Search

This paper presents the perception, mapping, and planning pipeline implemented on an autonomous race car. It was developed by the 2019 AMZ driverless team for the Formula Student Germany (FSG) 2019 driverless competition, where it won 1st place overall. The presented solution combines early fusion of camera and LiDAR data, a layered mapping approach, and a planning approach that uses Bayesian filtering to achieve high-speed driving on unknown race tracks while creating accurate maps. We benchmark the method against our team's previous solution, which won FSG 2018, and show improved accuracy when driving at the same speeds. Furthermore, the new pipeline makes it possible to reliably raise the maximum driving speed in unknown environments from 3~m/s to 12~m/s while still mapping with an acceptable RMSE of 0.29~m.

The aim was to investigate the applicability of a clinical decision support system in a real-world inpatient setting for patients with type 2 diabetes on general hospital wards. A total of 150 patients with type 2 diabetes requiring subcutaneous insulin therapy were treated with basal-bolus insulin therapy guided by a decision support system (GlucoTab) providing automated workflow tasks and suggestions for insulin dosing to health care professionals. By using the system, a mean daily blood glucose (BG) of 159 ± 32 mg/dL was achieved. 68.8% of measurements were in the target range (70 to 93% and 91%, respectively). The decision support system facilitates safe and efficacious inpatient diabetes care by standardizing treatment workflow and providing decision support for basal-bolus insulin dosing.

Visually poor scenarios are one of the main sources of failure in visual localization systems in outdoor environments. To address this challenge, we present MOZARD, a multi-modal localization system for urban outdoor environments using vision and LiDAR. By extending our preexisting key-point based visual multi-session local localization approach with the use of semantic data, an improved localization recall can be achieved across vastly different appearance conditions. In particular we focus on the use of curbstone information because of their broad distribution and reliability within urban environments. We present thorough experimental evaluations on several driving kilometers in challenging urban outdoor environments, analyze the recall and accuracy of our localization system and demonstrate in a case study possible failure cases of each subsystem. We demonstrate that MOZARD is able to bridge scenarios where our previous work VIZARD fails, hence yielding an increased recall performance, while a similar localization accuracy of 0.2m is achieved

Glycemic goals of inpatients should consider the prevention of hypoglycemia and hyperglycemia. To achieve glycemic control in the hospital, guidelines recommend basal-bolus insulin therapy as the preferred method for treatment of persistent hyperglycemia and the use of clinical decision support systems. The aim of this study was to investigate the efficacy, safety and usability of the GlucoTab® decision support system in hospital routine care of non-critically ill patients with type 2 diabetes at 3 general wards of a university medical center. In total, 150 patients with type 2 diabetes requiring insulin therapy (56 female, age 68 ± 11 years, HbA1c 77 ± 24 mmol/mol, diabetes duration 15 ± 11 years, BMI 30 ± 6 kg/m2, length of hospital stay 9 ± 7 days) were treated with basal-bolus insulin therapy guided by GlucoTab®, a mobile decision support system providing automated workflow tasks and suggestions for insulin dosing to nurses and physicians. In total 4879 capillary blood glucose values (BG) were documented. By using the GlucoTab® system in routine care, a mean daily BG of 159 ± 32 mg/dl could be achieved. 68.8% of all BG measurements were in the acceptable range (70-93% and 91%, respectively). The study confirmed that the GlucoTab® system supports a safe, efficacious and user-friendly glycemic management of basal-bolus insulin therapy in hospital routine care. Further research will include decision support for alternative therapy types for patients with less stringent goals to prevent hypoglycemia in hospital routine care. Disclosure K.M. Lichtenegger: None. F. Aberer: None. A.C. Tuca: None. K. Donsa: Stock/Shareholder; Self; decide Clinical Software GmbH. B. Höll: Employee; Self; decide Clinical Software GmbH. L. Schaupp: None. J. Plank: None. P. Beck: Employee; Self; decide Clinical Software GmbH. Stock/Shareholder; Self; decide Clinical Software GmbH. F. Fruhwald: None. T.R. Pieber: Advisory Panel; Self; ADOCIA, Arecor Limited, AstraZeneca, Novo Nordisk A/S, Sanofi. Speaker's Bureau; Self; Novo Nordisk A/S. J.K. Mader: Advisory Panel; Self; Boehringer Ingelheim International GmbH, Eli Lilly and Company, Prediktor Medical, Roche Diabetes Care, Sanofi. Speaker's Bureau; Self; Abbott, AstraZeneca, Dexcom, Inc., Novo Nordisk Inc. Stock/Shareholder; Self; decide Clinical Software GmbH. Funding Research Studio Austria "GlucoTab“ (844737)

Changes in appearance is one of the main sources of failure in visual localization systems in outdoor environments. To address this challenge, we present VIZARD, a visual localization system for urban outdoor environments. By combining a local localization algorithm with the use of multi-session maps, a high localization recall can be achieved across vastly different appearance conditions. The fusion of the visual localization constraints with wheel-odometry in a state estimation framework further guarantees smooth and accurate pose estimates. In an extensive experimental evaluation on several hundreds of driving kilometers in challenging urban outdoor environments, we analyze the recall and accuracy of our localization system, investigate its key parameters and boundary conditions, and compare different types of feature descriptors. Our results show that VIZARD is able to achieve nearly 100% recall with a localization accuracy below 0.5m under varying outdoor appearance conditions, including at night-time., Comment: 9 pages

We introduce a novel method for oriented place recognition with 3D LiDAR scans. A Convolutional Neural Network is trained to extract compact descriptors from single 3D LiDAR scans. These can be used both to retrieve near-by place candidates from a map, and to estimate the yaw discrepancy needed for bootstrapping local registration methods. We employ a triplet loss function for training and use a hard-negative mining strategy to further increase the performance of our descriptor extractor. In an extensive evaluation on the NCLT and KITTI datasets, we demonstrate that our method outperforms related state-of-the-art approaches based on both data-driven and handcrafted data representation in challenging long-term outdoor conditions., 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), ISBN:978-1-7281-4004-9, ISBN:978-1-7281-4003-2

Background: Diabetes management requires complex and interdisciplinary cooperation of health care professionals (HCPs). To support this complex process, IT-support is recommended by clinical guidelines. The aim of this article is to report on results from a clinical feasibility study testing the prototype of a mobile, tablet-based client-server system for computerized decision and workflow support (GlucoTab®) and to discuss its impact on hypoglycemia prevention. Methods: The system was tested in a monocentric, open, noncontrolled intervention study in 30 patients with type 2 diabetes mellitus (T2DM). The system supports HCPs in performing a basal-bolus insulin therapy. Diabetes therapy, adverse events, software errors and user feedback were documented. Safety, efficacy and user acceptance of the system were investigated. Results: Only 1.3% of blood glucose (BG) measurements were 300 mg/dl. The availability of the system (97.3%) and the rate of treatment activities documented with the system (>93.5%) were high. Only few suggestions from the system were overruled by the users (>95.7% adherence). Evaluation of the 3 anonymous questionnaires showed that confidence in the system increased over time. The majority of users believed that treatment errors could be prevented by using this system. Conclusions: Data from our feasibility study show a significant reduction of hypoglycemia by implementing a computerized system for workflow and decision support for diabetes management, compared to a paper-based process. The system was well accepted by HCPs, which is shown in the user acceptance analysis and that users adhered to the insulin dose suggestions made by the system.

Goal: Automated glucose control (AGC) has not yet reached the point where it can be applied clinically [3] . Challenges are accuracy of subcutaneous (SC) glucose sensors, physiological lag times, and both inter- and intraindividual variability. To address above issues, we developed a novel scheme for MPC that can be applied to AGC. Results: An individualizable generic whole-body physiology-based pharmacokinetic and dynamics (PBPK/PD) model of the glucose, insulin, and glucagon metabolism has been used as the predictive kernel. The high level of mechanistic detail represented by the model takes full advantage of the potential of MPC and may make long-term prediction possible as it captures at least some relevant sources of variability [4] . Robustness against uncertainties was increased by a control cascade relying on proportional-integrative derivative-based offset control. The performance of this AGC scheme was evaluated in silico and retrospectively using data from clinical trials. This analysis revealed that our approach handles sensor noise with a MARD of 10%–14%, and model uncertainties and disturbances. Conclusion: The results suggest that PBPK/PD models are well suited for MPC in a glucose control setting, and that their predictive power in combination with the integrated database-driven ( a priori individualizable) model framework will help overcome current challenges in the development of AGC systems. Significance: This study provides a new, generic, and robust mechanistic approach to AGC using a PBPK platform with extensive a priori (database) knowledge for individualization.

Objective Most preventable adverse drug events and medication errors occur during medication ordering. Medication order entry and clinical decision support are available on paper or as computerized systems. In this post-hoc analysis we investigated frequency and clinical impact of blood glucose (BG) documentation- and user-related calculation errors as well as workflow deviations in diabetes management. We aimed to compare a paper-based protocol to a computerized medication management system combined with clinical workflow and decision support. Methods Seventy-nine hospitalized patients with type 2 diabetes mellitus were treated with an algorithm driven basal-bolus insulin regimen. BG measurements, which were the basis for insulin dose calculations, were manually entered either into the paper-based workflow protocol (PaperG: 37 patients) or into GlucoTab ® —a mobile tablet PC based system (CompG: 42 patients). We used BG values from the laboratory information system as a reference. A workflow simulator was used to determine user calculation errors as well as workflow deviations and to estimate the effect of errors on insulin doses. The clinical impact of insulin dosing errors and workflow deviations on hypo- and hyperglycemia was investigated. Results The BG documentation error rate was similar for PaperG (4.9%) and CompG group (4.0%). In PaperG group, 11.1% of manual insulin dose calculations were erroneous and the odds ratio (OR) of a hypoglycemic event following an insulin dosing error was 3.1 (95% CI: 1.4–6.8). The number of BG values influenced by insulin dosing errors was eightfold higher than in the CompG group. In the CompG group, workflow deviations occurred in 5.0% of the tasks which led to an increased likelihood of hyperglycemia, OR 2.2 (95% CI: 1.1–4.6). Discussion Manual insulin dose calculations were the major source of error and had a particularly strong influence on hypoglycemia. By using GlucoTab ® , user calculation errors were entirely excluded. The immediate availability and automated handling of BG values from medical devices directly at the point of care has a high potential to reduce errors. Computerized systems facilitate the safe use of more complex insulin dosing algorithms without compromising usability. In CompG group, missed or delayed tasks had a significant effect on hyperglycemia, while in PaperG group insufficient precision of documentation times limited analysis. The use of old BG measurements was clinically less relevant. Conclusion Insulin dosing errors and workflow deviations led to measurable changes in clinical outcome. Diabetes management systems including decision support should address nurses as well as physicians in a computerized way. Our analysis shows that such systems reduce the frequency of errors and therefore decrease the probability of hypo- and hyperglycemia.

Aims To find an explanation for the lower potency of insulin detemir observed in humans compared with unmodified human insulin by investigating insulin detemir and human insulin concentrations directly at the level of peripheral insulin-sensitive tissues in humans in vivo. Methods Euglycaemic-hyperinsulinaemic clamp experiments were performed in healthy volunteers. Human insulin was administered i.v. at 6 pmol/kg/min and insulin detemir at 60 pmol/kg/min, achieving a comparable steady-state pharmacodynamic action. In addition, insulin detemir was doubled to 120 pmol/kg/min. Minimally invasive open-flow microperfusion (OFM) sampling methodology was combined with inulin calibration to quantify human insulin and insulin detemir in the interstitial fluid (ISF) of subcutaneous adipose and skeletal muscle tissue. Results The human insulin concentration in the ISF was ∼115 pmol/l or ∼30% of the serum concentration, whereas the insulin detemir concentration in the ISF was ∼680 pmol/l or ∼2% of the serum concentration. The molar insulin detemir interstitial concentration was five to six times higher than the human insulin interstitial concentration and metabolic clearance of insulin detemir from serum was substantially reduced compared with human insulin. Conclusions OFM proved useful for target tissue measurements of human insulin and the analogue insulin detemir. Our tissue data confirm a highly effective retention of insulin detemir in the vascular compartment. The higher insulin detemir relative to human insulin tissue concentrations at comparable pharmacodynamics, however, indicate that the lower potency of insulin detemir in humans is attributable to a reduced effect in peripheral insulin-sensitive tissues and is consistent with the reduced in vitro receptor affinity.

Background: This study investigated the efficacy, safety, and usability of standardized glycemic management by a computerized decision support system for non-critically ill hospitalized patients with type 2 diabetes on four different wards. Materials and Methods: In this open, noncontrolled intervention study, glycemic management of 99 patients with type 2 diabetes (62% acute admissions; 41 females; age, 67±11 years; hemoglobin A1c, 65±21 mmol/mol; body mass index, 30.4±6.5 kg/m2) on clinical wards (Cardiology, Endocrinology, Nephrology, Plastic Surgery) of a tertiary-care hospital was guided by GlucoTab® (Joanneum Research GmbH [Graz, Austria] and Medical University of Graz [Graz, Austria]), a mobile decision support system providing automated workflow support and suggestions for insulin dosing to nurses and physicians. Results: Adherence to insulin dosing suggestions was high (96.5% bolus, 96.7% basal). The primary outcome measure, percentage of blood glucose (BG) measurements in the range of 70–140 mg/dL, occurred in 50.2±22.2% of all measurements. The overall mean BG level was 154±35 mg/dL. BG measurements in the ranges of 60–70 mg/dL, 40–60 mg/dL, and

Inpatient glucose management is based on four daily capillary blood glucose (BG) measurements. The aim was to test the capability of continuous glucose monitoring (CGM) for assessing the clinical impact and safety of basal-bolus insulin therapy in non-critically ill hospitalized patients with type 2 diabetes mellitus (T2DM).Eighty-four patients with T2DM (age, 68±10 years; glycosylated hemoglobin, 72±28 mmol/mol; body mass index, 31±7 kg/m(2)) were treated with basal-bolus insulin. CGM was performed with the iPro(®)2 system (Medtronic MiniMed, Northridge, CA) and calibrated retrospectively.A remarkable consistency between CGM and BG measurements and therapy improvement was shown over the study period of 501 patient-days. The number of CGM and BG measurements (CGM/BG) in the range from 3.9-10 mmol/L increased from 67.7%/67.2% (on Day 1) to 77.5%/78.6% (on the last day) (P0.04). The number of low glycemic episodes (3.3 to3.9 mmol/L) during nighttime detected by CGM was 15-fold higher, and the number of episodes13.9 mmol/L detected by CGM during night was 12.5-fold higher than the values from the BG measurements. Ninety-nine percent of data points were in the clinically accurate or acceptable Clarke Error Grid Zones A+B, and the relative numbers of correctly identified episodes of3.9 and13.9 mmol/L detected by CGM (sensitivity) were 47.3% and 81.5%, respectively.Our data exhibit a good agreement between overall CGM and BG measurements, but there were a high number of missed hypo- and hyperglycemic episodes with BG measurements, particularly during nighttime. Overall assessment of glycemic control using CGM is feasible, whereas the use of CGM for individualized therapy decisions needs further improvement.

In this paper we describe the concept and in vivo results of a minimally invasive, chip-based near infrared (NIR) sensor, combined with microdialysis, for continuous glucose monitoring. The sensor principle is based on difference absorption spectroscopy in selected wavelength bands of the near infrared spectrum (1300 nm, 1450 nm, and 1550 nm) in the 1st overtone band. In vitro measurements revealed a linear relationship between glucose concentration and the integrated difference spectroscopy signal with a coefficient of determination of 99% in the concentration range of 0-400mg/dl. The absolute error in this case is about 5mg/dl, corresponding to a relative error of about 5% for glucose concentrations larger than 50mg/dl and about 12% in the hypoglycemic range (50mg/dl). In vivo measurements on 10 patients showed that the NIR-CGM sensor data reflects the blood reference values adequately, if a proper calibration and a signal drift correction is applied. The mean MARE (mean absolute relative error) value taken over all patient data is 13.8%. The best achieved MARE value is at 4.8%, whereas the worst lies at 25.8%, leading to a standard deviation of 5.5%.

Summary Background: Standardized insulin order sets for subcutaneous basal-bolus insulin therapy are recommended by clinical guidelines for the inpatient management of diabetes. The algorithm based GlucoTab system electronically assists health care personnel by supporting clinical workflow and providing insulin-dose suggestions. Objective: To develop a toolbox for improving clinical decision-support algorithms. Methods: The toolbox has three main components. 1) Data preparation: Data from several heterogeneous sources is extracted, cleaned and stored in a uniform data format. 2) Simulation: The effects of algorithm modifications are estimated by simulating treatment workflows based on real data from clinical trials. 3) Analysis: Algorithm performance is measured, analyzed and simulated by using data from three clinical trials with a total of 166 patients. Results: Use of the toolbox led to algorithm improvements as well as the detection of potential individualized subgroup-specific algorithms. Conclusion: These results are a first step towards individualized algorithm modifications for specific patient subgroups. Citation: Donsa K, Beck P, Plank J, Schaupp L, Mader JK, Truskaller T, Tschapeller B, Höll B, Spat S, Pieber TR. A toolbox to improve algorithms for insulin-dosing decision support. Appl Clin Inf 2014; 5: 548–556 http://dx.doi.org/10.4338/ACI-04-RA-0033

We report a novel approach to quantify interstitial analytes in living tissue by combining open-flow microperfusion (OFM) with a sensor and the re-circulation method. OFM is based on the unrestricted exchange of molecules between the interstitial fluid (ISF) and a perfusion medium through macroscopic perforations that enables direct access to the ISF. By re-circulating the perfusate and monitoring the changes of the analytes' concentration with a sensor, the absolute analyte concentration in the ISF can be calculated. In order to validate the new concept, the absolute electrical conductivity of the ISF was identified in six subjects to be 1.33 ± 0.08 S/m (coefficient of variation CV = 6 %), showing the robustness of this approach. The most striking feature of this procedure is the possibility to monitor several compounds simultaneously by applying different sensors which will allow not only the determination of the concentration of a single substance in vivo but also the simultaneous dynamics of different analytes. This will open new fields in analytical chemistry, pharmacology, as well as clinical experimental research.

A1 Meet the needs: Important questions for ICU relatives Magdalena M. Hoffmann1,2, Anna K. Holl3, Harald Burgsteiner4, Thomas Pieber1,5, Philipp Eller6, Karin Amrein1 A2 Daily use of the Surgical Safety Checklist: results of a real-time audit Gerald Sendlhofer1,2,3, Karina Leitgeb1, Veronika Gombotz1, Peter Tiefenbacher1, Lars-Peter Kamolz2,3 A3 Best-practice projects concerning patient safety in Austria Gerald Sendlhofer1,2,3,4, , Karina Leitgeb1,4, Magdalena Hoffmann1,4, Sabine Papst1,4, Susanne Gasteiner1,4, Lars Peter Kamolz2,3,4 A4 Comprehensive Cancer Center Graz: evaluation of availability of mandatory healthcare professionals in tumorboards Peter Tiefenbacher1,2, Yvonne Muller1,2, Gerald Sendlhofer1,2,3,4 A5 Comprehensive Cancer Center Graz: protocol to evaluate communication and recommendation finding in tumorboards Marlies Hart1, Peter Tiefenbacher1,2, Yvonne Muller1,2, Gerald Sendlhofer1,2,3,4 A6 Generating a new insulin prescription chart at a university hospital Julia Kopanz1, Katharina M. Neubauer1, Gerald Sendlhofer2,3,4, Barbara Semlitsch1, Andres Pak5, Gerald Cuder1, Thomas R. Pieber1, Johannes Plank6 A7 The Speech and Language Therapy teaching practice at the Institute for Speech and Language Therapy at FH JOANNEUM in Graz Annemarie Schinko1, Angelika Rother1 A8 Development and implementation of a progress test in undergraduate dental education: a prospective Austrian pilot project Barbara Kirnbauer1, Petra Rugani1, Norbert Jakse1, Johannes Bernhardt-Melischnig2, Rudolf Egger3 A9 What they need and what they get: protocol to identify the needs of information of patients and what they receive so far Magdalena M. Hoffmann1,2, Karina Leitgeb1,3, Esther Trampusch4,5, Angelika Hofer4, Regina Riedl6, Karin Amrein2, Christa Tax7, Gernot Brunner7, Gerald Sendlhofer,1,3,8 A 10 Medication-related problems resolved and money saved: results of a clinical pharmacy service evaluation in the surgical setting Emilie Tudela-Lopez1, Petra Polzleitner1, Klara Jadrna1, Christina Labut1, Maria Kundracikova1, Helga Fend1, Gunar Stemer1 A 11 Intercultural communication management in radiology Jutta Maria Lorenz1 A 12 Spotlight on data quality: comparison of data input by physicians vs. non-physicians in the German Resuscitation Registry Gerhard Prause1, Paul Zajic1, Philipp Zoidl1, Geza Gemes1 A 13 MRSA prevalence and eradication with octenidine Gerald Pichler1, Christian Pux1, Rita Babeluk2, Brigitte Hermann3, Eric Stoiser3, Antonella De Campo4, Andrea Grisold5, Ines Zollner-Schwetz6, Robert Krause6, Walter Schippinger4 A 14 QUIPS: a pain registry with impact in science and daily routine work Alexander Avian1,2, Brigitte Messerer3, Claudia Weinmann2, Winfried Meisner2 A 15 Patient empowerment for the youngest Karina Leitgeb1, Magdalena Hoffmann1, Gerald Sendlhofer1,2,3 A 16 Development of a pocket guide for parenteral nutrition in hospitalized adults Anna Maria Eisenberger1, Siegrid Fuchs2 A 17 Safety and efficacy of a clinical decision support system for blood glucose management in patients with diabetes mellitus type 2 at a plastic-surgical ward Alexandru-C. Tuca1, Katharina M. Neubauer2, Julia K. Mader2, Felix Aberer2 , David B. Lumenta1, Birgit Bauer1 ,Klaus Donsa3 , Thomas Augustin3 , Bernhard Holl3 , Lukas Schaupp2, Peter Beck4, Johannes Plank2 , Thomas R. Pieber2,3, Lars-P. Kamolz1,5 A 18 Urban dwellers show higher readiness in participating actively in the shaping of the healthcare system in Austria Christian Smolle1, Frederike Reischies2, Gudrun Pregartner3, Daryousch Parvizi2, Gerald Sendlhofer1,4, Lars-Peter Kamolz1,4

Background Methodologies for continuous sampling of lipophilic drugs and high-molecular solutes in the dermis are currently lacking. We investigated the feasibility of sampling a lipophilic topical drug and the locally released biomarker in the dermis of non-lesional and lesional skin of psoriatic patients over 25 h by means of membrane-free dermal open-flow microperfusion probes (dOFM) and novel wearable multi-channel pumps. Methods Nine psoriatic patients received a topical p-38 inhibitor (BCT194, 0.5% cream) on a lesional and a non-lesional application site once daily for 8 days. Multiple dOFM sampling was performed for 25 h from each site on day 1 and day 8. Patients were mobile as dOFM probes were operated by a novel light-weight push–pull pump. Ultrasound was used to verify intradermal probe placement, cap-LC–MS/MS for BCT194 and ELISA for TNFα analysis. Results dOFM was well tolerated and demonstrated significant drug concentrations in lesional as well as non-lesional skin after 8 days, but did not show significant differences between tissues. On day 8, TNFα release following probe insertion was significantly reduced compared to day 1. Conclusions Novel membrane-free probes and wearable multi-channel pumps allowed prolonged intradermal PK/PD profiling of a lipophilic topical drug in psoriatic patients. This initial study shows that dOFM overcomes limitations of microdialysis sampling methodology, and it demonstrates the potential for PK/PD studies of topical products and formulations in a clinical setting.

After decades of research, Automated Glucose Control (AGC) is still out of reach for everyday control of blood glucose. The inter- and intra-individual variability of glucose dynamics largely arising from variability in insulin absorption, distribution, and action, and related physiological lag-times remain a core problem in the development of suitable control algorithms. Here, we demonstrate the potential of approaches based on detailed models to improve controller performance by comparing two control algorithms. The performance of fading memory proportional derivative control (FMPD) with varying degrees of model-based information on the effects of insulin on board (EIOB) is compared to an ideal nonlinear model predictive control (MPC) scheme. We conclude, that large-scale in silico models of the glucose metabolism can provide a framework to improve diabetes research, the development of automatic control strategies for diabetes and ultimately every day diabetes management.

Glycemic control of intensive care patients can be beneficial for this patient group but the continuous determination of their glucose concentration is challenging. Current continuous glucose monitoring systems based on the measurement of interstitial fluid glucose concentration struggle with sensitivity losses, resulting from biofouling or inflammation reactions. Their use as decision support systems for the therapeutic treatment is moreover hampered by physiological time delays as well as gradients in glucose concentration between plasma and interstitial fluid. To overcome these drawbacks, we developed and clinically evaluated a system based on microdialysis of whole blood. Venous blood is heparinised at the tip of a double lumen catheter and pumped through a membrane based micro-fluidic device where protein-free microdialysate samples are extracted. Glucose recovery as an indicator of long term stability was studied in vitro with heparinised bovine blood and remained highly stable for 72 h. Clinical performance was tested in a clinical trial in eight healthy volunteers undergoing an oral glucose tolerance test. Glucose concentrations of the new system and the reference method correlated at a level of 0.96 and their mean relative difference was 1.9 +/- 11.2%. Clinical evaluation using Clark's Error Grid analysis revealed that the obtained glucose concentrations were accurate and clinically acceptable in 99.6% of all cases. In conclusion, results of the technical and clinical evaluation suggest that the presented device delivers microdialysate samples suitable for accurate and long term stable continuous glucose monitoring in blood.

OBJECTIVE To examine insulin's effect on the tissue glucose concentration at the site of subcutaneous insulin administration. RESEARCH DESIGN AND METHODS A CMA-60 microdialysis (MD) catheter and a 24-gauge microperfusion (MP) catheter were inserted into the subcutaneous adipose tissue of fasting, healthy subjects (n = 5). Both catheters were perfused with regular human insulin (100 units/ml) over a 6-h period and used for glucose sampling and simultaneous administration of insulin at sequential rates of 0.33, 0.66, and 1.00 units/h (each rate was used for 2 h). Before and after the insulin delivery period, both catheters were perfused with an insulin-free solution (5% mannitol) for 2 h and used for glucose sampling only. Blood plasma glucose was clamped at euglycemic levels during insulin delivery. RESULTS Start of insulin delivery with MD and MP catheters resulted in a decline of the tissue glucose concentration and the tissue-to-plasma glucose ratio (TPR) for ∼60 min (P < 0.05). However, during the rest of the 6-h period of variable insulin delivery, tissue glucose concentration paralleled the plasma glucose concentration, and the TPR for MD and MP catheters remained unchanged at 83.2 ± 3.1 and 77.1 ± 4.8%, respectively. After subsequent switch to insulin-free perfusate, tissue glucose concentration and TPR increased slowly and reattained preinsulin delivery levels by the end of the experiments. CONCLUSIONS The results show the attainment of a stable TPR value at the site of insulin administration, thus indicating that insulin delivery and glucose sensing may be performed simultaneously at the same adipose tissue site.

OBJECTIVE To simplify and improve the treatment of patients with type 1 diabetes, we ascertained whether the site of subcutaneous insulin infusion can be used for the measurement of glucose. RESEARCH DESIGN AND METHODS Three special indwelling catheters (24-gauge microperfusion [MP] catheters) were inserted into the subcutaneous adipose tissue of subjects with type 1 diabetes (n = 10; all C-peptide negative). One MP catheter was perfused with short-acting insulin (100 units/ml, Aspart) and used for insulin delivery and simultaneous glucose sampling during an overnight fast and after ingestion of a standard glucose load (75 g). As controls, the further two MP catheters were perfused with an insulin-free solution (5% mannitol) and used for glucose sampling only. Plasma glucose was measured frequently at the bedside. RESULTS Insulin delivery with the MP catheter was adequate to achieve and maintain normoglycemia during fasting and after glucose ingestion. Tissue glucose concentrations derived with the insulin-perfused catheter agreed well with plasma glucose levels. Median correlation coefficient and median absolute relative difference values were found to be 0.93 (interquartile range 0.91–0.97) and 10.9%, respectively. Error grid analysis indicated that the percentage number of tissue values falling in the clinically acceptable range is 99.6%. Comparable analysis results were obtained for the two mannitol-perfused catheters. CONCLUSIONS Our data suggest that estimation of plasma glucose concentrations from the glucose levels directly observed at the site of subcutaneous insulin infusion is feasible and its quality is comparable to that of estimating plasma glucose concentrations from glucose levels measured in insulin-unexposed subcutaneous tissue.

Intensive insulin therapy reduces mortality and morbidity in critically ill patients but imposes great demands on medical staff who must take frequent blood samples for the determination of glucose levels. A solution to this resourcing problem would be provided by an automated blood monitoring system. The aim of the present clinical study was to evaluate such a system comprising an automatic blood sampling unit linked to a glucose biosensor. Our approach was to determine the correlation and system error of the sampling unit alone and of the combined system with respect to reference levels over 12h in humans. Two venous cannulae were inserted to connect the automatic and reference systems to the subjects. Blood samples were taken at 15 and 30 min intervals. The median Pearson coefficient of correlation between manually and automatically withdrawn blood samples was 0.982 for the sampling unit alone and 0.950 for the complete system. The biosensor had a linear range up to 20 mmoll(-1) and a 95% response time of

A new optical fiber microsensor for continuous measurement of glucose in subcutaneous tissue is described. The hybrid sensor consists of two oxygen optodes wherein one optode contains immobilized glucose oxidase. The glucose signal which is a difference between p O 2 of both optodes allows the compensation of oxygen fluctuations in intercellular fluid. Based on the results of investigations of the oxygen dependence on sensor response, the optimization by thickening the diffusion barrier of glucose sensitive optode is done. The loss of the sensing resolution for glucose is compensated by the increased sensitivity of the oxygen transducer at the low oxygen concentrations. The optimized sensor is free of oxygen fluctuations in the expected p O 2 range between 53 and 95 hPa and the sensor output is linear in the physiological relevant glucose range up to 20 mM. The use of the reference oxygen sensor allows also the compensation of the slight temperature fluctuations.

In this work we present a low cost, minimally invasive, and chip-based near infrared (NIR) sensor, combined with subcutaneous microdialysis, for continuous glucose monitoring (CGM). The sensor principle is based on difference absorption spectroscopy in the 1st overtone band known to be dominated by glucose-specific absorption features. The device comprises a multi-emitter LED and InGaAs-photodiodes, which are located on a single electronic board (non-disposable part), connected to a personal computer via Bluetooth. The disposable part consists of a chip containing the fluidic connections for microdialysis, two fluidic channels acting as optical transmission cells and total internally reflecting mirrors for in- and out-coupling of the light to the chip and to the detectors. The use of the sensor in conjunction with a subcutaneous microdialysis catheter to separate the glucose from the cells and proteins has been demonstrated to be extremely useful and advantageous for obtaining continuous glucose monitoring data and detecting glycemic levels in real time for a long period. Several in vitro and in vivo experiments were conducted to test the reliability of the device. In vitro measurements showed a linear relationship between glucose concentration and the integrated difference signal with a coefficient of determination of 99 % at the physiological concentration range. Clinical trial on 6 subjects with Type 1 diabetes showed that the NIR-CGM sensor data reflects the blood reference values adequately, if a proper calibration and signal drift compensation is applied. The MARD (mean absolute relative difference) value taken on retrospective data over all subjects is 8.5 % (range 6-11.5 %).

Insulin’s action to stimulate glucose utilization is determined by the insulin concentration in interstitial fluid (ISF) of insulin-sensitive tissues. The concentration of interstitial insulin has been measured in human subcutaneous adipose tissue and skeletal muscle, however, never in parallel. The aim of this study was to compare interstitial insulin levels between both tissue beds by simultaneous measurements and to verify and quantify low peripheral ISF insulin fractions as found during moderate hyperinsulinemia. Nine healthy subjects (27.2 ± 0.8 yr) were investigated. A euglycemic-hyperinsulinemic clamp was started with a primed-constant intravenous insulin infusion of 1 mU·kg−1·min−1. For direct access to ISF, macroscopically perforated open-flow microperfusion catheters were inserted in both tissues. During steady-state conditions (9.5 h), interstitial effluents were collected in 30-min fractions using five different insulin concentrations in the inflowing perfusates (“no net flux” protocol). Regression analysis of insulin concentrations in perfusates and effluents yielded the relative recovery and the perfusate insulin concentration, which was in equilibrium with the surrounding tissue. Thus, in subcutaneous adipose tissue and skeletal muscle, the mean ISF-to-serum insulin level was calculated as 21.0% [95% confidence interval (CI) 17.5–24.5] and 26.0% (95% CI 19.1–32.8; P = 0.14), respectively. Recoveries for insulin averaged 51 and 64%, respectively. The data suggest that the concentrations of insulin arising in healthy subjects at the level of ISF per se are comparable between subcutaneous adipose and skeletal muscle tissue. The low interstitial insulin fractions seem to confirm reports of low peripheral insulin levels during moderate insulin clamps.

We investigated insulin lispro kinetics with bolus and continuous subcutaneous insulin infusion (CSII) modes of insulin delivery. Seven subjects with type-1 diabetes treated by CSII with insulin lispro have been studied during prandial and postprandial conditions over 12 hours. Eleven alternative models of insulin kinetics have been proposed implementing a number of putative characteristics. We assessed 1) the effect of insulin delivery mode, i.e., bolus or basal, on the insulin absorption rate, the effects of 2) insulin association state and 3) insulin dose on the rate of insulin absorption, 4) the remote insulin effect on its volume of distribution, 5) the effect of insulin dose on insulin disappearance, 6) the presence of insulin degradation at the injection site, and finally 7) the existence of two pathways, fast and slow, of insulin absorption. An iterative two-stage parameter estimation technique was used. Models were validated through assessing physiological feasibility of parameter estimates, posterior identifiability, and distribution of residuals. Based on the principle of parsimony, best model to fit our data combined the slow and fast absorption channels and included local insulin degradation. The model estimated that 67(53-82)% [mean (interquartile range)] of delivered insulin passed through the slow absorption channel [absorption rate 0.011(0.004-0.029) min/sup -1/] with the remaining 33% passed through the fast channel [absorption rate 0.021(0.011-0.040) min/sup -1/]. Local degradation rate was described as a saturable process with Michaelis-Menten characteristics [V/sub MAX/=1.93(0.62-6.03) mU min/sup -1/, K/sub M/=62.6(62.6-62.6) mU]. Models representing the dependence of insulin absorption rate on insulin disappearance and the remote insulin effect on its volume of distribution could not be validated suggesting that these effects are not present or cannot be detected during physiological conditions.

We investigated the dynamic relationship between interstitial glucose (IG) in the subcutaneous adipose tissue and plasma glucose (PG) during physiologic conditions in type 1 diabetes mellitus (T1DM). Nine subjects with T1DM (5/4 M/F; age, 33 ± 13 years; body mass index, 26.6 ± 4.3 kg/m 2 ; glycosylated hemoglobin [HbA 1c ], 8.6% ± 0.9%; mean ± SD) treated by continuous subcutaneous insulin infusion (CSII) with insulin lispro were studied over 12 hours after a standard meal (40 g carbohydrate [CHO]) and prandial insulin. IG was measured by open flow microperfusion. Nine compartment models were postulated to account for temporal variations in the IG/PG ratio. The models differed in the inclusion of physiologically motivated alterations of pathways entering/leaving the IG compartment in the adipose tissue. The best model included zero order (constant) glucose disposal from the interstitial fluid (ISF) and insulin-stimulated glucose transfer from plasma to the ISF. The former effect is expressed by a positive association between the IG/PG ratio and PG, eg, a decrease in PG from 9 to 3.3 mmol/L lowers the IG/PG ratio by 0.1. The latter effect results in the IG/PG ratio to be increased by 0.03 per 10 mU/L of plasma insulin. We were not able to detect the stimulatory effect of insulin on glucose disappearance from the ISF. In conclusion, we developed and quantified a model of IG kinetics in the adipose tissue applicable to physiologic conditions in subjects with T1DM.

The objective of the project Advanced Insulin Infusion using a Control Loop (ADICOL) was to develop a treatment system that continuously measures and controls the glucose concentration in subjects with type 1 diabetes. The modular concept of the ADICOL's extracorporeal artificial pancreas consisted of a minimally invasive subcutaneous glucose system, a handheld PocketPC computer, and an insulin pump (D-Tron, Disetronic, Burgdorf, Switzerland) delivering subcutaneously insulin lispro. The present paper describes a subset of ADICOL activities focusing on the development of a glucose controller for semi-closed-loop control, an in silico testing environment, clinical testing, and system integration. An incremental approach was adopted to evaluate experimentally a model predictive glucose controller. A feasibility study was followed by efficacy studies of increasing complexity. The ADICOL project demonstrated feasibility of a semi-closed-loop glucose control during fasting and fed conditions with a wearable, modular extracorporeal artificial pancreas.

We studied the kinetics of glucose exchange between plasma and interstitial fluid (ISF) in human skeletal muscle and adipose tissue under fasting conditions. Five normal human subjects received an intravenous [6,6-2H2]glucose infusion in a prime-continuous fashion. During the tracer infusion, the open-flow microperfusion technique was employed to frequently sample ISF from quadriceps muscle and subcutaneous adipose tissue. The tracer glucose kinetics observed in muscle and adipose tissue ISF were found to be well described by a capillary-tissue exchange model. As a measure of transcapillary glucose exchange efficiency, the 95% equilibrium time was calculated from the identified model parameters. This time constant was similar for skeletal muscle and adipose tissue (28.6 ± 3.2 vs. 26.8 ± 3.6 min; P = 0.60). Furthermore, we found that the (total) interstitial glucose concentration was significantly lower ( P < 0.01) in muscle (3.32 ± 0.46 mmol/l) and adipose tissue (3.51 ± 0.17 mmol/l) compared with arterialized plasma levels (5.56 ± 0.13 mmol/l). Thus the observed gradients and dynamic relationships between plasma and ISF glucose in muscle and adipose tissue provide evidence that transcapillary exchange of glucose is limited in these two tissues under fasting conditions.

The absolute concentration of albumin was measured in the interstitial fluid of subcutaneous adipose tissue and skeletal muscle in six healthy volunteers by combining the method of open-flow microperfusion and the no-net-flux calibration technique. By use of open-flow microperfusion, four macroscopically perforated double lumen catheters were inserted into the tissue regions of interest and constantly perfused. Across the macroscopic perforations of the catheters interstitial fluid was partially recovered in the perfusion fluid. Catheters were perfused with five solutions, each containing different concentrations of albumin. Absolute interstitial albumin concentrations were calculated by applying linear regression analysis to perfusate vs. sampled albumin concentration (no-net-flux calibration technique). Interstitial albumin concentrations were significantly lower ( P < 0.0001) in adipose tissue (7.36 g/l; r = 0.99, P < 0.0003; range: 4.3–10.7 g/l) and in skeletal muscle (13.25 g/l; r = 0.99, P < 0.0012; range: 9.7 to 15.7 g/l) compared with the serum concentration (48.9 ± 0.7 g/l, mean ± SE, n = 6; range: 46.4–50.4 g/l). Furthermore, interstitial albumin concentrations were significantly higher in skeletal muscle compared with adipose tissue ( P < 0.01). The study indicates that open-flow microperfusion allows stable sampling of macromolecules from the interstitial space of peripheral tissue compartments. Moreover, the present data report for the first time in healthy humans in vivo the true albumin concentrations of interstitial fluid of adipose tissue and skeletal muscle.

The objective of the study was to evaluate the performance of an amperometric enzyme based lactate sensor and to investigate the possibility of replacing a double lumen catheter based blood withdrawal system with a heparin coated single lumen system. The inner lumen of a double lumen catheter which was placed in a peripheral vein was perfused with heparin solution. The outer lumen was used to collect heparinized blood samples at a defined flow rate. The single lumen system was attached to a heparinized catheter which was also placed in a peripheral vein. The undiluted blood samples were collected at a specified flow rate. A sensor flow chamber incorporating an amperometric thin-film lactate microbiosensor was placed in the sampling line for real-time lactate monitoring. Plasma lactate concentrations were measured during frequently performed hyperlactatemia bicycle ergometer experiments in six healthy volunteers (age 25.8 +/- 2.8 years, BMI 22.7 +/- 1 kg/m2). Additionally, plasma lactate was measured in real-time using the lactate sensors. The first three experiments were performed with a double lumen based catheter system whereas the following three experiments were performed with a heparin coated catheter system. The correlation coefficients of sensor readings and laboratory analyzer results in all six experiments were between 0.93 and 0.99, respectively (P < 0.001). The miniaturized lactate sensors showed a linear range up to 25 mmol/l lactate concentration and 95% response times < 30 s in undiluted serum. During the experiments maximum lactate concentrations of 14 mmol/l were achieved. Improvements of system performance using heparin coated catheter systems could be shown. The overall SD of the sensor readings compared to laboratory results using three double lumen catheter based systems was 0.91 mmol/l whereas the SD using three heparin coated systems was 0.65 mmol/l. In summary, real-time monitoring of lactate in human whole blood is feasible with such a device and can be improved by using heparin coated catheter systems.

The present study was carried out to investigate in vivo in healthy humans the method of open-flow microperfusion for monitoring of the subcutaneous (s.c.) lactate concentration during rest and cycle ergometer exercise. Using open-flow microperfusion, a perforated double lumen catheter with an inflow and an outflow connection is inserted into the s.c. adipose tissue and perfused with a sterile, isotonic, ionfree fluid. Due to the low flow rate, the fluid partially equilibrates with the surrounding tissue. The equilibrated perfusate passes a sensor flow chamber where the substance of interest and the rate of recovery (i.e. the ratio of sampled concentration to interstitial concentration) are continuously monitored. Within this study, the method was evaluated in four healthy volunteers during cycle ergometer exercise. The relative increase of the lactate concentration was approximately a third in the s.c. tissue compared to the capillary blood and the peak time was delayed on average by 10 min. The correlation coefficient between blood and s.c. tissue lactate concentration ranged from r = 0.41 to r = 0.90 (n = 29) in the individual experiments. The combination of open-flow microperfusion and lactate and conductivity sensors enables on-line monitoring of the s.c. lactate concentration without in vivo calibration during steady-state and cycle ergometer exercise.

OBJECTIVE To evaluate the clinical and analytical accuracy of home blood glucose meters. RESEARCH DESIGN AND METHODS Six blood glucose meters—Reflolux S (Boehringer Mannheim, Mannheim, Germany), One Touch II (LifeScan, Milpitas, CA), Glucocard Memory (Menarini, Florence, Italy), Precision QID (Medisense, Cambridge, U.K.), HaemoCue (HaemoCue, Ängelholm, Sweden), and Accutrend a (Boehringer Mannheim, Mannheim, Germany)—were compared with a reference method (Beckman Glucose Analyzer II) under controlled conditions (glucose clamp technique). Validation of the blood glucose meters was accomplished by clinically oriented approaches (error grid analysis), statistical approaches (variance components analysis), and by the criteria of the American Diabetes Association (ADA), which recommend a target variability of RESULTS A total of 1,794 blood glucose monitor readings and 299 reference values ranging from 2.2 to 18.2 mmol/1 were analyzed (705 readings 9.99 mmol/1). According to error grid analysis, only Reflolux S and Glucocard M had 100% of estimations within the clinically acceptable zones A and B. Assessment of analytical accuracy revealed substantial differences between the glucose meters after separation of the data into defined glycemic ranges. None of the devices met the ADA criteria. CONCLUSIONS To evaluate accuracy of blood glucose meters, error grid analysis, as well as statistical models, are helpful means and should be performed together. Analytical performance of currently available home blood glucose meters differs substantially within defined glycemic ranges.

Hyperglycaemia in hospitalized patients is a common and costly health care problem. The GlucoTab system is a mobile workflow and decision support system, aiming to facilitate efficient and safe glycemic control of non-critically ill patients. Being a medical device, the GlucoTab requires extensive and reproducible testing. A framework for high-volume, reproducible and automated system testing of the GlucoTab system was set up applying several Open Source tools for test automation and system time handling. The REACTION insulin titration protocol was investigated in a paper-based clinical trial (PBCT). In order to validate the GlucoTab system, data from this trial was used for simulation and system tests. In total, 1190 decision support action points were identified and simulated. Four data points (0.3%) resulted in a GlucoTab system error caused by a defective implementation. In 144 data points (12.1%), calculation errors of physicians and nurses in the PBCT were detected. The test framework was able to verify manual calculation of insulin doses and detect relatively many user errors and workflow anomalies in the PBCT data. This shows the high potential of the electronic decision support application to improve safety of implementation of an insulin titration protocol and workflow management system in clinical wards.