Introduction: Malaysia has one of the highest diabetes prevalence rates in Asia (1 in 5), with >50% fail to achieve optimal glycemic control. Continuous glucose monitoring (CGM) provides detailed insights beyond hemoglobin A1c (HbA1c), capturing daily glucose fluctuations and variability. Realworld data describing CGM patterns and their clinical associations in Malaysian adults with type 2 diabetes mellitus (T2DM) are limited. Methodology: This cross-sectional study analyzed CGM data from 25 Malaysian adults with T2DM, standardized over 14 days. CGM metrics included time in range (TIR), time above range (TAR), time below range (TBR), mean glucose, and coefficient of variation (CV). Weekday–weekend comparisons were performed, and correlations with age, diabetes duration, HbA1c, body mass index (BMI), treatment regimen, and complications were assessed. Results: Participants had a mean age of 56.2 ± 13.2 years (52% male), mean HbA1c 9.4 ± 2.7%, diabetes duration 10.7 ± 9.0 years, and BMI 29.8 ± 9.4 kg/m². Mean TIR, TAR, TBR, mean glucose, and CV were similar between weekdays and weekends (p >0.34 for all). TIR >70% was achieved by 52% of participants on weekdays and 48% on weekends, with no statistically significant difference (p = 0.75). Longer diabetes duration correlated with lower TIR (r = −0.62, p <0.001), higher mean glucose (r = 0.58, p = 0.002), and greater variability (r = 0.54, p = 0.004). Older age was associated with lower TIR (r = −0.48, p = 0.014) and higher mean glucose (r = 0.44, p = 0.025). Higher HbA1c correlated with lower TIR (r = −0.36, p = 0.04), higher TAR (r = 0.35, p = 0.05), greater variability (r = 0.51, p = 0.006), and increased target organ damage (ρ = 0.55, p = 0.004). BMI was not associated with TIR (r = −0.18, p = 0.38). Participants with ≥2 complications had lower TIR (55.8 ± 28.4% vs 76.7 ± 19.2%, p = 0.073), while insulin the Conclusion In Malaysian adults with T2DM, CGM metrics were similar between weekdays and weekends, indicating lifestyle differences had minimal impact on glycemic control. Longer diabetes duration, older age, higher HbA1c, multiple complications, and insulin therapy identified patients at highest risk for poor glycemic control, greater variability, and hypoglycemia. These findings support the use of CGM for risk stratification, individualized monitoring, and therapy optimization to reduce complications and hypoglycemia risk.
Adult ; Blood Glucose ; Blood Glucose Self-Monitoring ; Continuous Glucose Monitoring ; Diabetes Mellitus, Type 2

Introduction: ISPAD guidelines recommend continuous glucose monitoring (CGM) as the standard of care for paediatric type 1 diabetes (T1DM). However, a “real-world” adoption gap persists, particularly in resource-limited settings. The Introductions of the study were to evaluate CGM adoption prevalence, identify documented barriers, and compare glycemic outcomes between active and non-active users. Methodology: This retrospective review analyzed electronic medical records (EMR) of 125 paediatric T1DM patients at Hospital Putrajaya (2025). Data included CGM status, insulin delivery method, and documented barriers. Independent T-tests compared mean hemoglobin A1c (HbA1c) between groups, and multivariable logistic regression identified independent predictors of adoption. Results: Cohort mean age was 11.8 ± 3.8 years. Active CGM users were 32.8% (n = 41), of whom 29.3% (n = 12) utilized predominantly automated insulin delivery (AID) systems. The remaining 67.2% (n = 84) were classified as non-active users, comprising both never-users and ex-users (discontinued use). Active users achieved significantly lower mean HbA1c than non-active users (8.73% vs 9.86%; p <0.001), with no significant difference in rates of DKA (p = 0.564) or severe hypoglycemia (p = 0.250). Among never-users, 42.9% lacked documented technology counselling (p = 0.001). Multivariable analysis identified funding source as the sole independent predictor of CGM adoption (adjusted OR = 7.76, p <0.001). While the primary documented barrier was financial (31.0%), a lack of documented barriers was noted in 61.9% of non-active users. Conclusion A substantial technology gap exists, primarily driven by financial access rather than clinical demographics. The difference of 1.13% in HbA1c between groups underscores the need to address financial setbacks to improve technology access in Malaysia and prevent diabetes complications.
Child ; Blood Glucose ; Blood Glucose Self-Monitoring ; Continuous Glucose Monitoring ; Diabetes Mellitus, Type 1

Humans ; Drug Monitoring/methods* ; Child ; Inflammatory Bowel Diseases/drug therapy* ; Immunosuppressive Agents/therapeutic use* ; China ; Biological Products/therapeutic use*

Cardiotocography (CTG) is a non-invasive and important tool for diagnosing fetal distress during pregnancy. To meet the needs of intelligent fetal heart monitoring based on deep learning, this paper proposes a TWD-MOAL deep active learning algorithm based on the three-way decision (TWD) theory and multi-objective optimization Active Learning (MOAL). During the training process of a convolutional neural network (CNN) classification model, the algorithm incorporates the TWD theory to select high-confidence samples as pseudo-labeled samples in a fine-grained batch processing mode, meanwhile low-confidence samples annotated by obstetrics experts were also considered. The TWD-MOAL algorithm proposed in this paper was validated on a dataset of 16 355 prenatal CTG records collected by our group. Experimental results showed that the algorithm proposed in this paper achieved an accuracy of 80.63% using only 40% of the labeled samples, and in terms of various indicators, it performed better than the existing active learning algorithms under other frameworks. The study has shown that the intelligent fetal heart monitoring model based on TWD-MOAL proposed in this paper is reasonable and feasible. The algorithm significantly reduces the time and cost of labeling by obstetric experts and effectively solves the problem of data imbalance in CTG signal data in clinic, which is of great significance for assisting obstetrician in interpretations CTG signals and realizing intelligence fetal monitoring.
Humans ; Pregnancy ; Female ; Cardiotocography/methods* ; Deep Learning ; Neural Networks, Computer ; Algorithms ; Fetal Monitoring/methods* ; Heart Rate, Fetal ; Fetal Distress/diagnosis* ; Fetal Heart/physiology*

Heart failure (HF) is the end-stage of all cardiac diseases, characterized by high prevalence, high mortality, and heavy social and economic burden. Early warning of HF exacerbation is of great value for outpatient management and reducing readmission rates. Currently, remote dynamic monitoring technology, which captures changes in hemodynamic and physiological parameters of HF patients, has become the primary method for early warning and is a hot research topic in clinical studies. This paper systematically reviews the progress in this field, which was categorized into invasive monitoring based on implanted devices, non-invasive monitoring based on wearable devices, and other monitoring technologies based on audio and video. Invasive monitoring primarily involves direct hemodynamic parameters such as left atrial pressure and pulmonary artery pressure, while non-invasive monitoring covers parameters such as thoracic impedance, electrocardiogram, respiration, and activity levels. These parameters exhibit characteristic changes in the early stages of HF exacerbation. Given the clinical heterogeneity of HF patients, multi-source information fusion analysis can significantly improve the prediction accuracy of early warning models. The results of this study suggest that, compared with invasive monitoring, non-invasive monitoring technology, with its advantages of good patient compliance, ease of operation, and cost-effectiveness, combined with AI-driven multimodal data analysis methods, shows significant clinical application potential in establishing an outpatient management system for HF.
Humans ; Heart Failure/physiopathology* ; Monitoring, Physiologic/methods* ; Wearable Electronic Devices ; Remote Sensing Technology ; Early Diagnosis ; Electrocardiography ; Hemodynamics

BACKGROUND: In the real world, the plasma drug concentration range of Icotinib treated with epidermal growth factor receptor (EGFR) gene mutant non-small cell lung cancer (NSCLC) is not yet clear, and there may be a correlation between drug concentration and its efficacy, as well as adverse reactions. This study conducted therapeutic drug monitoring (TDM) of Icotinib. The aim of this study was to analyze the drug exposure of Icotinib in targeted therapy for NSCLC, and to investigate the relationship between Icotinib drug concentration and its efficacy and safety. METHODS: Prospective blood samples were collected from NSCLC patients with EGFR-sensitive mutations who received treatment with Icotinib in Peking University Cancer Hospital from April 2022 to July 2024. The drug trough concentration of Icotinib in plasma was detected, and the correlation between drug concentration and efficacy, as well as the toxic side effects, were further analyzed based on the patient's clinical medical records. RESULTS: 22 patients who were treated with Icotinib underwent TDM, but one of them did not acquire the data due to prolonged discontinuation. The remaining 21 patients, each with 1-7 blood draws, obtained a total of 32 plasma drug concentration data. The drug concentration of icotinib is a range of 126.9-2317.1 ng/mL. Among the 21 patients, 18 cases were female (85.7%), and 3 cases were male (14.3%), with an age range of 44-85 years old. The pathological types are all lung adenocarcinoma. Except for 5 patients receiving postoperative adjuvant therapy, 16 patients had assessable tumors. The objective response rate was 43.8% (7/16), and the disease control rate reached 100.0% (16/16). The median value of drug concentration is 805.5 ng/mL among those 21 patients. Compared with the patients who achieved stable disease, the median value of drug concentrations of Icotinib in patients who achieved partial response were 497.2 and 1195.5 ng/mL, respectively (P=0.017). The median value of drug concentrations for patients who did not experience adverse reactions during treatment and those who experienced adverse reactions were 997.0 and 828.6 ng/mL, respectively (P=0.538). CONCLUSIONS Icotinib demonstrates good therapeutic effect and tolerable toxicity on the EGFR gene mutant NSCLC. There is a certain negative correlation between the plasma drug concentration of Icotinib and its efficacy, while there seems no significant correlation with safety.
Humans ; Carcinoma, Non-Small-Cell Lung/genetics* ; ErbB Receptors/metabolism* ; Lung Neoplasms/genetics* ; Male ; Female ; Crown Ethers/blood* ; Middle Aged ; Drug Monitoring ; Aged ; Quinazolines/blood* ; Mutation ; Adult ; Aged, 80 and over ; Antineoplastic Agents/blood* ; Prospective Studies

The application of small molecule targeted drugs for lung cancer has significantly improved the survival of lung cancer patients. However, these drugs have a wide variety of types, fast development and market launch of new drugs, complex adverse reactions, and are mostly used at home, which increases the risk of irrational drug use. At the same time, insufficient monitoring of efficacy and safety is also prone to occur, ultimately affecting treatment outcomes. This consensus focuses on 43 small molecule targeted drugs or combinations for lung cancer, providing standardized recommendations for rational drug use and monitoring of efficacy/adverse reactions in clinical practice. The recommendations are regarding drug selection, dosage adjustment, efficacy monitoring, adverse reaction monitoring, and improvement of patient compliance. This consensus aims to improve the rational use and efficacy/safety monitoring quality of small molecule targeted drugs for lung cancer, ensure the effectiveness and safety of drug treatment, prolong the survival of lung cancer patients and improve their quality of life.
.
Humans ; Lung Neoplasms/drug therapy* ; Antineoplastic Agents/adverse effects* ; Consensus ; Molecular Targeted Therapy ; Drug Monitoring ; Small Molecule Libraries/therapeutic use*

Sleep-related hypermotor epilepsy （SHE） is a rare type of epilepsy with a prevalence rate of approximately 1.8/100 000. This disease mainly manifests as complex motor behaviors during non-rapid eye movement sleep， such as leg kicking， arm waving， and sitting up. Since such symptoms are similar to non-epileptic disorders such as night terrors and sleepwalking and abnormal discharges may not be observed on electroencephalography， the diagnosis of SHE is quite challenging. Currently， there is still a lack of evidence from large-scale randomized controlled studies to support pharmacological treatment strategies for SHE， and related data in China remain scarce. This article reports a case of SHE， in order to provide a clinical reference for the diagnosis and medication treatment of this disease.
Polysomnography

Objective To investigate the sleep architecture of sleep related bruxism（SB）in adults and the sleep architecture of SB comorbid with obstructive sleep apnea hypopnea syndrome（OSAHS），as well as their correlation with age and other factors. Methods A total of 51 subjects with SB and 67 controls were included in this study to analyze the sleep architecture of SB and compare the sleep architecture of SB comorbid with different severities of OSAHS. Results Compared with the control group，the SB group had a younger age，increases in N1（%TST）and N2（%TST），a reduction in N3（%TST），and an increase in arousal index. The SB group was divided into non-OSAHS group（group 1），mild OSAHS group（group 2），and moderate-to-severe OSAHS group（group 3）. Group 1 had a younger age than group 2 and group 3，and group 3 had increases in body mass index（BMI），N1（%TST），oxygen desaturation index（ODI），and arousal index and a reduction in N3（%TST）. The Spearman's rank correlation analysis showed that BMI，N1（%TST），arousal index，and ODI increased with the increase in apnea-hypopnea index（AHI），while N3（%TST）decreased with the increase in AHI. The binary logistic regression analysis showed that SB was negatively correlated with age and was positively correlated with arousal index. Conclusion SB may affect sleep architecture by increasing light sleep，reducing deep sleep，and increasing the number of awakenings. There are changes in sleep architecture in case of SB comorbid with different severities of OSAHS. SB is negatively correlated with age and is positively correlated with arousal index.
Polysomnography

INTRODUCTION: Obstructive sleep apnoea (OSA) is a serious but underdiagnosed condition. Demand for the gold standard diagnostic polysomnogram (PSG) far exceeds its availability. More efficient diagnostic methods are needed, even in tertiary settings. Machine learning (ML) models have strengths in disease prediction and early diagnosis. We explored the use of ML with oximetry, demographic and anthropometric data to diagnose OSA. METHODS: A total of 2,996 patients were included for modelling and divided into test and training sets. Seven commonly used supervised learning algorithms were trained with the data. Sensitivity (recall), specificity, positive predictive value (PPV) (precision), negative predictive value, area under the receiver operating characteristic curve (AUC) and F1 measure were reported for each model. RESULTS: In the best performing four-class model (neural network model predicting no, mild, moderate or severe OSA), a prediction of moderate and/or severe disease had a combined PPV of 94%; one out of 335 patients had no OSA and 19 had mild OSA. In the best performing two-class model (logistic regression model predicting no-mild vs. moderate-severe OSA), the PPV for moderate-severe OSA was 92%; two out of 350 patients had no OSA and 26 had mild OSA. CONCLUSION Our study showed that the prediction of moderate-severe OSA in a tertiary setting with an ML approach is a viable option to facilitate early identification of OSA. Prospective studies with home-based oximeters and analysis of other oximetry variables are the next steps towards formal implementation.
Humans ; Oximetry/methods* ; Sleep Apnea, Obstructive/diagnosis* ; Male ; Female ; Middle Aged ; Machine Learning ; Polysomnography ; Adult ; Anthropometry ; ROC Curve ; Aged ; Algorithms ; Predictive Value of Tests ; Sensitivity and Specificity ; Neural Networks, Computer ; Demography