained for clinical indications and during the annual follow-up examinations. Event rate findings are based on objective tests. Data were obtained from the Michigan Vena Cava Filter Registry.

RESULTS

Filters were placed in 385 patients with trauma; 249 of these filters were prophylactic (group P). Event rates were similar in the two groups. New pulmonary embolism was diagnosed in 1.5% of the patients in group P and 2% of the patients in group T. Caval occlusion rates were 3.5% for group P and 2.3% for group T. In all, 15.6% of the patients in group P had deep venous thrombosis or pulmonary embolism after placement. The frequencies of lower extremity swelling and use of support hose were higher in group T than in group P (43% vs 25% and 25% vs 3.5%, respectively; P <.005). Outcomes were comparable in the two groups with respect to mechanical stability of the filter.

CONCLUSIONS

The prophylactic indication for vena caval filter placement in patients with trauma is associated with a low incidence of adverse outcomes while providing protection from fatal pulmonary embolism. The current challenge is to limit the number of unnecessary placements through improved methods of risk stratification.

Outcome from injury: general health, work status, and satisfaction 12 months after trauma

OBJECTIVE

We evaluated outcomes 12 months after trauma in terms of general health, satisfaction, and work status.

METHODS

Two hundred forty-seven patients without severe neurotrauma were evaluated by interview during admission and by mailed self-report 6 and 12 months after trauma. Data were obtained from the Trauma Registry, interviews, and survey instruments. Baseline assessment was obtained with the Short Form 36 (SF36) and the Sickness Impact Profile (SIP) work scale. Outcome measures were the SF36, SIP work scale, Brief Symptom Inventory (BSI) depression scale, the Civilian Mississippi Scale for Posttraumatic Stress Disorder (PTSD), and a satisfaction questionnaire. Three regressions were determined for outcome. The dependent variables were general health and work status (linear) and satisfaction (logistic). Each regression controlled for baseline status and mental health, Injury Severity Score (ISS), and 12-month SF36 physical function before evaluating the effect of outcome mental health.

RESULTS

Follow-up data were available for 75% of the patients at 6 months and 51% at 12 months. The mean age of patients was 37.2 +/- 0.9 years (+/-SEM), and 73% were male. Their average ISS was 13.9 +/- 0.6. Seventy percent of injuries were blunt force, 13.5 % were penetrating, and 16.5 % were burn injuries (mean total body surface area, 13.3 +/- 1.5%). Sixty-four percent of the patients had returned to work at 12 months. Follow-up SF36 mental health was associated with the dependent outcome in each regression. After controlling for baseline status and mental health, ISS, and outcome SF36 physical function, outcome mental health was associated with outcome SF36 general health (p < 0.001), SIP work status (p = 0.017), and satisfaction with recovery (p = 0.005). Outcome SF36 mental health was related to baseline mental health, 12-month PTSD and BSI depression scores, and increased drug and alcohol use.

CONCLUSIONS

Twelve months after trauma, patients' work status, general health, and overall satisfaction with recovery are dependent on outcome mental health. This dependency persists despite measured baseline status, ISS, or physical recovery. The mental disease after trauma is attributable to poor mental health, the development of symptoms of PTSD and depression, and increased substance abuse. Trauma centers that fail to recognize, assess, and treat these injury-related mental health outcomes are not fully assisting their patients to return to optimal function.

Posttraumatic stress disorder in injured adults: etiology by path analysis

OBJECTIVE

Posttraumatic Stress Disorder (PTSD) impairs outcome from injury. We present a path analysis of factors related to the development of PTSD in injured adults.

METHODS

A prospective cohort of 250 patients without severe neurotrauma was evaluated by interview during admission and by mailed self-report 6 months later. Data were gathered from the trauma registry (age, injury mechanism, and Injury Severity Score), social history (gender, income, education, and social support), and survey instruments. Baseline assessment used the Michigan Critical Events Perception Scale (peritraumatic dissociation and subjective threat to life), the Life Experience Survey (stressful exposure history), and the SF36 (general and mental health). PTSD at 6 months was identified with the civilian Mississippi Scale for PTSD. Data are listed as mean +/- SEM or percent (%). Path analysis was conducted by linear regression and significant (p<0.05) variables are shown. Factors are listed with the standardized beta. A negative beta suggests a protective effect.

RESULTS

The 176 patients (72%) who completed the 6-month follow-up were 37.7+/-0.88 years old; 75% were men; and blunt (70%), penetrating (13.5%), and burn (16.4%) mechanisms caused the injuries. Assault was involved in 14.5% of the cases. Average income was $44,300+/-2,700/yr, education was 13.0+/-0.15 years, and Injury Severity Score was 13.9+/-0.50. A total of 42.3% of the patients developed PTSD. The 39.7% of the variance in PTSD explained by the model was due to intentional injury (beta = 0.27), male gender (beta = -0.21), age (beta = -0.20), peritraumatic dissociation (beta = 0.174), baseline mental health (beta = -0.21), and prior life-threatening illness (beta = -0.29). Peritraumatic dissociation was due to the patient's sense of threat to life (beta = -0.47), and threat was related to Injury Severity Score (beta = 0.2), assault(beta = 0.14), education (beta = -0.15), and age (beta = -0.19). Baseline SF36 mental health was related to social support (beta = 0.27) and income (beta = 0.21). Income was contingent on education (beta = 0.21).

CONCLUSION

PTSD occurred in 42.3% of injured adults 6 months after trauma and was related to assault, dissociation, female gender, youth, poor mental health, and prior illness. By modeling PTSD, we may learn more of the etiology, risk stratification, and potentials for the treatment of this common and important morbidity of injury.

Posttraumatic stress disorder after injury: impact on general health outcome and early risk assessment

OBJECTIVE

To evaluate prospectively components of general health outcome after trauma and to report on the further validation of the Michigan Critical Events Perception Scale (MCEPS), an instrument that predicts increased risk for posttraumatic stress disorder (PTSD).

METHODS

Adults without neurologic injury admitted to a Level I trauma center in 1997 were interviewed during hospitalization. Baseline data included demographics, injury mechanism, Injury Severity Score, the Short Form 36 (SF36), and the MCEPS, which measures peri-traumatic dissociation (the sense of depersonalization or derealization during an injury event). Surveys sent by mail and completed 6 months later included the SF36 and civilian Mississippi Scale for PTSD.

RESULTS

A total of 140 patients were interviewed; the 70% (n = 100 patients) who completed the 6-month assessment form the study group. Injuries were categorized as 71% blunt, 13% penetrating, and 16% burn. Mean Injury Severity Score was 13.7+/-0.52. PTSD at 6 months occurred in 42% of the patients and was directly related to MCEPS dissociation (p = 0.001; odds ratio = 3.1; 95% confidence interval, 1.6, 5.9). A stepwise linear regression explains 40% of the variance in 6-month SF36 general health outcome (adjusted R2 = 0.402). The model controls for individual factors related to dissociation, PTSD, and general health outcome. Development of PTSD was independently and inversely related to general health outcome as measured by the SF36 at 6 months (p < 0.001, beta = -0.404). The R2 change of 0.132 for PTSD (vs. 0.082 for 6-month physical function) illustrates that PTSD contributes more to the patient's perceived general health at 6 months than the degree of physical function or injury severity.

CONCLUSIONS

Within hours of injury, the MCEPS identifies patients who are three times more likely to develop PTSD. PTSD compromises self-reported general health outcome in injured adults independent of baseline status, Injury Severity Score, or degree of physical recovery. These data suggest that psychological morbidity is an important part of the patient's perceived general health.

Blunt thoracic aortic injury: delayed or early repair?

OBJECTIVE

To evaluate the safety and benefit of delayed repair of blunt thoracic aortic injury (BTAI) in trauma patients with multiple injuries and to assess the financial impact of delayed repair.

METHODS

A retrospective review of charts was performed on 55 patients with the diagnosis of BTAI from January 1, 1992, through December 31, 1997, at our Level I trauma center. Early repair was defined as operative repair of BTAI within 12 hours of admission. Seven patients were excluded from analysis due to death before BTAI diagnosis (two deaths were from rupture in the emergency department and five were from massive blunt trauma without rupture). The groups were compared by using a McNemar chi2 test, for which p less than or equal to 0.05 is significant.

RESULTS

There were 30 patients in the early repair (ER) group repaired at 5.3+/-2.4 hours, and 18 patients in the delayed repair (DR) group repaired at 8.5 days (range, 17 hours-67 days). There were no significant differences between the ER and DR groups in age (37+/-18 years vs. 41+/-19 years), Injury Severity Score (39+/-15 vs. 45+/-14), intensive care unit days (12+/-14 days vs. 18+/-11 days), hospital length of stay (21+/-19 days vs. 28+/-14 days), or mortality rates (7% vs. 6%). There was a trend toward longer lengths of stay in the DR group. Most DR patients required beta-blocker therapy and/or other antihypertensives for systolic BP more than 120 mm Hg during admission. There were no deaths from aortic rupture in either group. By using financial data that was available from July of 1994 onward, we performed a subset analysis of the direct costs associated with BTAI. Total direct and variable direct costs for patients undergoing delayed repair were over two times the costs for early repair patients (p < 0.05).

CONCLUSION

The management of trauma patients with multiple injuries requires prioritization of injuries so that the outcomes from these injuries can be optimized. Although delayed aortic repair was safely practiced in this series, there was not an obvious outcome benefit to delayed repair. The patients undergoing late repair required increased attention to hemodynamics, and there was a trend toward increased length of stay. In addition, analysis of the costs associated with delayed repair demonstrated a twofold increase in the direct costs for delayed repair compared with early repair.

Extracorporeal life support in pulmonary failure after trauma

OBJECTIVE

To present a series of 30 adult trauma patients who received extracorporeal life support (ECLS) for pulmonary failure and to retrospectively review variables related to their outcome.

METHODS

In a Level I trauma center between 1989 and 1997, ECLS with continuous heparin anticoagulation was instituted in 30 injured patients older than 15 years. Indication was for an estimated mortality risk greater than 80%, defined by a PaO2: FIO2 ratio less than 100 on 100% FIO2, despite pressure-mode inverse ratio ventilation, optimal positive end-expiratory pressure, reasonable diuresis, transfusion, and prone positioning. Retrospective analysis included demographic information (age, gender, Injury Severity Score, injury mechanism), pulmonary physiologic and gas-exchange values (pre-ECLS ventilator days [VENT days], PaO2:FIO2 ratio, mixed venous oxygen saturation [SvO2], and blood gas), pre-ECLS cardiopulmonary resuscitation, complications of ECLS (bleeding, circuit problems, leukopenia, infection, pneumothorax, acute renal failure, and pressors on ECLS), and survival.

RESULTS

The subjects were 26.3+/-2.1 years old (range, 15-59 years), 50% male, and had blunt injury in 83.3%. Pulmonary recovery sufficient to wean the patient from ECLS occurred in 17 patients (56.7%), and 50% survived to discharge. Fewer VENT days and more normal SvO2 were associated with survival. The presence of acute renal failure and the need for venoarterial support (venoarterial bypass) were more common in the patients who died. Bleeding complications (requiring intervention or additional transfusion) occurred in 58.6% of patients and were not associated with mortality. Early use of ECLS (VENT days < or = 5) was associated with an odds ratio of 7.2 for survival. Fewer VENT days was independently associated with survival in a logistic regression model (p = 0.029). Age, Injury Severity Score, and PaO2:FIO2 ratio were not related to outcome.

CONCLUSION

ECLS has been safely used in adult trauma patients with multiple injuries and severe pulmonary failure. In our series, early implementation of ECLS was associated with improved survival. Although this may represent selection bias for less intractable forms of acute respiratory distress syndrome, it is our experience that early institution of ECLS may lead to improved oxygen delivery, diminished ventilator-induced lung injury, and improved survival.

Trauma service cost: the real story

OBJECTIVE

The objective was to define and characterize the costs associated with trauma care at a level I trauma center. Once the costs were identified, attending physician-led teams were designed to reduce costs within each cost center.

SUMMARY BACKGROUND DATA

The location and magnitude of the costs on a trauma service remain largely unknown. Focused cost-containment strategies remain difficult to implement because the expected return on these interventions is unknown.

METHODS

Cost center data were reviewed for the 40 major DRGs admitted for the first 6 months of the fiscal years 1996 and 1997. Data were obtained from the hospital finance department using the Transition Systems Inc. accounting system. We focused on variable direct costs, those that vary with patient volume (e.g., staff nursing expense and medical/surgical supplies). To address issues of inflation, pay raises, and changing costs, a proxy value was created for 1996 and costs were held constant for the 1997 calculation. The major services that constitute cost centers identified in the system were nursing, surgical, pharmacy, laboratory, radiology, and emergency services. Attendings were assigned to develop and oversee customized cost-reduction modalities specific to each cost center. The cost-reduction modalities used to achieve significant savings were as follows: nursing, case management approach focusing on early discharge; surgical, meeting with operating room (OR) purchasing to modify expensive behavior patterns; pharmacy, integrating clinical pharmacist with direct attending support; laboratory, enforcing protocol for lab draws; radiology, increasing the use of emergency room ultrasound and accepting outside x-rays; and emergency services, 24-hour in-house attending staff to reduce emergency room time. The surgical and emergency services cost centers predominately generate costs by the length of time care is delivered in that area.

RESULTS

For each period, data from 363 patients were compared. Mean length of stay decreased between the study periods from 8.72 to 7.06 days, while the average injury severity score was unchanged. Together, these cost centers constituted 87.4% of the total cost of care delivered. Significant cost reduction was achieved in all six variable cost centers: nursing (24%), surgical (5%), pharmacy (57%), laboratory (27), radiology (7%), and emergency (36). The mean cost per case was reduced by 25%.

CONCLUSIONS

Identification of the true cost centers and directed attending surgeon involvement are essential to the development and implementation of a successful cost-reduction process.

Psychosocial factors limit outcomes after trauma

BACKGROUND

Psychological morbidity compromises return to work after trauma. We demonstrate this relationship and present methods to identify risks for significant psychological morbidity.

METHODS

Thirty-five adults were evaluated prospectively for return to functional employment after injury using demographic data, validated psychological and health measures, and the Michigan Critical Events Perception Scale. Evaluation was conducted at admission and at 1 and 5 months after injury.

RESULTS

Poor return to work at 5 months was attributable to physical disability (p < 0.05) and psychological disturbance (p < 0.05) in a regression model that controlled for preinjury employment and psychopathologic factors as well as injury severity. A high score on the Impact of Events Scale administered during acute admission predicted development of acute stress disorder at 1 month (p < 0.01, odds ratio (OR) = 9.4) and posttraumatic stress disorder at 5 months (p < 0.05, OR = 6.7). Peritraumatic dissociation on the Michigan Critical Events Perception Scale was predictive for development of acute stress disorder (p < 0.05, OR = 5.8) at 1 month and posttraumatic stress disorder (p < 0.05, OR = 7.5) at 5 months.

CONCLUSION

Psychological morbidity after injury compromises return to work independent of preinjury employment and psychopathologic condition, Injury Severity Score, or ambulation. A high Impact of Events Scale score or peritraumatic dissociation at admission predicts this morbidity.

Physician resource utilization after geriatric trauma

OBJECTIVE

As health care resources become increasingly strained, the value of physician consultation has come under heightened scrutiny. This report reviews the value of early consultation by the physical medicine and rehabilitation (PMR) service to an integrated trauma service for geriatric patients with multiple trauma.

METHODS

We retrospectively reviewed the records of 110 geriatric trauma patients (age > 60 years) with an Injury Severity Score > or = 15 to evaluate the effects of PMR consultation. Patients in group 1 were admitted to a general surgical service, and those in group 2 were admitted to a multidisciplinary trauma service. Demographic and physiologic factors, as well as short-term and long-term outcomes, were evaluated, and a subgroup analysis was performed to compare early (< or =3 days) versus late (>3 days) consultation by PMR.

RESULTS

Although there were significant differences in Glasgow Coma Scale score and length of stay, no differences were found within groups in other demographic, physiologic, or outcome data. Focused review of PMR intervention based on early versus late consultation revealed no significant difference between the two groups. Furthermore, an after-discharge phone survey revealed no significant group differences in dependence on a care provider or nursing home placement, readmission to hospital, employment status, or current functional activity status.

CONCLUSIONS

Long-term patient functional outcome and the in-house rehabilitation process are not affected by integration of PMR into a multidisciplinary trauma team or early PMR consultation.

Blunt force injury of the abdominal aorta

OBJECTIVE

To review the clinical presentation, diagnosis, and management of injury to the abdominal aorta after blunt force trauma.

DESIGN

This study was a retrospective review.

RESULTS

A total of 5,676 patients were admitted to the University of Michigan Medical Center with traumatic injury. Seven had injuries to the abdominal aorta after a blunt force mechanism. Five patients had operative repair of the aortic injury, of which four involved orthotopic graft placement and one had an extra-anatomic bypass. Two patients had the aortic injury repaired by endovascular stent placement in the angiography suite. One patient died, and lower extremity amputations were performed in three patients.

CONCLUSIONS

Surgical repair of abdominal aortic injury is preferable for the unstable patient or those with threatened extremities. In the stable patient with viable limbs, treatment with radiologic placement of endovascular stents may provide a nonoperative option for management.