Printer Friendly

Revised evaluation of hemodilution response in the semi-closed loop infusion system.

I. INTRODUCTION

A mini volume loading test (mVLT) is used for decision making in a semi-closed loop infusion system (SCLIS) [1]-[5]. The mVLT fluid protocol implies hemodilution guided mini fluid challenges. These are 2.5 ml [kg.sup.-1]-5 ml [kg.sup.-1] boluses infused over 2 min-5 min, each followed by a 5 min period with no fluids. The assessment of hemodilution response requires hemoglobin measures before and after each mini fluid challenge (MFC). Hemoglobin in arteries (invasive, aHb) and capillaries (non-invasive, SpHb) are used to calculate the arterial and capillary plasma dilution efficacies (aPDE and cPDE, respectively) and arteriocapillary plasmadilution efficacy difference (acPED) [1], [3]. Based on observations in healthy volunteers [1] and patients [2], a transcapillary reflux model suggests that acPED [less than or equal to] 0 p.d.u. is an indication hemodilution nonresponse and imminent edema. This signifies that fluid loading should be stopped. However, the need for invasive measurements is a major drawback in the clinical implication of our novel automated clinical decision support system (ACDSS) and the SCLIS [3]-[5]. A solely noninvasive assessment of hemodilution response would be possible if the cPDE could reliably predict the acPED value or the hemodilution response (the presence of acPED value within a range of acPED > 0) or non-response (acPED [less than or equal to] 0).

The primary objective was to assess the accuracy of cPDE in predicting the hemodilution responsiveness using the receiver operating characteristic (ROC) curve and gray zone approaches [6]. The secondary objective was to investigate the feasibility of a statistical optimization method (SOM) in predicting the acPED using the measured cPDE.

II. MATERIALS AND METHODS

We retreived and retrospectively analysed data from our previous randomized controlled trial (Clinicaltrials.Gov, identifier: NCT01355900) in patients who had elective primary total knee arthroplasty (TKA) surgery [2], [7]. The adhesive spectrophotometry sensor (ReSposable R2-25a; Masimo Inc., Irvine, CA) connected to the Radical-7 Pulse CO-Oximeter (Masimo Inc.; software version 7.6.2.1) was used to measure SpHb. A radial artery cannula was placed in the same arm on which the sensor was attached. This line was used to extract blood for the analysis of aHb by CO-Oximetry (COULTER[R] LH750, Beckman Coulter Inc., Chicago, IL) with a coefficient of variation [less than or equal to] 0.8 %. An antecubital cannula for infusion was placed in the same arm.

The stepwise infusion was administered according to a revised goal directed therapy (revGDT) protocol consisting of conventional GDT supplemented by the mVLT [2]. The cPDE and acPED were obtained in 48 subjects who had two mVLT sessions--before the induction of anesthesia (preoperative session) and again 24 hours later (post-operative). In each mVLT session they received six MFCs with 2.5 ml [kg.sup.-1] boluses of Ringer's acetate solution. The aHb and SpHb were measured immediately before and after each MFC. These variables were used to calculate aPDE, cPDE (Fig. 1) and acPED (Fig. 2) using the Windows[R] platform-based software of our prototypes ACDSS and SCLIS [3]-[5].

III. STATISTICAL ANALYSIS

A. Sample Size

As previously suggested [8], assuming that 50 % of MFCs are likely to be hemodilution 'responder' cases, a total of 200 MFCs would be necessary for the reliable determination of the ROC curve and cut-off value. With an aim to perform bootstrapping, we included data from 480 MFCs, a number close to the previously used sample size [9]. We analysed the pooled data related to the last five MFCs in each of the two perioperative mVLT sessions in 48 subjects. Thus, there were 480 paired values of cPDE and acPED from 96 mVLT sessions. The MFCs were divided in 2 groups according to the acPED: the hemodilution responders were defined as acPED > 0 and non-responders as acPED < 0.

B. Receiver Operating Characteristic (ROC) Curve

A receiver operating characteristic (ROC) curve (Fig. 3) was created to assess the discriminative power of cPDE to predict the presence of the acPED value within a range of hemodilution non-response (acPED [less than or equal to] 0). The area under the ROC curve was also created by using bootstraping [9], [10]. It creates multiple samples (1,000) by randomly drawing instances from the study population. Since the unique best cut-off cPDE value cannot be obtained by the bootstrap method, the best cut-off value with the ROC curve was obtained without bootstrap.

C. Threshold and Gray Zone Determination

The gray zone approach [6], [9], [11]-[13] was used to determine a range of cPDE for which no conclusion may be drawn concerning hemodilution response. We used a two-step procedure. In the 1st step, the best threshold for cPDE was determined in a bootstrapped population. The 95 % CI of the best threshold was defined using the distribution of the thresholds [2]. The best threshold for the ROC curve was defined as that which maximizes the Youden index (sensitivity + specificity -1).

The aim of the 2nd step was to determine the cPDE values for which no conclusive information could be provided. The two-curve representation illustrates this step (Fig. 4(a)). We defined inconclusive values with the sensitivity of < 90 % and specificity < 90 %. The gray zone was defined as the values that comply with a 10 % diagnostic tolerance. When the characteristics of the study population produced a 95 % CI of the best threshold and it is larger than the inconclusive zone, the values obtained during the first step were retained as the gray zone. The two-curve representation illustrates this step (Fig. 4(a)). As previously applied to prediction of fluid responsiveness for the flow-related dynamic target parameters in GDT [9], the optimal threshold for cPDE to predict the hemodilution non-response was then determined as the one that minimized the explicit cost ratio (R = 1), which is equivalent to maximizing the Youden's index. By determining this value for all the 1,000 resampled variables, a distribution for the optimal thresholds of the 1,000 samples was obtained. The optimal threshold was 0.015, and the 95 % CI of this distribution (the gray zone) was 0.013-0.027. Considering that the determination of the gray zone is prone to slightly different results between assessments [9], the results of the two approaches were merged as a gray zone from 0.013 to 0.028 (p.d.u.).

We used the SOM as an alternative method for predicting hemodilution response by calculating the predicted acPED.

IV. Statistical Optimization Method (SOM) The SOM was based on the inverse relationship between cPDE and acPED in individual patient's records (Fig. 5).

The SOM was applied to calculate the predicted acPED value ([acPEDs.sub.om]) from the non-invasively measured (actual) cPDE by using two approaches:

1. Invertion of the cPDE

[acPED.sub.som] = (-cPDE). (1)

2. Applying an offset (coefficient)

[acPED.sub.of] = [K.sub.of] - cPDE = [K.sub.of] + [acPED.sub.som], (2)

where [K.sub.of]--empiric coefficient determined by the brute-force search method.

The proportion of hemodilution response patterns was compared between the actual (based on the measured acPED) and the predicted (based on the calculated acPEDsom and [acPED.sub.of]) hemodilution responses. Chi-square test was used to compare proportions of actual and predicted hemodilution responses in all approaches. The statistical analysis was performed with SAS 9.2 (SAS Institute Inc., Cary, NC). Results are expressed according to variable distributions (mean and 95 % CI) or median [25th; 75th percentiles] for quantitative variables.

V. RESULTS

Data from 48 TKA patients (41 women) were analysed. Subjects were 69.5 (6.0) years old, their height (m) was 1.62 [1.56 to 1.65], weight (kg) was 87.0 [75.5 to 97.5] and the body mass index (BMI, kg/[m.sup.2]) was 33.5 [31.0 to 35.1]. We analyzed cPDE and acPED records, 480 values of each. The variables were recorded during five mini fluid challenges in each of the 48 pre- and 48 post-operative mVLT sessions.

A. The ROC Curve and Gray Zone Approaches

The area under the ROC curve with 95 % CI was 0.936 (CI: 0.915 to 0.958). The best cut-off value for cPDE was 0.015 (specificity = 0.84, sensitivity = 0.90, positive likelihood ratio = 8.34, negative likelihood ratio = 0.12). The 95 % CI for the best threshold value was 0.013 to 0.027 (p.d.u.). The results of the ROC curve and gray zone approaches were merged and suggested a range of inconclusive cPDE values from 0.013 to 0.028 (p.d.u.). The two-curve representation is used to illustrate this approach (Fig. 4(a)). There were 69 (14 %) cPDEs in this gray zone. There were 184 (38.3 %) cPDEs below this threshold. The previous version of mVLT algorithm [2] was then revised by implementing these findings (Fig. 5). The software of our prototype ACDSS within SCLIS [3], [4] was upgraded, accordingly (Fig. 7).

B. Actual Hemodilution Response

The actual acPED [less than or equal to] 0 (p.d.u.) was found in 296 (61.7 %) of the mini fluid challenges defined as non-responders. There were 82 (34.2 %) responders in pre-operative, and 102 (42.5 %) in post-operative mVLT sessions. There were 158 (65.8 %) non-responders in pre-operative, and 138 (57.5 %) in post-operative sessions.

There was no difference between the proportions of response patterns in pre- and post-operative mVLT sessions: there were 82 pre-operative MFCs with acPED [less than or equal to] 0 and 158 with acPED > 0 while there were 102 post-operative MFCs with acPED [less than or equal to] 0 and 138 with acPED > 0 (P = 0.074).

C. SOM Predicted Hemodilution Response

We found that 91 (18.9 %) [acPED.sub.som] was in the range of hemodilution 'response' ([acPED.sub.som] > 0 p.d.u.) and 389 (81.1 %) were in the range of 'non-response' ([acPED.sub.som] < 0).

Also, 181 (37.7 %) [acPED.sub.of] was a responder ([acPED.sub.of] > 0 p.d.u.) and 299 (62.3 %) was non-responder ([acPED.sub.of] < 0).

D. Actual vs. SOM Predicted Hemodilution Response

The number of actual and SOM predicted ([acPED.sub.som]) was [n.sub.1] = 480 and [n.sub.2] = 480, and the proportion of response patterns was [p.sub.1] = 184 and [p.sub.2] = 91, respectively. There was a statistically significant difference (P = 3.16*10-11).

The 0.015 offset ([K.sub.of] = 0.015) was used in the calculation of [acPED.sub.of] (2). The number of samples was [n.sub.1] = 480, [n.sub.2] = 480, and the proportions of hemodilution response patterns were [p.sub.1] = 184, [p.sub.2] = 181, respectively. There was no significant difference between the proportions (P = 0.849).

VI. DISCUSSION

This retrospective observational study used the ROC curve and gray zone approaches to examine the accuracy of cPDE in the evaluation of hemodilution response to mini fluid challenges with crystalloids in the mVLT protocol.

Our results show the three patterns of the cPDE predicted hemodilution responses: negative, inconclusive and positive. The gray zone approach has proposed two cut-offs: subjects with a cPDE > 0.028 p.d.u. exclude positive response with near certainty (negative response), whereas one with a cPDE < 0.013 p.d.u. will nearly always respond to a MFC (positive response). The cPDE between 0.013 and 0.028 (p.d.u.) cannot reliably predict the response (inconclusive response). There were 14 % MFCs with inconclusive cPDE predictions of response in two perioperative mVLT sessions.

The area under the ROC curve for cPDE (Fig. 3) was the first approach in our analysis. The ROC curve was created by using a bootstrap methodology which limits the impact of outliers [9], [10], but the best cut-off value (0.015 p.d.u.) was obtained without bootstrapping. The second approach was the gray zone method. This approach was proposed to avoid the binary constraint of a "black-or-white" decision of the ROC curve approach that often does not fit the reality of clinical practice [6].

In this study, we also used a previously proposed ratio of costs (R) defined as R = false positive cost/false negative cost [9]. The R < 1 denotes that not treating a false-negative result is worse than treating a false-positive result which may lead to unnecessary fluid loading [potential risk/fluid overload]. The R > 1 denotes that to treat a false-positive result is worse than to miss a false-negative [9]. The results of both approaches were merged as a gray zone for cPDE from 0.013 to 0.028 (p.d.u.). The two-step procedure in the gray zone approach was used with an aim to minimize the impact of potential outliers.

A difference between the aPDE and cPDE, the acPED, is dependent on the interaction between a transcapillary fluid filtration-absorption ratio (FAR) which affects hemodilution in capillaries, and net fluid elimination which affects the hemodilution in arteries. The acPED may decrease when the aPDE decreases, e.g., due to activation of renal elimination, and/or when the cPDE increases due to an increase in fluid shift from tissues to capillaries (increase in FAR). Therefore, since activation of renal elimination requires about 20 min, and distribution of crystalloid throughout the extracellular fluid space requires about 30 min for completion [14], we have excluded the first mini fluid challenge related data from the present analysis.

Despite the obvious possibility of different combinations of the FAR and net fluid elimination rate, the revised mVLT algorithm (Fig. 6) implies that fluid loading is not indicated when hemodilution in capillaries is higher than in arteries as suggested by cPDE > 0.028 p.d.u., according to the results of the present study. The background concept of the mVLT is depicted on physiological basis in the transcapillary reflux model [1], [2]. It explains the changes in acPED during the incremental hemodilution in the the context of changes in viscoelastic properties of the tissues that surround capillaries and the rate of systemic fluid elimination from circulation. The phenomenon of negative acPED is explained by the level of tissue hydration in which interstitium releases excessive fluid into capillaries during the 5 min period with no fluids after a bolus. Our model, which is described in detail within previous reports [1], [2], is similar to the recently proposed 'physiological model' outlined within the 'Discussion' of the retrospective analysis of data from previous volume kinetic studies in healthy volunteers and patients [14]. The feasibility of the models needs evidence from the direct measurments of changes in pertinent tissues as they expand during the goal directed fluid infusion. Further research in this area is urgently required.

We used the findings of our present study to upgrade the mVLT algorithm (Fig. 6) in our Windows[R] platform-based software of the prototype SCLIS [3]-[5] (Fig. 7). We defined the positive hemodilution response as cPDE < 0.013 p.d.u. since it predicts the acPED > 0 p.d.u.. However, according to the mVLT method, both cPDE < 0.013 p.d.u. and acPED > 0 p.d.u. may be observed in both under- and overhydration states. Thus, a static cPDE value is not a reliable factor when deciding on the need for additional MFCs. Meanwhile, a trend of cPDE can provide useful insights, e.g., boluses are indicated if cPDE within that range decreases, but should be stopped if cPDE increases. Moreover, in cases when the cPDE is within the gray zone during several consecutive MFCs, it remains unclear as to how many MFCs are feasible. Theoretically, boluses could be indicated if the cPDE within the gray zone decreases, but should be stopped if it increases. Future research into this issue is needed.

The off-label use of Pulse CO-Oximetry for non-invasive measurement of capillary hemodilution remains an isue that has to be addressed in discussions with industry in the light of current evidence that non-invasive SpHb cannot reliably predict hemoglobin concentration in large vessels [15].

VII. CLINICAL IMPLICATIONS

When a given value of cPDE is in the gray zone, physicians cannot use this index as a factor in deciding whether to proceed with or to exclude the need for additional mini fluid challenges. Conversely, when a measured value is outside of the gray zone, the necessity of performing a fluid challenge can be confirmed (value above the upper limit of the grey zone) or excluded (value below the lower limit of the gray zone) with less than 10 % chance of error (specificity and sensitivity > 90 %). At the bedside, the gray zone appears to be more informative than the confidence interval of the best threshold value obtained using the ROC curve method.

The future implications of the mVLT within the revGDT protocol and SCLIS may include the evaluation of the body hydration levels and detection of imminent edema during infusion, as well as monitoring the redistribution of fluids following administration of a diuretic. These expectations are supported by previous reports that mVLT method was able to discriminate between differrent levels of hydration status in healthy volunteers who received 3 mini fluid challenges with the 2.5 ml [kg.sup.-1] boluses of Ringer's acetate [1] and detect pre-operative dehydration in 36 TKA patients who received 3 mini fluid challenges with the 5 ml [kg.sup.-1] boluses of the same crystalloid [2]. In 48 TKA patients, who received six 2.5 ml [kg.sup.-1] boluses of Ringer's acetate in pre-and post-operative mVLTs, the positive 24-h fluid balance and higher arterio-capillary plasma dilution difference in pre-operative mVLT compared to post-operative suggested higher post-operative hydration level [16]. These results are consistent with the findings of our present study. We retrospectively analysed data that was obtained in 5 out of 6 mini fluid challenges in these 48 TKA patients and found that there was a trend towards lower number of negative hemodilution responses in pre-operative mVLT compared to post-operative: 82 pre- vs. 102 post-operative MFCs with the acPED [less than or equal to] 0 (P = 0.074).

VIII. KEY MESSAGES

--In 480 measurements, a gray zone approach showed that cPDE cannot reliably predict hemodilution response when its value is between 0.013 and 0.028 (p.d.u.).

--In the TKA surgery setting, about 14 % of mini fluid challenges with the 2.5 ml [kg.sup.-1] boluses of a crystalloid had the cPDE values within a gray zone.

--An isolated value of cPDE should be considered as a tool rather than a target for titrating the fluid infusion. Future studies are necessary to evaluate the feasibility of targeting cPDE as a resuscitation endpoint.

IX. CONCLUSIONS

The non-invasive evaluation of hemodilution response is acceptably reliable in perioperative TKA setting. The mVLT algorithm within the prototype SCLIS was upgraded by implementing the results of the present study.

http://dx.doi.org/10.5755/j01.eee.22.1.12826

Manuscript received 26 January, 2015; accepted 10 June, 2015.

This work received financial support from the European Social Fund under the Global Grant measure (VP1-3.1-SMM-07-K-03-019; the EU Structural Funds).

CONFLICT OF INTERESTS

A. A. has received a consultant's fee, travel funding and an honorarium for an expert report from Masimo Corp. (Irvine, CA, USA). A. A. is an inventor in US patent US 7,788,045 B2, US patent application US 61/470,224 and International application US 13/973,747.

ACKNOWLEDGEMENT

We thank Dr. Christer H. Svensen, MD, PhD (Professor, Senior Consultant, Director for Doctoral Education at Karolinska Institutet, Department of Clinical Science and Education, Section of Anaesthesiology and Intensive Care, Sodersjukhuset, Stockholm, Sweden) for his valuable ideas, consulting and advices during the study.

APPENDIX A

We have retrieved and pooled the data from the first part of the primary RCT [7] where 3 mini fluid challenges with the 5 ml [kg.sup.-1] boluses of Ringer's acetate were used in each of the 36 pre- and 36 post-operative mVLT sessions [2]. Since we analysed only the data from the last two MFCs (second and third), a total of 144 measures were analysed. The cut-off cPDE value was 0.026 p.d.u. (specificity = 0.93, sensitivity = 0.77). The gray zone was a range of cPDE values between 0.015 and 0.080 (p.d.u.) (Fig. (b)). The difference in findings compared to the ones based on the 480 measures in 48 patients may be related to the use of a twice larger volume (doubled) but smaller number of boluses. However, the significantly smaller sample size (144 vs. 480) is probably the main source of discrepancy. Future research is needed for more conclusive results.

Previously, data from 36 subjects who received three 5 ml [kg.sup.-1] boluses have been published addressing blood loss [7], accuracy of non-invasive SpHb measurements [3] and the use of the mVLT to detect pre-operative dehydration [2].

REFERENCES

[1] C. H. Svensen, E. Stankevicius, J. Broms, V. Markevicius, A. Andrijauskas, "Evaluation of hydration status calculated from differences in venous and capillary plasma dilution during stepwise crystalloid infusions: A randomized crossover study in healthy volunteers", Medicina, vol. 50, no. 5, pp. 255-62, 2014. [Online]. Available: http://dx.doi.org/10.1016/j.medici.2014.09.007

[2] A. Andrijauskas, J. Ivaskevicius, N. Porvaneckas, E. Stankevicius, C. H. Svensen, V. Uvarovas, et al., "A mini volume loading test for indication of preoperative dehydration in surgical patients", Medicina, vol. 51, no. 2, pp. 81-91, 2015. [Online]. Available: http://dx.doi.org/10.1016/j.medici.2015.02.001

[3] V. Markevicius, A. Andrijauskas, D. Navikas, C. Svensen, N. Porvaneckas, D. Andriukaitis, et al., "Statistically biased calibration method for the real-time adjustment of noninvasive haemoglobin measurements in a semi-automated infusion system", Elektronika ir Elektrotechnika, vol. 19, no. 7, pp. 65-71, 2013. [Online]. Available: http://dx.doi.org/10.5755/j01.eee.19.7.5165

[4] V. Markevicius, A. Andrijauskas, D. Navikas, N. Dubauskiene, N. Porvaneckas, E. Stankevicius, et al., "In silico testing of the semi-closed loop infusion system with a new simulator", Elektronika ir Elektrotechnika, vol. 20, no. 9, pp. 19-24, 2014. [Online]. Available: http://dx.doi.org/10.5755/j01.eee.20.9.8520

[5] A. Andrijauskas, V. Markevicius, D. Navikas, N. Porvaneckas, D. Andriukaitis, E. Stankevicius, et al., "In vivo testing of the semi-closed loop infusion system: the preliminary observations", Elektronika ir Elektrotechnika, vol. 21, no. 1, pp. 28-32, 2015. [Online]. Available: http://dx.doi.org/10.5755/j01.eee.21.L2458

[6] P. Ray, Y. Le Manach, B. Riou, T. T. Houle, "Statistical evaluation of a biomarker", Anesthesiology, vol. 112, pp. 1023-1040, 2010. [Online]. Available: http://dx.doi.org/10.1097/ALN.0b013e3181d 47604

[7] G. Kvederas, N. Porvaneckas, A. Andrijauskas, C. H. Svensen, J. Ivaskevicius, J. Mazunaitis, et al., "A randomized double-blind clinical trial of tourniquet application strategies for total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc", vol. 21, no. 12, pp. 2790-9, 2013. [Online]. Available: http://dx.doi.org/ 10.1007/s00167-012-2221-1

[8] M. Biais, S. Ehrmann, A. Mari, et al., "Clinical relevance of pulse pressure variations for predicting fluid responsiveness in mechanically ventilated intensive care unit patients: the grey zone approach", Critical Care, 2014, vol. 18, p. 587. [Online]. Available: http://ccforum.com/content/18/6/587

[9] M. Cannesson, "Assessing the diagnostic accuracy of pulse pressure variations for the prediction of fluid responsiveness: a "gray zone" approach", Anesthesiology, 2011, vol. 115, pp. 231-241. [Online]. Available: http://dx.doi.org/10.1097/ALN.0b013e318225b80a

[10] J. Carpenter, "Bootstrap confidence intervals: when, which, what? A practical guide for medical statisticians", Stat Med, vol. 19, pp. 1141-1164, 2000. [Online]. Available: http://dx.doi.org/ 10.1002/(SICI)1097-0258(20000515)19:9<1141::AID-SIM479>3.0. CO;2-F

[11] J. Coste, P. Jourdain, J. Pouchot, "A gray zone assigned to inconclusive results of quantitative diagnostic tests: Application to the use of brain natriuretic peptide for diagnosis of heart failure in acute dyspneic patients", Clin Chem, 2006, vol. 52, pp. 2229-2235. [Online]. Available: http://dx.doi.org/10.1373/clinchem.2006.072280

[12] J Coste, J. Pouchot, "A grey zone for quantitative diagnostic and screening tests", Int J Epidemiol, vol. 32, pp. 304-313, 2003. [Online]. Available: http://dx.doi.org/10.1093/ije/dyg054

[13] Y. Le Manach, C. K. Hofer, J. J. Lehot, B. Vallet, J. P. Goarin, B. Tavernier, M. Cannesson, "Can changes in arterial pressure be used to detect changes in cardiac output during volume expansion in the perioperative period?", Anesthesiology, vol. 117, pp. 1165-1174, 2012. [Online]. Available: http://dx.doi.org/10.1097/ALN.0b013 e318275561d

[14] R. G. Hahn, D. Drobin, J. Zdolsek, "Distribution of crystalloid fluid changes with the rate of infusion: a population-based study", Acta Anaesth Scand, 2016. [Online]. Available: http://dx.doi.org/10.1111/ aas.12686. [Epub ahead of print]

[15] A. J. Broderick, "Point-of-care haemoglobin measurement--state of the art or a bleeding nuisance?", Anaesthesia, vol. 70, no. 11, pp. 1225-9, 2015. [Online]. Available: http://dx.doi.org/10.1111/ anae.13231

[16] A. Andrijauskas, N. Porvaneckas, J. Sipylaite, E. Stankevicius, D. Cincikas, V. Uvarovas, et al., "A mini volume loading test using crystalloid boluses for indication of perioperative changes in hydration status. Medicina, vol. 52, no. 3, 2016, to be published.

Vytautas Markevicius (1), Darius Cincikas (2), Narunas Porvaneckas (3), Edgaras Stankevicius (4), Dangirutis Navikas (1), Darius Andriukaitis (1), Jurate Sipylaite (2), Saule Svediene (2), Valentinas Uvarovas (3), Igoris Satkauskas (2), Giedrius Kvederas (3), Algimantas Valinevicius (1), Mindaugas Zilys (1), Audrius Andrijauskas (2)

(1) Department of Electronics Engineering, Kaunas University of Technology, Studentu St. 50-418, LT-51368 Kaunas, Lithuania

(2) Vilnius University Faculty of Medicine, Clinic of Anaesthesiology and Intensive Care, Siltnamiu St. 29, LT-04130 Vilnius, Lithuania

(3) Vilnius University Faculty of Medicine, Clinic of Rheumatology, Orthopaedics-Traumatology and Reconstructive Surgery, Siltnamiu St. 29, LT-04130 Vilnius, Lithuania

(4) Institute of Physiology and Pharmacology, Lithuanian University of Health Sciences, A. Mickeviciaus St. 9, LT-44307 Kaunas, Lithuania

audrius.andrijauskas@mf.vu.lt
COPYRIGHT 2016 Kaunas University of Technology, Faculty of Telecommunications and Electronics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Markevicius, Vytautas; Cincikas, Darius; Porvaneckas, Narunas; Stankevicius, Edgaras; Navikas, Dangi
Publication:Elektronika ir Elektrotechnika
Article Type:Report
Geographic Code:1USA
Date:Feb 1, 2016
Words:4284
Previous Article:Serial-parallel IGBT connection method based on overvoltage measurement.
Next Article:Switched scalar optimal [mu]-law quantization with adaptation performed to both the variance and the distribution of speech signal.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters