Strange attractors for the family of orientation preserving Lozi maps.

We extend the result of Michal Misiurewicz assuring the existence of strange attractors for the parametrized family {f(a,b)} of orientation reversing Lozi maps to the orientation preserving case. That is, we rigorously determine an open subset of the parameter space for which an attractor A(a,b) of f(a,b) always exists and exhibits chaotic properties. Moreover, we prove that the attractor is maximal in some open parameter region and arises as the closure of the unstable manifold of a fixed point on which f(a,b)|A(a,b) is mixing. We also show that A(a,b) vary continuously with parameter (a,b) in the Hausdorff metric.

Reference Ranges of Glycemic Variability in Infants after Surgery-A Prospective Cohort Study.

We aimed to define reference ranges of glycemic variability indices derived from continuous glucose monitoring data for non-diabetic infants during post-operative intensive care treatment after cardiac surgery procedures. We performed a prospective cohort intervention study in a pediatric intensive care unit (PICU). Non-diabetic infants aged 0-12 months after corrective cardiovascular surgery procedures were fitted upon arrival to the PICU with a continuous glucose monitoring system (iPro2, Medtronic, Minneapolis, MN, USA). Thirteen glycemic variability indices were calculated for each patient. Complete recordings of 65 patients were collected on the first postoperative day. During the first three postsurgical days 5%, 24% and 43% of patients experienced at least one hypoglycemia episode, and 40%, 10% and 15%-hyperglycemia episode, respectively, in each day. Due to significant differences between the first postoperative day (mean glycemia 130 ± 31 mg/dL) and the second and third day (105 ± 18 mg/dL, 101 ± 22.2 mg/dL; p < 0.0001), we proposed two separate reference ranges-for the acute and steady state patients. Thus, for individual glucose measurements, we proposed a reference range between 85 and 229 mg/dL and 69 and 149 mg/dL. For the mean daily glucose level, ranges between 122 and 137 mg/dL and 95 and 110 mg/dL were proposed. In conclusion, rt-CGM revealed a very high likelihood of hyperglycemia in the first postsurgical day. The widespread use of CGM systems in a pediatric ICU setting should be considered as a safeguard against dysglycemic episodes; however, reference ranges for those patients should be different to those used in diabetes care.

Boosting Intelligent Data Analysis in Smart Sensors by Integrating Knowledge and Machine Learning.

The presented paper proposes a hybrid neural architecture that enables intelligent data analysis efficacy to be boosted in smart sensor devices, which are typically resource-constrained and application-specific. The postulated concept integrates prior knowledge with learning from examples, thus allowing sensor devices to be used for the successful execution of machine learning even when the volume of training data is highly limited, using compact underlying hardware. The proposed architecture comprises two interacting functional modules arranged in a homogeneous, multiple-layer architecture. The first module, referred to as the knowledge sub-network, implements knowledge in the Conjunctive Normal Form through a three-layer structure composed of novel types of learnable units, called L-neurons. In contrast, the second module is a fully-connected conventional three-layer, feed-forward neural network, and it is referred to as a conventional neural sub-network. We show that the proposed hybrid structure successfully combines knowledge and learning, providing high recognition performance even for very limited training datasets, while also benefiting from an abundance of data, as it occurs for purely neural structures. In addition, since the proposed L-neurons can learn (through classical backpropagation), we show that the architecture is also capable of repairing its knowledge.

Interactive Timeline Approach for Contextual Spatio-Temporal ECT Data Investigation.

This paper presents a novel approach to a complex process of electrical capacitance tomography (ECT) measurement data analysis. ECT is frequently employed for non-invasive monitoring of industrial process phenomena. Proposed methodology is based on the premeditated integration of the spatial and temporal relations inherent in the measurement records into the workflow of the analysis procedure. We propose a concept of interactive timeline that enables arranging data visualization according to the user's current focus along the process of analysis. We evaluated the proposed method using a prototype system in a task-based user study conducted with a group of domain experts. The evaluation is based on gravitational silo flow measurement datasets. Proposed prototype system enables diverse data manipulation in a more natural way allowing the user to switch back and forth between space and time domains along the data analysis trail. Experiments with the prototype system showed that the accuracy and completion times have significantly improved in comparison to the performance measured in the baseline condition. Additionally, the participants reported decreased physical load with improved efficiency measured with NASA task load index. Finally, a short discussion coupled with directions for the future of interactive spatio-temporal ECT measurement data analysis conclude the paper.

A systemic approach to screening high-throughput RT-qPCR data for a suitable set of reference circulating miRNAs.

BACKGROUND: The consensus on how to choose a reference gene for serum or plasma miRNA expression qPCR studies has not been reached and none of the potential candidates have yet been convincingly validated. We proposed a new in silico approach of finding a suitable reference for human, circulating miRNAs and identified a new set of endogenous reference miRNA based on miRNA profiling experiments from Gene Expression Omnibus. We used 3 known normalization algorithms (NormFinder, BestKeeper, GeNorm) to calculate a new normalization score. We searched for a universal set of endogenous miRNAs and validated our findings on 2 new datasets using our approach. RESULTS: We discovered and validated a set of 13 miRNAs (miR-222, miR-92a, miR-27a, miR-17, miR-24, miR-320a, miR-25, miR-126, miR-19b, miR-199a-3p, miR-30b, miR-30c, miR-374a) that can be used to create a reliable reference combination of 3 miRNAs. We showed that on average the mean of 3 miRNAs (p = 0.0002) and 2 miRNAs (p = 0.0031) were a better reference than single miRNA. The arithmetic means of 3 miRNAs: miR-24, miR-222 and miR-27a was shown to be the most stable combination of 3 miRNAs in validation sets. CONCLUSIONS: No single miRNA was suitable as a universal reference in serum miRNA qPCR profiling, but it was possible to designate a set of miRNAs, which consistently contributed to most stable combinations.

Resistance to Data Loss of Glycemic Variability Measurements in Long-Term Continuous Glucose Monitoring.

Background: Continuous glucose monitoring (CGM) is a method of estimating blood glucose values from those recorded in the interstitial fluid. Because increasingly longer CGM measurements are possible, errors and data loss become more and more likely and potentially more damaging to accurate calculations of glycemic variability (GV) indices. Our research investigates the resistance of the CGM recording to data loss. Methods: We collected 71 CGM recordings (duration of min: 2, max: 265, median: 42 days) from patients with type 1 diabetes and used three algorithms to introduce missing data. We calculated mean and standard deviation (SD) of absolute percentage error of 12 variability indices and correlated those with the percentage of missing data and duration of the measurements. Results: Mean absolute percentage error of variability indices increased linearly with the percentage of missing data along with SD of absolute percentage error. Except for mean amplitude of glycemic excursions and time spent in hypoglycemia, all absolute errors never exceeded 25%, while mean absolute errors stayed below 5%. The gradient removal algorithm introduced errors larger than the single datapoint and block removal algorithms. The absolute percentage error of variability indices correlated negatively with the duration of the CGM measurements. Conclusions: Standard GV measurements in long-term glucose monitoring are robustly resistant to data loss.