Diabetes self-management plays a vital role in improving clinical outcomes and the quality of life of individuals living with diabetes. Despite considerable research on its impact on clinical outcomes, diabetes self-management continues to be challenging for many individuals living with the condition. As part of the growth in digital health technologies for diabetes care, smartphone applications present potential opportunities to bridge the existing gaps in self-management and improve patient outcomes.

Participants (N = 3241 people with diabetes) were recruited to answer questions about diabetes self-management, including their use of digital tools, their preferences for smartphone applications for diabetes, and the preferred functions of these applications they found useful. Frequency distributions and chi-square analyses were performed to examine the demographic differences among users of diabetes and general wellness applications.

Among participants, 30.2% reported using health applications specifically made for diabetes management, while 33.9% reported using health applications that were not diabetes-specific. Considerable differences in demographic characteristics were found between users and nonusers of both diabetes-specific and general health applications groups. The most preferred applications provided the opportunity to engage with continuous glucose monitoring data (i.e., continuous measurement; 47.4%) followed by glucose monitoring (i.e., single reading measurement; 20.9%), food intake trackers (23.6%), and fitness goal trackers (22.8%).

These findings suggest that the use of digital health technologies is popular for people living with diabetes, but more needs to be done to ensure wider adoption and sustained use.

Continuous biosensors have the potential to transform medicine, enabling healthcare to be more preventative and personalized as compared to the current standard of reactive diagnostics. Realizing this transformative potential requires biosensors that can function continuously in vivo without sample preparation and deliver molecular specificity, sensitivity, and high temporal resolution. Molecular switches stand out as a promising solution for creating such sensors for the continuous detection of many different types of molecules. Molecular switches are target-binding receptors designed such that binding causes a conformational change in the switch's structure. This structure switching induces a measurable signal change via reporters incorporated into the molecular switch, enabling highly specific, label-free sensing. However, there remains an outstanding need for generalizable switch designs that can be adapted for the detection of a wide range of molecular targets. In this Account, we chronicle the work our lab has done to develop generalizable molecular switch designs that allow more rapid development of high-performance biosensors across a broad range of biomarkers. Pioneering efforts toward molecular switch-based biosensing have employed aptamers─nucleic acid-based receptors with sequence-specific target affinity. However, most of these early demonstrations relied upon aptamers with intrinsic structure-switching capabilities. To accelerate aptamer switch design for more targets, we have applied rational design and knowledge of an aptamer's structure to engineer switching functionality into pre-existing aptamers. Our designs contained several structural parameters that enabled us to easily tune the sensitivity and binding kinetics of the resulting switches. Using such rationally designed aptamer switches, we demonstrated continuous optical detection of cortisol and dopamine at physiologically relevant concentrations in complex media. In an effort to move beyond aptamers with well-characterized structural properties, we developed a high-throughput screening method that allowed us to simultaneously screen millions of candidates derived from a single aptamer to find sensitive switches without any prior structural knowledge of the parent aptamer. In subsequent work, we reasoned that we could enhance our ability to design a broader range of biosensors by leveraging other classes of receptors besides aptamers. Antibodies offer excellent affinity and specificity for a wide range of targets, but lack the capacity for intrinsic structure switching. We therefore developed a set of strategies to augment antibodies with the capacity to act as molecular switches with a diverse range of target-binding properties. We combined both the high binding affinity of an antibody with the structure-switching capabilities of an aptamer to develop a chimeric switch with 100-fold enhanced sensitivity for a protein target and improved function in interferent-rich samples. In a second design, we developed a competitive immunoassay-inspired scheme to engineer switching behavior into an antibody for minutes-scale temporal resolution with nanomolar sensitivity. We used this competitive antibody-switch to demonstrate the first continuous detection of cortisol directly in whole blood. Together, these advances in molecular switch development have expanded our capability to rapidly engineer new continuous biosensors, thereby increasing opportunities to track health via a wide range of biomarkers to deliver more personalized and preventative medicine.

We examined associations between daily variations in objectively measured sleep and blood glucose in a sample of non-diabetic young adults to complement laboratory studies on how sleep affects blood glucose levels.

One hundred and nineteen university students underwent sleep measurement using an Oura Ring 2 and continuous glucose monitoring (CGM) for up to 14 days. In 69 individuals who consumed a standardized diet across the study, multilevel models examined associations between sleep duration, timing, efficiency, and daily CGM profiles. Separately, in 58 individuals, multilevel models were used to evaluate postprandial glycaemic responses to a test meal challenge on 7 days. Participants also underwent oral glucose tolerance testing once after a night of ad libitum sleep, and again following a night of sleep restriction by 1-2 hours relative to that individual's habitual sleep duration. Between-condition glucose and insulin excursions, HOMA-IR and Matsuda index were compared.

Nocturnal sleep did not significantly influence following-day CGM profiles, postprandial glucose, or nocturnal mean glucose levels (all ps > .05). Longer sleep durations were associated with lower same-night glucose variability (all ps < .001). However, the range of variation in sugar levels was small and unlikely to be of functional significance. Considering naps in the analysis did not alter the findings. Sleep restriction by an average of 1.73 hours (SD = 0.97) did not significantly impact excursions in glucose or insulin or insulin sensitivity the following morning (all ps > .05).

Glucose handling in young, healthy adults may be more resilient to real-life fluctuations in sleep patterns than previously thought.

Monitoring Sleep and Glucose Among University Students https://clinicaltrials.gov/study/NCT04880629, ID: NCT04880629.

Continuous Glucose Monitoring sensors (CGMs) have revolutionized type 1 diabetes (T1D) management. In particular, in several cases, the retrospective analysis of CGM recordings allows clinicians to review and adjust patients' therapy. However, in this set-up, the artifacts that are often present in CGM data could lead to incorrect therapeutic actions. To mitigate this risk, we investigate how to detect one of the most common of these artifacts, the so-called pressure induced sensor attenuations, by means of anomaly detection algorithms. Specifically, these methods belong to the class of unsupervised techniques, which is particularly appealing since it does not require a labeled dataset, hardly available in practice. After having designed five features to highlight the anomalous state of the sensor, 8 different methods (e.g. Isolation Forest and Histogram-based Outlier Score) are assessed both in silico using the UVa/Padova Type 1 Diabetes Simulator and on real data of 36 subjects monitored for about 10 days. In the in silico scenario, the best results are achieved with Isolation Forest, which recognized the 74% of the failures generating on average only 2 false alerts during the whole monitoring time. In real data, Isolation Forest is confirmed to be effective in the detection of failures, achieving a recall of 55% and generating 3 false alarms in 10 days. By allowing to detect more than 50% of the artifacts while discarding only a few portions of correct data in several days of monitoring, the proposed approach could effectively improve the quality of CGM data used by clinicians to retrospectively evaluate and adjust T1D therapy.

Despite the existence of an increasing array of digital technologies and tools for diabetes management, there are disparities in access to and uptake and use of continuous glucose monitoring (CGM) devices, particularly for those most at risk of poor diabetes outcomes.

This study aims to assess communication technology and CGM access, literacy, and use among patients receiving treatment for diabetes at an inner-city safety-net hospital.

A survey on digital technology ownership and use was self-administered by 75 adults with type 1 and type 2 diabetes at the diabetes clinic of Grady Memorial Hospital in Atlanta, Georgia. In-depth interviews were conducted with 16% (12/75) of these patient participants and 6 health care providers (HCPs) to obtain additional insights into the use of communication technology and CGM to support diabetes self-management.

Most participants were African American (66/75, 88%), over half (39/75, 52%) were unemployed or working part time, and 29% (22/75) had no health insurance coverage, while 61% (46/75) had federal coverage. Smartphone ownership and use were near universal; texting and email use were common (63/75, 84% in both cases). Ownership and use of tablets and computers and use and daily use of various forms of media were more prevalent among younger participants and those with type 1 diabetes, who also rated them as easier to use. Technology use specifically for diabetes and health management was low. Participants were supportive of a potential smartphone app for diabetes management, with a high interest in such an app helping them track blood sugar levels and communicate with their care teams. Younger participants showed higher levels of interest, perceived value, and self-efficacy for using an app with these capabilities. History of CGM use was reported by 56% (42/75) of the participants, although half (20/42, 48%) had discontinued use, above all due to the cost of the device and issues with its adhesive. Nonuse was primarily due to not being offered CGM by their HCP. Reasons given for continued use included convenience, improved blood glucose control, and better tracking of blood glucose. The in-depth interviews (n=18) revealed high levels of satisfaction with CGM by users and supported the survey findings regarding reasons for continued use. They also highlighted the value of CGM data to enhance communication between patients and HCPs.

Smartphone ownership was near universal among patients receiving care at an inner-city hospital. Alongside the need to address barriers to CGM access and continued use, there is an opportunity to leverage increased access to communication technology in combination with CGM to improve diabetes outcomes among underresourced populations.

This work evaluates the accuracy and agreement between the FreeStyle Libre sensor (FSL) and an off-label converted real-time continuous glucose monitor (c-rtCGM) device consisting of the MiaoMiao transmitter and the xDrip+ application which can be coupled to the FSL.

Four weeks of glucose data were collected from 21 participants with type 1 diabetes using the c-rtCGM and FSL: two weeks with a single initial calibration (uncalibrated) and two weeks with a daily calibration (calibrated). Accuracy and agreement evaluation included mean absolute relative difference (MARD), the %20/20 rule, Bland-Altman plots, and the Consensus Error Grid analysis.

Values reported by the c-rtCGM system compared with the FSL resulted in an overall MARD of 12.06% and 84.71% of the results falling within Consensus Error Grid Zone A when the device is calibrated. For uncalibrated devices, an overall MARD of 17.49% was obtained. Decreased accuracy was shown in the hypoglycemic range and for rates of change greater than 2 mg/dL/min. The between-device bias also incremented with increasing glucose values.

Measurements recorded by the c-rtCGM were found to be accurate when compared with FSL data only when performing daily c-rtCGM device calibrations. High drops in accuracy and agreement between devices occurred when the c-rtCGM was not calibrated.

The development of receptors suitable for the continuous detection of analytes in complex, interferent-rich samples remains challenging. Antibodies are highly sensitive but difficult to engineer in order to introduce signaling functionality, while aptamer switches are easy to construct but often yield only a modest target sensitivity. We present here a programmable antibody and DNA aptamer switch (PANDAS), which combines the desirable properties of both receptors by using a nucleic acid tether to link an analyte-specific antibody to an internal strand-displacement (ISD)-based aptamer switch that recognizes the same target through different epitopes. The antibody increases PANDAS analyte binding due to its high affinity, and the effective concentration between the two receptors further enhances two-epitope binding and fluorescent aptamer signaling. We developed a PANDAS sensor for the clotting protein thrombin and show that a tuned design achieves a greater than 300-fold enhanced sensitivity compared to that of using an aptamer alone. This design also exhibits reversible binding, enabling repeated measurements with a temporal resolution of ∼10 min, and retains excellent sensitivity even in interferent-rich samples. With future development, this PANDAS approach could enable the adaptation of existing protein-binding aptamers with modest affinity to sensors that deliver excellent sensitivity and minute-scale resolution in minimally prepared biological specimens.

Throughout the COVID-19 pandemic, there has been a notable increase in the incidence of new-onset diabetes and diabetic ketoacidosis (DKA). Simultaneously, children diagnosed with type 1 diabetes (T1D) have encountered difficulties in maintaining optimal blood glucose levels. The mechanisms underpinning these correlations still remain a puzzle. We reviewed the studies that examined changes in incidence during the pandemic. These studies utilized various metrics for comparison, which encompassed the timing of data collection, diagnostic criteria, as well as the numbers and incidence rates of diabetes and DKA. We found the incidence of diabetes and DKA was higher during the pandemic. As to mechanisms, the invivo and invitro study revealed the factors such as direct viral damage, metabolic dysfunction, and immune responses all attribute to the process of T1D after suffering from COVID-19. Furthermore, we provide some useful strategies to prevent and treat children suffering from diabetes and COVID-19.

Strong correlations have been observed between new-onset diabetes and COVID-19. Insights gleaned from clinical descriptions and basic research can offer valuable experience and recommendations for the treatment and prevention of diabetes during future pandemics.

Accurate blood glucose (BG) prediction are key in next-generation tools for type 1 diabetes (T1D) management, such as improved decision support systems and advanced closed-loop control. Glucose prediction algorithms commonly rely on black-box models. Large physiological models, successfully adopted for simulation, were little explored for glucose prediction, mostly because their parameters are hard to individualize. In this work, we develop a BG prediction algorithm based on a personalized physiological model inspired by the UVA/Padova T1D Simulator. Then we compare white-box and advanced black-box personalized prediction techniques.

A personalized nonlinear physiological model is identified from patient data through a Bayesian approach based on Markov Chain Monte Carlo technique. The individualized model was integrated within a particle filter (PF) to predict future BG concentrations. The black-box methodologies considered are non-parametric models estimated via gaussian regression (NP), three deep learning methods: long-short-term-memory (LSTM), gated recurrent unit (GRU), temporal convolutional networks (TCN), and a recursive autoregressive with exogenous input model (rARX). BG forecasting performances are assessed for several prediction horizons (PH) on 12 individuals with T1D, monitored in free-living conditions under open-loop therapy for 10 weeks.

NP models provide the most effective BG predictions by achieving a root mean square error (RMSE), RMSE = 18.99 mg/dL, RMSE = 25.72 mg/dL and RMSE = 31.60 mg/dL, significantly outperforming: LSTM, GRU (for PH = 30 minutes), TCN, rARX, and the proposed physiological model for PH=30, 45 and 60 minutes.

Black-box strategies remain preferable for glucose prediction even when compared to a white-box model with sound physiological structure and individualized parameters.

Aptamers have been spotlighted as promising bio-recognition elements because they can be tailored to specific target molecules, bind to targets with a high affinity and specificity, and are easy to chemically synthesize and introduce functional groups to. In particular, fluorescent aptasensors are widely used in biological applications to diagnose diseases as well as prevent diseases by detecting cancer cells, viruses, and various biomarkers including nucleic acids and proteins as well as biotoxins and bacteria from food because they have the advantages of a high sensitivity, selectivity, rapidity, a simple detection process, and a low price. We introduce screening methods for isolating aptamers with q high specificity and summarize the sequences and affinities of the aptamers in a table. This review focuses on aptamer-based fluorescence detection sensors for biological applications, from fluorescent probes to mechanisms of action and signal amplification strategies.

Diabetes mellitus is the most challenging and fastest-growing global public health concern. Approximately 10.5% of the global adult population is affected by diabetes, and almost half of them are undiagnosed. The growing at-risk population exacerbates the shortage of health resources, with an estimated 10.6% and 6.2% of adults worldwide having impaired glucose tolerance and impaired fasting glycemia, respectively. All current diabetes screening methods are invasive and opportunistic and must be conducted in a hospital or laboratory by trained professionals. At-risk participants might remain undetected for years and miss the precious time window for early intervention to prevent or delay the onset of diabetes and its complications.

We aimed to develop an artificial intelligence solution to recognize elevated blood glucose levels (≥7.8 mmol/L) noninvasively and evaluate diabetic risk based on repeated measurements.

This study was conducted at KK Women's and Children's Hospital in Singapore, and 500 participants were recruited (mean age 38.73, SD 10.61 years; mean BMI 24.4, SD 5.1 kg/m2). The blood glucose levels for most participants were measured before and after consuming 75 g of sugary drinks using both a conventional glucometer (Accu-Chek Performa) and a wrist-worn wearable. The results obtained from the glucometer were used as ground-truth measurements. We performed extensive feature engineering on photoplethysmography (PPG) sensor data and identified features that were sensitive to glucose changes. These selected features were further analyzed using an explainable artificial intelligence approach to understand their contribution to our predictions.

Multiple machine learning models were trained and assessed with 10-fold cross-validation, using participant demographic data and critical features extracted from PPG measurements as predictors. A support vector machine with a radial basis function kernel had the best detection performance, with an average accuracy of 84.7%, a sensitivity of 81.05%, a specificity of 88.3%, a precision of 87.51%, a geometric mean of 84.54%, and F score of 84.03%.

Our findings suggest that PPG measurements can be used to identify participants with elevated blood glucose measurements and assist in the screening of participants for diabetes risk.

Among guidelines on gestational diabetes mellitus, there is an incongruity about the threshold of maternal hyperglycemia to diagnose gestational diabetes mellitus.

This study aimed to ascertain the association between continuous glucose monitoring metrics and adverse outcomes among individuals undergoing gestational diabetes mellitus screening.

This was a prospective study (from June 2020 to January 2022) of individuals who underwent 2-step gestational diabetes mellitus screening at ≤30 weeks of gestation. The participants wore a blinded continuous glucose monitoring device (Dexcom G6 Pro; Dexcom, Inc, San Diego, CA) for 10 days starting when they took the 50-g glucose challenge test. The primary outcome was a composite of adverse neonatal outcomes (large for gestational age, shoulder dystocia or neonatal injury, respiratory distress, need for intravenous glucose treatment for hypoglycemia, or fetal or neonatal death). The secondary neonatal outcomes included preterm birth, neonatal intensive care unit admission, hypoglycemia, mechanical ventilation or continuous positive airway pressure, hyperbilirubinemia, and hospital length of stay. The secondary maternal outcomes included weight gain during pregnancy, hypertensive disorders of pregnancy, induction of labor, cesarean delivery, and postpartum complications. Time within the target range (63-140 mg/dL), time above the target range (>140 mg/dL) expressed as a percentage of all continuous glucose monitoring readings, and mean glucose level were analyzed. The Youden index was used to choose the threshold of ≥10% for the time above the target range and association with adverse outcomes.

Of 136 participants recruited, data were available from 92 individuals (67.6%). The 2-step method diagnosed gestational diabetes mellitus in 2 individuals (2.2%). Continuous glucose monitoring indicated that 17 individuals (18.5%) had time above the target range of ≥10%. Individuals with time above the target range of ≥10% had a significantly higher likelihood of composite adverse neonatal outcomes than individuals with time above the target range of <10% (63% vs 18%; P=.001). Furthermore, compared with neonates born to individuals with time above the target range of <10%, neonates born to individuals with time above the target range of ≥10% had an increased likelihood for hypoglycemia (14.5% vs 47%; P=.009) and had a longer length of stay (2 vs 4 days; P=.03). No difference in maternal outcomes was noted between the groups.

In this prospective study of individuals undergoing gestational diabetes mellitus screening, a cutoff of the time above the target range of ≥10% using continuous glucose monitoring was associated with a higher rate of neonatal adverse outcomes. A randomized trial of continuous glucose monitoring vs 2-step screening for gestational diabetes mellitus to lower the rate of adverse outcomes is underway (identification number: NCT05430204).

Continuous glucose monitoring (CGM) sensors measure interstitial glucose concentration every 1-5 min for days or weeks. New CGM-based diabetes therapies are often tested in in silico clinical trials (ISCTs) using diabetes simulators. Accurate models of CGM sensor inaccuracies and failures could help improve the realism of ISCTs. However, the modeling of CGM failures has not yet been fully addressed in the literature. This work aims to develop a mathematical model of CGM gaps, i.e., occasional portions of missing data generated by temporary sensor errors (e.g., excessive noise or artifacts).

Two datasets containing CGM traces collected in 167 adults and 205 children, respectively, using the Dexcom G6 sensor (Dexcom Inc., San Diego, CA) were used. Four Markov models, of increasing complexity, were designed to describe three main characteristics: number of gaps for each sensor, gap distribution in the monitoring days, and gap duration. Each model was identified on a portion of each dataset (training set). The remaining portion of each dataset (real test set) was used to evaluate model performance through a Monte Carlo simulation approach. Each model was used to generate 100 simulated test sets with the same size as the real test set. The distributions of gap characteristics on the simulated test sets were compared with those observed on the real test set, using the two-sample Kolmogorov-Smirnov test and the Jensen-Shannon divergence.

A six-state Markov model, having two states to describe normal sensor operation and four states to describe gap occurrence, achieved the best results. For this model, the Kolmogorov-Smirnov test found no significant differences between the distribution of simulated and real gap characteristics. Moreover, this model obtained significantly lower Jensen-Shannon divergence values than the other models.

A Markov model describing CGM gaps was developed and validated on two real datasets. The model describes well the number of gaps for each sensor, the gap distribution over monitoring days, and the gap durations. Such a model can be integrated into existing diabetes simulators to realistically simulate CGM gaps in ISCTs and thus enable the development of more effective and robust diabetes management strategies.

We present a generalizable approach for designing biosensors that can continuously detect small-molecule biomarkers in real time and without sample preparation. This is achieved by converting existing antibodies into target-responsive "antibody-switches" that enable continuous optical biosensing. To engineer these switches, antibodies are linked to a molecular competitor through a DNA scaffold, such that competitive target binding induces scaffold switching and fluorescent signaling of changing target concentrations. As a demonstration, we designed antibody-switches that achieve rapid, sample preparation-free sensing of digoxigenin and cortisol in undiluted plasma. We showed that, by substituting the molecular competitor, we can further modulate the sensitivity of our cortisol switch to achieve detection at concentrations spanning 3.3 nanomolar to 3.3 millimolar. Last, we integrated this switch with a fiber optic sensor to achieve continuous sensing of cortisol in a buffer and blood with <5-min time resolution. We believe that this modular sensor design can enable continuous biosensor development for many biomarkers.

The coronavirus disease (COVID-19) pandemic affected the management and follow-up of several chronic ailments, including pediatric type 1 diabetes mellitus (T1DM). Restricted access to healthcare and fear of contracting the virus during medical facility visits resulted in poor compliance, irregular follow-up visits, treatment, and delayed diagnosis of complications in pediatric diabetes such as diabetic ketoacidosis (DKA). As such, the incidence of complicated DKA in resource-limited settings is high due to delayed presentation, poor compliance with therapy, and associated comorbidities such as malnutrition and sepsis. The pandemic had only added to the woes. The increased surge in DKA, in the face of limited resources, prompted clinicians to find alternative solutions to manage these children effectively. In this narrative review, we discuss the key challenges faced globally while caring for children with T1DM and DKA during the COVID-19 pandemic, and the lessons learned thereof.

Advanced decision support systems for type 1 diabetes (T1D) management often embed prediction modules, which allow T1D people to take preventive actions to avoid critical episodes like hypoglycemia. Real-time prediction of blood glucose (BG) concentration relies on a subject-specific model of glucose-insulin dynamics. Model parameter identification is usually based on the mean square error (MSE) cost function, and the model is usually used to predict BG at a single prediction horizon (PH). Finally, a hypo-alarm is raised if the predicted BG crosses a threshold. This work aims to show that real-time hypoglycemia forecasting can be improved by leveraging: a glucose-specific mean square error (gMSE) cost function in model's parameters identification, and a "prediction-funnel," that is, confidence intervals (CIs) for multiple PHs, within the hypo-alarm-raising strategy.

Autoregressive integrated moving average with exogenous input (ARIMAX) models are selected to illustrate the proposed solution (use of gMSE and prediction-funnel) and its assessment against the conventional approach (MSE and single PH). The gMSE penalizes the model misfit in unsafe BG ranges (e.g., hypoglycemia), and the prediction-funnel allows raising an alarm by monitoring if the CIs cross a suitable threshold. The algorithms were evaluated by measuring precision (P), recall (R), F1-score (F1), false positive per day (FP/day), and time gain (TG) on a real dataset collected in 11 T1D individuals.

The best performance is achieved exploiting both the gMSE and the prediction-funnel: P = 65%, R = 88%, F1 = 75%, FP/day = 0.29, and mean TG = 15 minutes.

The combined use of a glucose-specific metric and an alarm-raising strategy based on the prediction-funnel allows achieving a more effective and reliable hypoglycemia prediction algorithm.

The high need for optimal diabetes management among an ever-increasing number of patients dictates the development and implementation of new digital sensors for continuous glucose monitoring. The purpose of this work is to systematize the global patenting trends of digital sensors for continuous glucose monitoring and analyze their effectiveness in controlling the treatment of diabetes patients of different ages and risk groups. The Lens database was used to build the patent landscape of sensors for continuous glucose monitoring. Retrospective analysis showed that the patenting of sensors for continuous glucose monitoring had positive trend over the analyzed period (2000-2022). Leading development companies are Dexcom Inc., Abbott Diabetes Care Inc., Medtronic Minimed Inc., Roche Diabetes Care Inc., Roche Diagnostics Operations Inc., Roche Diabetes Care Gmbh, and Ascensia Diabetes Care Holdings Ag, among others. Since 2006, a new approach has emerged where digital sensors are used for continuous glucose monitoring, and smartphones act as receivers for the data. Additionally, telemedicine communication is employed to facilitate this process. This opens up new opportunities for assessing the glycemic profile (glycemic curve information, quantitative assessment of the duration and amplitude of glucose fluctuations, and so on), which may contribute to improved diabetes management. A number of digital sensors for minimally invasive glucose monitoring are patented, have received FDA approval, and have been on the market for over 10 years. Their effectiveness in the clinic has been proven, and advantages and disadvantages have been clarified. Digital sensors offer a non-invasive option for monitoring blood glucose levels, providing an alternative to traditional invasive methods. This is particularly useful for patients with diabetes who require frequent monitoring, including before and after meals, during and after exercise, and in other scenarios where glucose levels can fluctuate. However, non-invasive glucose measurements can also benefit patients without diabetes, such as those following a dietary treatment plan, pregnant women, and individuals during fasting periods like Ramadan. The availability of non-invasive monitoring is especially valuable for patients in high-risk groups and across different age ranges. New world trends have been identified in the patenting of digital sensors for non-invasive glucose monitoring in interstitial skin fluid, saliva, sweat, tear fluid, and exhaled air. A number of non-invasive devices have received the CE mark approval, which confirms that the items meet European health, safety, and environmental protection standards (TensorTip Combo-Glucometer, Cnoga Medical Ltd.; SugarBEAT, Nemaura Medical; GlucoTrack, GlucoTrack Inc.), but are not FDA-approved yet. The above-mentioned sensors have characteristics that make them popular in the treatment of diabetes: they do not require implantation, do not cause an organism reaction to a foreign body, and are convenient to use. In the EU, in order to increase clinical safety and the level of transparency about medical devices, manufacturers must obtain certificates in accordance with Regulation (EU) 2017/745, taking into account the transition period. The development of systems, which include digital sensors for continuous glucose monitoring, mobile applications, and web platforms for professional analysis of glycemic control and implementation of unified glycemic assessment principles in mobile healthcare, represent promising approaches for controlling glycaemia in patients.

Advances in diabetes technology, especially in the last few decades, have transformed our ability to deliver care to persons with diabetes (PWD). Developments in glucose monitoring, especially continuous glucose monitoring systems (CGM), have revolutionized diabetes care and empowered our patients to manage their disease. CGM has also played an integral role in advancing automated insulin delivery systems. Currently available and upcoming advanced hybrid-closed loop systems aim to decrease patient involvement and are approaching the functionality of a fully automated artificial pancreas. Other advances, such as smart insulin pens and daily patch pumps, offer more options for patients and require less complicated and costly technology. Evidence to support the role of diabetes technology is growing, and PWD and clinicians must choose the right type of technology with a personalized strategy to manage diabetes effectively. Here, we review currently available diabetes technologies, summarize their individual features and highlight key patient factors to consider when creating a personalized treatment plan. We also address current challenges and barriers to the adoption of diabetes technologies.

This study aims to reveal the use and management status of continuous glucose monitoring (CGM) in tertiary hospitals in China and to determine the potential factors affecting the application of CGM, based on which more effective solutions would be produced and implemented.

An online, cross-sectional study was conducted from October 2021 to December 2021.

Eighty-three tertiary hospitals in China were involved.

Eighty-three head nurses and 281 clinical nurses were obtained.

Current condition of CGM use and management, the factors that hinder the use and management of CGM, scores of current CGM use and management, as well as their influencing factors, were collected.

Among the 83 hospitals surveyed, 57 (68.7%) hospitals used CGM for no more than 10 patients per month. Seventy-three (88.0%) hospitals had developed CGM standard operating procedures, but only 29 (34.9%) hospitals devised emergency plans to deal with adverse effects related to CGM. Comparably, maternal and children's hospitals were more likely to have a dedicated person to assign install CGM than general hospitals (52.2% vs 26.7%). As for the potential causes that hinder the use and management of CGM, head nurses' and nurses' perceptions differed. Head nurses perceived patients' limited knowledge about CGM (60.2%), the high costs of CGM and inaccessibility to medical insurance (59.0%), and imperfect CGM management systems (44.6%) as the top three factors. Different from head nurses, CGM operation nurses considered the age of CGM operators, the type of hospital nurses worked in, the number of patients using CGM per month and the number of CGM training sessions as potential factors (p<0.05).

The study provides a broad view of the development status of CGM in China. Generally speaking, the use and management of CGM in China are not yet satisfactory, and more efforts are wanted for improvement.

Diabetes can lead to multiple complications that can reduce the quality of life, impose additional costs on the healthcare systems and ultimately lead to premature death. Proper self-care in diabetic patients can impede or delay the onset of diabetes complications. This study aimed to investigate diabetes complications and their association with diet knowledge, skills, and self-care barriers.

This was a cross-sectional study. A total of 1139 patients with Type 2 Diabetes Mellitus (T2DM) referring to health centers in Tabriz, Iran, were included from January to July 2019. Data were collected using two questionnaires: (1) a sociodemographic questionnaire and (2) a Personal Diabetes Questionnaire (PDQ). Data were analyzed using SPSS software version 22. χ 2 test was used to examine the association between the socioeconomic and disease-related variables and the prevalence of diabetes complications. T-test was used to examine the association between diet knowledge and skills, self-care barriers, and the incidence of diabetes complications.

In this study, 76.1% of patients had at least one complication, and 30.2% had a history of hospitalization due to diabetes complications during the past year. Approximately 49% and 43% were diagnosed with high blood pressure and hyperlipidemia, respectively. Cardiovascular disease was the most common diabetes complication (15.9%) and the cause of hospitalization (11.01%) in patients with diabetes. Barriers to diet adherence, blood glucose monitoring, and exercise were significantly associated with self-reported diabetes complications (p < 0.001). Our results showed no significant association between the number of complications and diet knowledge and skills (p = 0.44).

This study indicated that the prevalence of diabetes complications was higher among patients with more barriers to self-care. In light of these findings, taking appropriate measures to reduce barriers to self-care can prevent or delay the onset of diabetes complications.

Continuous Glucose Monitoring (CGM) sensors micro-invasively provide frequent glucose readings, improving the management of Type 1 diabetic patients' life and making available reach data-sets for retrospective analysis. Unlikely, CGM sensors are subject to failures, such as compression artifacts, that might impact on both real-time and respective CGM use. In this work is focused on retrospective detection of compression artifacts. An in-silico dataset is generated using the T1D UVa/Padova simulator and compression artifacts are subsequently added in known position, thus creating a dataset with perfectly accurate faulty/not-faulty labels. The problem of compression artifact detection is then faced with supervised data-driven techniques, in particular using Random Forest algorithm. The detection performance guaranteed by the method on in-silico data is satisfactory, opening the way for further analysis on real-data.

With the first commercially available smart insulin pens, the predominant insulin delivery device for millions of people living with diabetes is now coming into the digital age. Smart insulin pens (SIPs) have the potential to reshape a connected diabetes care ecosystem for patients, providers, and health systems. Existing SIPs are enhanced with real-time wireless connectivity, digital dose capture, and integration with personalized dosing decision support. Automatic dose capture can promote effective retrospective review of insulin dose data, particularly when paired with glucose data. Patients, providers, and diabetes care teams will be able to make increasingly data-driven decisions and recommendations, in real time, during scheduled visits, and in a more continuous, asynchronous care model. As SIPs continue to progress along the path of digital transformation, we can expect additional benefits: iteratively improving software, machine learning, and advanced decision support. Both these technological advances, and future care delivery models with asynchronous interactions, will depend on easy, open, and continuous data exchange between the growing number of diabetes devices. SIPs have a key role in modernizing diabetes care for a large population of people living with diabetes.

Background. Continuous glucose monitoring (CGM) offers an opportunity for patients with diabetes to modify their lifestyle to better manage their condition and for clinicians to provide personalized healthcare and lifestyle advice. However, analytic tools are needed to standardize and analyze the rich data that emerge from CGM devices. This would allow glucotypes of patients to be identified to aid clinical decision-making.Methods. In this paper, we develop an analysis pipeline for CGM data and apply it to 148 diabetic patients with a total of 8632 days of follow up. The pipeline projects CGM data to a lower-dimensional space of features representing centrality, spread, size, and duration of glycemic excursions and the circadian cycle. We then use principal components analysis and k -means to cluster patients' records into one of four glucotypes and analyze cluster membership using multinomial logistic regression.Results. Glucotypes differ in the degree of control, amount of time spent in range, and on the presence and timing of hyper- and hypoglycemia. Patients on the program had statistically significant improvements in their glucose levels.Conclusions. This pipeline provides a fast automatic function to label raw CGM data without manual input.

Abrupt implementation of lockdowns during the coronavirus disease 2019 (COVID-19) pandemic affected the management of diabetes mellitus in patients worldwide. Limited access to health facilities and lifestyle changes potentially affected metabolic parameters in patients at risk. We conducted a meta-analysis to determine any differences in the control of metabolic parameters in patients with diabetes, before and during lockdown.

We performed searches of five databases. Meta-analyses were carried out using random- or fixed-effect approaches to glycaemic control parameters as the primary outcome: glycosylated hemoglobin (HbA1c), random blood glucose (RBG), fasting blood glucose (FBG), time-in-range (TIR), time-above-range (TAR), time-below-range (TBR). Mean difference (MD), confidence interval (CI), and P value were calculated. Lipid profile was a secondary outcome and is presented as a descriptive analysis.

Twenty-one studies enrolling a total of 3,992 patients with type 1 or type 2 diabetes mellitus (T1DM or T2DM) were included in the study. Patients with T1DM showed a significant improvement of TIR and TAR (MD=3.52% [95% CI, 0.29 to 6.74], I2=76%, P=0.03; MD=-3.36% [95% CI, -6.48 to -0.25], I2=75%, P=0.03), while FBG among patients with T2DM significantly worsened (MD=3.47 mg/dL [95% CI, 1.22 to 5.73], I2=0%, P<0.01). No significant difference was found in HbA1c, RBG, and TBR. Use of continuous glucose monitoring in T1DM facilitated good glycaemic control. Significant deterioration of lipid parameters during lockdown, particularly triglyceride, was observed.

Implementation of lockdowns during the COVID-19 pandemic did not worsen glycaemic control in patients with diabetes. Other metabolic parameters improved during lockdown, though lipid parameters, particularly triglyceride, worsened.

A soft and flexible wearable sweat epidermal microfluidic device capable of simultaneously stimulating, collecting, and electrochemically analyzing sweat is demonstrated. The device represents the first system integrating an iontophoretic pilocarpine delivery system around the inlet channels of epidermal polydimethylsiloxane (PDMS) microfluidic device for sweat collection and analysis. The freshly generated sweat is naturally pumped into the fluidic inlet without the need of exercising. Soft skin-mounted systems, incorporating non-invasive, on-demand sweat sampling/analysis interfaces for tracking target biomarkers, are in urgent need. Existing skin conformal microfluidic-based sensors for continuous monitoring of target sweat biomarkers rely on assays during intense physical exercising. This work demonstrates the first example of combining sweat stimulation, through transdermal pilocarpine delivery, with sample collection through a microfluidic channel for real-time electrochemical monitoring of sweat glucose, in a fully integrated soft and flexible multiplexed device which eliminates the need of exercising. The on-body operational performance and layout of the device were optimized considering the fluid dynamics and evaluated for detecting sweat glucose in several volunteers. Furthermore, the microfluidic monitoring device was integrated with a real-time wireless data transmission system using a flexible electronic board PCB conformal with the body. The new microfluidic platform paves the way to real-time non-invasive monitoring of biomarkers in stimulated sweat samples for diverse healthcare and wellness applications.

In this study by Alva et al, accuracy of a second-generation factory calibrated continuous glucose monitoring system is evaluated. Compared to the first-generation FreeStyle Libre 14-day system (FSL), accuracy was improved throughout the 14-day wear period, including improved accuracy in hypoglycemia for adults and youth. The addition of optional real-time alerts for hypoglycemia and hyperglycemia as well as an integrated continuous glucose monitor (iCGM) designation by the FDA may further enable users to benefit from using CGM in real time, including in future automated insulin delivery systems. As CGM accuracy, affordability, and accessibility improve, we anticipate increased uptake of CGM by people on intensive insulin therapy, and also potential benefits and expansion into a broader patient population. There are growing opportunities to leverage cloud-connected CGM devices in the increasingly virtual, continuous telehealth-driven diabetes care model, which will require more focus on development and use of data interoperability standards.

People with type 1 diabetes (T1D) need exogenous insulin administrations several times a day. The amount of injected insulin is key for maintaining the concentration of blood glucose (BG) within a physiological safe range. According to well-established clinical guidelines, insulin dosing at mealtime is calculated through an empirical formula which, however, does not take advantage of the knowledge of BG trend provided in real-time by continuous glucose monitoring (CGM) sensors. To overcome suboptimal insulin dosage, we recently used machine learning techniques to build two new models, one linear and one nonlinear, which incorporate BG trend information.In this work, we propose an ensemble learning method for mealtime insulin bolus estimation based on dynamic voting, which combines the two models by taking advantage of where each alternative performs better. Being the resulting model black-box, a tool that enables its interpretability was applied to evaluate the contribution of each feature. The proposed model was trained using a synthetic dataset having information on 100 virtual subjects with different mealtime conditions, and its performance was evaluated within a simulated environment.The benefit given by the ensemble method compared to the single models was confirmed by the high time within the target glycemic range, and the trade-off reached in terms of time spent below and above this range. Moreover, the model interpretation pointed out the key role played by the information on BG dynamics in the estimation of insulin dosage.

Type-1 diabetes (T1D) is a metabolic disease, characterized by impaired blood glucose (BG) regulation, which forces patients to multiple daily therapeutic actions, the most critical of which is exogenous insulin administration. T1D management can considerably benefit of mathematical models enabling accurate BG predictions and effective/safe automated insulin delivery. In building these models, dealing with large inter- and intra-patient variability in glucose-insulin dynamics represents a major challenge. The aim of the present work is to assess linear black-box methods, including a novel non-parametric methodology, for learning individualized models of glucose response to insulin and meal, suitable for model-based prediction and control.

We focus on data-driven techniques for linear model-learning and compare the state-of-art parametric pipeline, exploring all its degrees of freedom (including population vs. individualized parameter identification, model class chosen among ARX/ARMAX/ARIMAX/Box-Jenkins, model order selection criteria, etc.), with a novel non-parametric approach based on Gaussian regression and stable spline kernel. By using data collected in 11 T1D individuals, we evaluate effectiveness of the different models by measuring root mean squared error (RMSE), coefficient of determination (COD), and time gain of the associated BG predictors.

Among the tested approaches, the non-parametric technique results in the best prediction performance: median RMSE=29.8mg/dL, and median COD=57.4%, for a prediction horizon (PH) of 60 min. With respect to the state-of-the-art parametric techniques, the non-parametric approach grants a COD improvement of about 2%, 7%, 21%, and 41% for PH = 30, 60, 90, and 120 min (paired-sample t-test p 0.001, p=0.003, p=0.03, and p=0.07 respectively).

Non-parametric linear model-learning grants statistically significant improvement with respect to the state-of-art parametric approach. The higher PH, the more pronounced the improvement.

The use of a linear non-parametric model-learning approach for model-based prediction and control could bring to a more prompt, safe and effective T1D management.

The concept of implantable glucose sensors has been promulgated for more than 40 years. It is now accepted that continuous glucose monitoring (CGM) increases quality of life by allowing informed diabetes management decisions as a result of more optimized glucose control. The focus of this article is to provide a brief overview of the CGM market history, emerging technologies, and the foreseeable challenges for the next CGM generations as well as proposing possible solutions in an effort to advance the next generation of implantable sensor.

A 54-year-old male with liver cirrhosis (Child-Pugh score 5) presented with severe hepatogenous diabetes (HbA1c 12.6%). Contrast-enhanced CT showed a large portosystemic shunt from the inferior mesenteric vein to the left internal iliac vein. Glucose monitoring showed postprandial hyperglycemia and reactive hypoglycemia. After balloon-occluded retrograde transvenous obliteration (BRTO) and partial splenic transarterial embolization, postprandial hyperglycemia was diminished. Seven months later, HbA1c had improved from 12.6% to 6.7%. In this case, postprandial hyperglycemia occurred by direct delivery of glucose into the systemic circulation via the shunt, and fasting hypoglycemia occurred during treatment with oral antidiabetic agents and insufficient gluconeogenesis. BRTO of the portosystemic shunt resulted in improvement in hepatogenous diabetes.

Due to their widespread use among the population and their wide range of functions and sensors, smartphones are suitable for data collection for medical purposes. App-supported input masks, patient diaries, and patient information systems, mobile access to the patient file as well as telemedical services will continue to find their way into our field of expertise in the future. In addition, the use of smartphone sensors (GPS and motion sensors, touch display, microphone) and coupling possibilities with biosensors (for example with Continuous Glucose Monitoring [CGM] systems), advanced camera technology, the possibility of regular and appointment independent checking of the visual system (visual acuity/contrast vision) as well as real-time data transfer offer interesting possibilities for patient treatment and clinical research. The present review deals with the current status and future perspectives of smartphone-based data collection and possible applications in ophthalmology.

Type 1 diabetes (T1D) therapy requires multiple daily insulin injections to compensate the lack of endogenous insulin production due to β-cells destruction. An empirical standard formula (SF) is commonly used for such a task. Unfortunately, SF does not include information on glucose dynamics, e.g. the glucose rate-of-change (ROC) provided by continuous glucose monitoring (CGM) sensor. Hence, SF can sometimes lead to under/overestimations that can cause critical hypo/hyperglycemic episodes during/after the meal. Recently, to overcome this limitation, we proposed new linear regression models, integrating ROC information and personalized features. Despite the first encouraging results, the nonlinear nature of the problem calls for the application of nonlinear models. In this work, random forest (RF) and gradient boosting tree (GBT), nonlinear machine learning methodologies, were investigated. A dataset of 100 virtual subjects, opportunely divided into training and testing sets, was used. For each individual, a single-meal scenario with different meal conditions (preprandial ROC, BG and meal amounts) was simulated. The assessment was performed both in terms of accuracy in estimating the optimal bolus and glycemic control. Results were compared to the best performing linear model previously developed. The two tree-based models proposed lead to a statistically significant improvement of glycemic control compared to the linear approach, reducing the time spent in hypoglycemia (from 32.49% to 27.57-25.20% for RF and GBT, respectively). These results represent a preliminary step to prove that nonlinear machine learning techniques can improve the estimation of insulin bolus in T1D therapy. Particularly, RF and GBT were shown to outperform the previously linear models proposed.Clinical Relevance- Insulin bolus estimation with nonlinear machine learning techniques reduces the risk of adverse events in T1D therapy.

The aim of this systematic review was to evaluate the clinical implications of continuous glucose monitoring (CGM) among patients with diabetes mellitus using variables that include glycated hemoglobin (HbA1c), estimated A1c, glucose variability, and users' perspectives.

This study analyzed 17 articles that were identified and studied according to the research question criteria. PRISMA guidelines were used for identification and screening of the literature. The required data were searched using Medscape, PubMed, PROSPERO, Wiley Library, Scopus, Clinical Trial Registry, and Trip.

The articles reviewed were on the use of CGM in type 1 and type 2 diabetes mellitus, which showed significant improvement in the levels of HbA1c as compared to non-CGM. The application of CGM on acute sudden onset type of adverse drug reactions (i.e., hypoglycemia) is better than fasting blood sugar or self-monitoring of blood glucose or capillary blood glucose (random blood glucose monitoring). CGM is beneficial for use in patients with type 2 diabetes mellitus including elderly patients as it gives information regarding glucose variability as well as HbA1c levels. The health-care providers require full spectrum of patients' CGM data to design a better therapeutic plan. However, the patients experienced inconvenience on wearing the device on the body for longer periods. The findings also stated the fact that more education and training is required for the patients to interpret their own glycemic data using CGM and modify their lifestyle accordingly. Use of CGM along with HbA1c has also been used to achieve better glycemic results and it allows the health care professional to guide patients in terms of their glucose level; whether they are hypoglycemic or hyperglycemic, however its use has some controversies that minimize its application.

The study concluded that CGM has significant potential in the management of not only patients with type 1 diabetes mellitus but also patients with type 2 diabetes mellitus in spite of the few limitations that are being improvised in the upcoming years. However, limited literature of CGM among patients with type 2 diabetes mellitus and pregnant women reduces the practice scope.

Wearable continuous glucose monitoring (CGM) sensors are revolutionizing the treatment of type 1 diabetes (T1D). These sensors provide in real-time, every 1-5 min, the current blood glucose concentration and its rate-of-change, two key pieces of information for improving the determination of exogenous insulin administration and the prediction of forthcoming adverse events, such as hypo-/hyper-glycemia. The current research in diabetes technology is putting considerable effort into developing decision support systems for patient use, which automatically analyze the patient's data collected by CGM sensors and other portable devices, as well as providing personalized recommendations about therapy adjustments to patients. Due to the large amount of data collected by patients with T1D and their variety, artificial intelligence (AI) techniques are increasingly being adopted in these decision support systems. In this paper, we review the state-of-the-art methodologies using AI and CGM sensors for decision support in advanced T1D management, including techniques for personalized insulin bolus calculation, adaptive tuning of bolus calculator parameters and glucose prediction.

Many predictive models for incident type 2 diabetes (T2D) exist, but these models are not used frequently for public health management. Barriers to their application include (1) the problem of model choice (some models are applicable only to certain ethnic groups), (2) missing input variables, and (3) the lack of calibration. While (1) and (2) drives to missing predictions, (3) causes inaccurate incidence predictions. In this paper, a combined T2D risk model for public health management that addresses these three issues is developed.

The combined T2D risk model combines eight existing predictive models by weighted average to overcome the problem of missing incidence predictions. Moreover, the combined model implements a simple recalibration strategy in which the risk scores are rescaled based on the T2D incidence in the target population. The performance of the combined model was compared with that of the eight existing models using data from two test datasets extracted from the Multi-Ethnic Study of Atherosclerosis (MESA; n=1031) and the English Longitudinal Study of Ageing (ELSA; n=4820). Metrics of discrimination, calibration, and missing incidence predictions were used for the assessment.

The combined T2D model performed well in terms of both discrimination (concordance index: 0.83 on MESA; 0.77 on ELSA) and calibration (expected to observed event ratio: 1.00 on MESA; 1.17 on ELSA), similarly to the best-performing existing models. However, while the existing models yielded a large percentage of missing predictions (17%-45% on MESA; 63%-64% on ELSA), this was negligible with the combined model (0% on MESA, 4% on ELSA).

Leveraging on existing literature T2D predictive models, a simple approach based on risk score rescaling and averaging was shown to provide accurate and robust incidence predictions, overcoming the problem of recalibration and missing predictions in practical application of predictive models.

Mathematical physiological models of type 1 diabetes (T1D) glucose-insulin dynamics have been of great help in designing and preliminary assessing new algorithm for glucose control. Derivation of models at the individual level is however difficult because of identifiability issues. Recently, fitting these models against data of real patients with T1D has been made possible by both the use of Bayesian estimation techniques and the availability of individual datasets including plasma glucose and insulin concentration samples gathered in clinical protocols. The aim of this work is to make a step further and develop a methodology able to estimate the parameters of T1D physiological models using easily accessible data only, i.e. continuous glucose monitoring (CGM) sensor, carbohydrate intakes (CHO), and exogenous insulin infusion (I) data. The methodology is tested on synthetic data of 100 patients generated by a composite model of glucose-insulin dynamics. To solve identifiability problems, a Bayesian approach numerically implemented by Markov Chain Monte Carlo (MCMC) has been used to obtain point estimates and confidence intervals of model unknown parameters exploiting a priori knowledge available from the literature. Results show goodness of model fit and acceptable precision of parameter estimates. The methodology is also successful in reconstructing of "non-accessible" glucose-insulin fluxes, i.e. glucose rate of appearance and plasma insulin. These preliminary results encourage further development of this framework and its assessment in more challenging setups.

Minimally-invasive continuous glucose monitoring (CGM) sensors are used in diabetes therapy to monitor interstitial glucose (IG) concentration almost continuously (e.g. every 5 min) and detect/predict dangerous hypo/hyperglycemic episodes. When compared with frequent blood glucose (BG) concentration references, CGM measurements are unavoidably affected by error. Models of the CGM error can be important in several applications, e.g. for testing in simulation the safety and effectiveness of CGM-based artificial pancreas algorithms. In this work, we model the error of the Dexcom G6, a CGM sensor that recently entered the market and does not require in vivo calibrations. The dataset includes CGM and BG data collected in 11 subjects wearing two Dexcom G6 sensors in parallel. The model is derived applying a methodology to dissect and model 3 main CGM error components: BG-to-IG kinetics, calibration error and measurement noise. An aspect of novelty of the method is its capability of handling factory-calibrated CGM sensor data. Results of model identification show that the time-variability of sensor calibration error during the sensor lifetime (10 days) can be well represented by a regression model with time-variant parameters described by 2^nd-order polynomials in time.

Previous literature has demonstrated that hypoglycemic events in patients with type 1 diabetes (T1D) are associated with measurable scalp electroencephalography (EEG) changes in power spectral density. In the present study, we used a dataset of 19-channel scalp EEG recordings in 34 patients with T1D who underwent a hyperinsulinemic-hypoglycemic clamp study. We found that hypoglycemic events are also characterized by EEG complexity changes that are quantifiable at the single-channel level through empirical conditional and permutation entropy and fractal dimension indices, i.e., the Higuchi index, residuals, and tortuosity. Moreover, we demonstrated that the EEG complexity indices computed in parallel in more than one channel can be used as the input for a neural network aimed at identifying hypoglycemia and euglycemia. The accuracy was about 90%, suggesting that nonlinear indices applied to EEG signals might be useful in revealing hypoglycemic events from EEG recordings in patients with T1D.

Factory-calibrated continuous glucose monitoring (FC-CGM) sensors are new devices used in type 1 diabetes (T1D) therapy to measure the glucose concentration almost continuously for 10-14 days without requiring any in vivo calibration. Understanding and modelling CGM errors is important when designing new tools for T1D therapy. Available literature CGM error models are not suitable to describe the FC-CGM sensor error, since their domain of validity is limited to 12-h time windows, i.e., the time between two consecutive in vivo calibrations. The aim of this paper is to develop a model of the error of FC-CGM sensors. The dataset used contains 79 FC-CGM traces collected by the Dexcom G6 sensor. The model is designed to dissect the error into its three main components: effect of plasma-interstitium kinetics, calibration error, and random measurement noise. The main novelties are the model extension to cover the entire sensor lifetime and the use of a new single-step identification procedure. The final error model, which combines a first-order linear dynamic model to describe plasma-interstitium kinetics, a second-order polynomial model to describe calibration error, and an autoregressive model to describe measurement noise, proved to be suitable to describe FC-CGM sensor errors, in particular improving the estimation of the physiological time-delay.

Artificial pancreas (AP) technology has been proven to improve glucose and patient-centered outcomes for people with type 1 diabetes (T1D). Several approaches to implement the AP have been described, clinically evaluated, and in one case, commercialized. However, none of these approaches has shown a clear superiority with respect to others. In addition, several challenges still need to be solved before achieving a fully automated AP that fulfills the users' expectations. We have introduced the Bio-inspired Artificial Pancreas (BiAP), a hybrid adaptive closed-loop control system based on beta-cell physiology and implemented directly in hardware to provide an embedded low-power solution in a dedicated handheld device. In coordination with the closed-loop controller, the BiAP system incorporates a novel adaptive bolus calculator which aims at improving postprandial glycemic control. This paper focuses on the latest developments of the BiAP system for its utilization in the home environment.

The hardware and software architectures of the BiAP system designed to be used in the home environment are described. Then, the clinical trial design proposed to evaluate the BiAP system in an ambulatory setting is introduced. Finally, preliminary results corresponding to two participants enrolled in the trial are presented.

Apart from minor technical issues, mainly due to wireless communications between devices, the BiAP system performed well (~88% of the time in closed-loop) during the clinical trials conducted so far. Preliminary results show that the BiAP system might achieve comparable glycemic outcomes to the existing AP systems (~73% time in target range 70-180 mg/dL).

The BiAP system is a viable platform to conduct ambulatory clinical trials and a potential solution for people with T1D to control their glucose control in a home environment.

Background: The standard treatment for hypoglycemia recommended by the American Diabetes Association (ADA) suggests patients with diabetes to take small amounts of carbohydrates, the so-called hypotreatments (HTs), as soon as blood glucose concentration goes below 70 mg/dL. However, prevention, or at least mitigation, of hypoglycemic events could be achieved by triggering HTs ahead of time thanks to the use of the predictive capabilities of suitable real-time algorithms fed by continuous glucose monitoring (CGM) sensor data. Materials and Methods: The algorithm proposed in this article to trigger HTs for preventing forthcoming hypoglycemic events is based on the computation of the "dynamic risk", there is a nonlinear function combining current glycemia with its rate-of-change, both provided by CGM. A comparison of performance of the proposed algorithm against the ADA guidelines is made, in silico, on datasets of 100 virtual patients undergoing a single-meal experiment, with induced postmeal hypoglycemia, generated by the UVA/Padova type 1 diabetes simulator. Results: On noise-free CGM data, the proposed algorithm reduces the time spent in hypoglycemia, on median [25th-75th percentiles] from 36 [29-43] to 0 [0-11] min (P < 0.0001), with a concomitant decrease of the post-treatment rebound (PTR) in glucose concentration, on median [25th-75th percentiles] from 136 [121-148] to 121 [116-127] mg/dL (P < 0.0001). On noisy CGM data, there is still a reduction of both time spent in hypoglycemia from 41 [28-49] min to 25 [0-41] min (P < 0.0001) and PTR from 174 [146-189] mg/dL to 137 [123-151] mg/dL (P < 0.0001). Conclusions: The potentiality of the new algorithm in generating preventive HTs, which can allow significant reduction of hypoglycemia without concomitant increase of hyperglycemia, suggests its further development and test in silico, for example, simulating both insulin pump and multiple-daily-injection therapies.

Recent development of automated closed-loop (CL) insulin delivery systems, the so-called artificial pancreas (AP), improved the quality of type 1 diabetes (T1D) therapy. As new technologies emerge, patients put increasing trust in their therapeutic devices; therefore, it becomes increasingly important to detect malfunctioning affecting such devices. In this work, we explore a new paradigm to detect insulin pump faults (IPFs) that use unsupervised anomaly detection.

We generated CL data corrupted with IPFs using the latest version of the T1D Padova/UVA simulator. From the data, we extracted several features capable to describe the patient dynamics and making more apparent suspicious data portions. Then, a feature selection is performed to determine the optimal feature set. Finally, the performance of several popular unsupervised anomaly detection algorithms is analyzed and compared on the identified optimal feature set.

Using the identified optimal configuration, the best performance is obtained by the Histogram-Based Outlier Score (HBOS) algorithm, which detected 87% of the IPF with only 0.08 false positives per day on average. Isolation forest is the best algorithm that offers more conservative performances, detection of 85% of the faults but only 0.06 false positives per day on average.

Unsupervised anomaly detection algorithms can be used effectively to detect IPFs and improve the safety of the AP. Future studies will be dedicated to test the presented method inside dedicated clinical trials.

In type 1 diabetes management, maintaining nocturnal blood glucose within target range can be challenging. Although semi-automatic systems to modulate insulin pump delivery, such as low-glucose insulin suspension and the artificial pancreas, are starting to become a reality, their elevated cost and performance below user expectations is hindering their adoption. Hence, a decision support system that helps people with type 1 diabetes, on multiple daily injections or insulin pump therapy, to avoid undesirable overnight blood glucose fluctuations (hyper- or hypoglycaemic) is an attractive alternative. In this paper, we introduce a novel data-driven approach to predict the quality of overnight glycaemic control in people with type 1 diabetes by analyzing commonly gathered data during the day-time period (continuous glucose monitoring data, meal intake and insulin boluses). The proposed approach is able to predict whether overnight blood glucose concentrations are going to remain within or outside the target range, and therefore allows the user to take the appropriate preventive action (snack or change in basal insulin). For this purpose, a number of popular established machine learning algorithms for binary classification were evaluated and compared on a publicly available clinical dataset (i.e., OhioT1DM). Although there is no clearly superior classification algorithm, this study indicates that, by using commonly gathered data in type 1 diabetes management, it is possible to predict the quality of overnight glycaemic control with reasonable accuracy (AUC-ROC = 0.7).