Fast Linde-Buzo-Gray (FLBG) Algorithm for Image Compression through Rescaling Using Bilinear Interpolation.

Vector quantization (VQ) is a block coding method that is famous for its high compression ratio and simple encoder and decoder implementation. Linde-Buzo-Gray (LBG) is a renowned technique for VQ that uses a clustering-based approach for finding the optimum codebook. Numerous algorithms, such as Particle Swarm Optimization (PSO), the Cuckoo search algorithm (CS), bat algorithm, and firefly algorithm (FA), are used for codebook design. These algorithms are primarily focused on improving the image quality in terms of the PSNR and SSIM but use exhaustive searching to find the optimum codebook, which causes the computational time to be very high. In our study, our algorithm enhances LBG by minimizing the computational complexity by reducing the total number of comparisons among the codebook and training vectors using a match function. The input image is taken as a training vector at the encoder side, which is initialized with the random selection of the vectors from the input image. Rescaling using bilinear interpolation through the nearest neighborhood method is performed to reduce the comparison of the codebook with the training vector. The compressed image is first downsized by the encoder, which is then upscaled at the decoder side during decompression. Based on the results, it is demonstrated that the proposed method reduces the computational complexity by 50.2% compared to LBG and above 97% compared to the other LBG-based algorithms. Moreover, a 20% reduction in the memory size is also obtained, with no significant loss in the image quality compared to the LBG algorithm.

Generative deep learning for the development of a type 1 diabetes simulator.

BACKGROUND: Type 1 diabetes (T1D) simulators, crucial for advancing diabetes treatments, often fall short of capturing the entire complexity of the glucose-insulin system due to the imprecise approximation of the physiological models. This study introduces a simulation approach employing a conditional deep generative model. The aim is to overcome the limitations of existing T1D simulators by synthesizing virtual patients that more accurately represent the entire glucose-insulin system physiology. METHODS: Our methodology utilizes a sequence-to-sequence generative adversarial network to simulate virtual T1D patients causally. Causality is embedded in the model by introducing shifted input-output pairs during training, with a 90-min shift capturing the impact of input insulin and carbohydrates on blood glucose. To validate our approach, we train and evaluate the model using three distinct datasets, each consisting of 27, 12, and 10 T1D patients, respectively. In addition, we subject the trained model to further validation for closed-loop therapy, employing a state-of-the-art controller. RESULTS: The generated patients display statistical similarity to real patients when evaluated on the time-in-range results for each of the standard blood glucose ranges in T1D management along with means and variability outcomes. When tested for causality, authentic causal links are identified between the insulin, carbohydrates, and blood glucose levels of the virtual patients. The trained generative model demonstrates behaviours that are closer to reality compared to conventional T1D simulators when subjected to closed-loop insulin therapy using a state-of-the-art controller. CONCLUSIONS: These results highlight our approach's capability to accurately capture physiological dynamics and establish genuine causal relationships, holding promise for enhancing the development and evaluation of therapies in diabetes.

New therapies and treatments for type 1 diabetes (T1D) are often first tested on specialized computer programs called simulators before being tried on actual patients. Traditionally, these simulators rely on mathematical equations to mimic real-life patients, but they sometimes fail to provide reliable results because they do not consider everything that affects individuals with diabetes, such as lifestyle, eating habits, time of day, and weather. In our research, we suggest using computer programs based on artificial intelligence that can directly learn all these factors from real patient data. We tested our programs using information from different groups of patients and found that they were much better at predicting what would happen with a patient's diabetes. These new programs can understand how insulin, food, and blood sugar levels interact in the body, which makes them valuable for developing therapies for T1D.

Incidence of postoperative pulmonary congestion as diagnosed by lung ultrasound in surgeries performed under general anaesthesia: A prospective, observational study.

Background and Aims: Administering liberal fluid raises concerns about pulmonary congestion postoperatively. Bedside ultrasonography is a valuable tool for the early detection of pulmonary congestion. In this study, we have used it to ascertain the impact of the duration of surgery and intraoperative fluid volume on the causation of pulmonary congestion. Our objective was to determine the incidence of pulmonary congestion as diagnosed by lung ultrasound in patients undergoing general anaesthesia with varied fluid administration. Methods: Seventy participants of American Society of Anesthesiologists physical status I and II, aged between 18 and 60 years, undergoing elective extrathoracic surgeries of over 3 h under general anaesthesia were included. Preoperative lung ultrasound was carried out in all patients, and a postoperative lung ultrasound was carried out at 1 h after extubation. The appearance of three or more "B"-lines was considered positive for lung congestion. Results: Significant differences (P < 0.001) were found in the duration of surgery and the appearance of B-lines in the postoperative period. Participants who developed B lines received, on average, 150% more fluid (1148.16 ± 291.79 ml) than those who did not (591.29 ± 398.42 ml) (P = 0.0240). Net fluid balance was also significantly different in patients who developed B lines (P = 0.0014). None of the patients developed symptoms of lung congestion postoperatively. Conclusion: Long duration of surgery under general anaesthesia (>3 h) with the administration of large volumes of intraoperative fluid and a large net fluid balance are associated with lung congestion as diagnosed by lung ultrasound.

Mode of Mechanical Ventilation in a Case of Venolymphatic Malformation: Spontaneous-Saves, Positive-Precludes.

Mediastinal venolymphatic malformations (VLM) are rare tumours, with very few reported cases in the literature. Arising often from the anterior mediastinum, VLM manifests symptoms based on invaded surrounding structures. Masses from the anterior and superior mediastinum pose an anaesthetic challenge for airway and hemodynamic management. A 7-month-old male child presented with a progressively growing mass over the left anterior chest wall for one month, about 4x4 cm, with diffuse margins and now expanded to involve the root of the neck and into the axilla. The patient was free from any apparent systemic illness. The breathing difficulty worsened in the past week with noisy respiration associated with feeding difficulty and hence sought medical admission to the paediatrics emergency unit. In conclusion, such huge mediastinal masses are managed better under spontaneous ventilation with an adequate surgical depth of anaesthesia to maintain appropriate respiratory compliance and necessitate lower peak inspiratory pressure. Given rare cases reported in the literature, similar topics would help choose the modus of ventilation and their safe management.

Generation of Individualized Synthetic Data for Augmentation of the Type 1 Diabetes Data Sets Using Deep Learning Models.

In this paper, we present a methodology based on generative adversarial network architecture to generate synthetic data sets with the intention of augmenting continuous glucose monitor data from individual patients. We use these synthetic data with the aim of improving the overall performance of prediction models based on machine learning techniques. Experiments were performed on two cohorts of patients suffering from type 1 diabetes mellitus with significant differences in their clinical outcomes. In the first contribution, we have demonstrated that the chosen methodology is able to replicate the intrinsic characteristics of individual patients following the statistical distributions of the original data. Next, a second contribution demonstrates the potential of synthetic data to improve the performance of machine learning approaches by testing and comparing different prediction models for the problem of predicting nocturnal hypoglycemic events in type 1 diabetic patients. The results obtained for both generative and predictive models are quite encouraging and set a precedent in the use of generative techniques to train new machine learning models.

The Relationship of Respiratory Rate-Oxygenation (ROX) and Modified ROX Index With High-Flow Nasal Cannula Oxygen Therapy in COVID-19 Patients: An Observational Pilot Study.

Background and aim Respiratory Rate-Oxygenation (ROX) and modified ROX (mROX) indexes have been proposed to detect early high-flow nasal cannula (HFNC) therapy failure. We evaluated the utility and relationship of ROX and mROX indexes in COVID-19 patients started on HFNC oxygen therapy. Methods This pilot study collected data from adult COVID-19 patients requiring HFNC oxygenation from 29 Jan - 29 Jun 2021. The patients were divided into two cohorts based on HFNC therapy success. ROX and mROX were compared using statistical diagnostic testing, including receiver operating characteristics and area under the curve (AUC) using online Epitools (https://epitools.ausvet.com.au/) and MedCalc software (MedCalc Software Ltd, Ostend, Belgium, https://www.medcalc.org/); p<0.05 was considered significant. Results Twenty-seven patients fulfilled the inclusion criteria; 48.15% of therapy failed. The cohort's mean ± standard deviation age was 53.93 ± 10.67 years; 74.1% were male. The accuracy of predicting failure for mean ROX versus mROX at baseline and six-hour values was 59.81 versus 70.68 and 67.42 versus 74.88, respectively (all p>0.05). The AUC for ROX and mROX at baseline and at six hours were statistically indifferent. Only an mROX of 4.05 (mean value) and 3.34 (Youden's J cut-off) had a sensitivity plus specificity at 156% and 163%, respectively. Conclusion Both ROX and mROX at baseline and six hours had fair-to-good accuracies and AUC; the differences were insignificant. Both ROX and mROX had better accuracies at six hours. However, only mROX < 4.05 at six hours fulfilled the sensitivity plus specificity criteria to be a clinically valuable screener.

Machine Learning Techniques for Hypoglycemia Prediction: Trends and Challenges.

(1) Background: the use of machine learning techniques for the purpose of anticipating hypoglycemia has increased considerably in the past few years. Hypoglycemia is the drop in blood glucose below critical levels in diabetic patients. This may cause loss of cognitive ability, seizures, and in extreme cases, death. In almost half of all the severe cases, hypoglycemia arrives unannounced and is essentially asymptomatic. The inability of a diabetic patient to anticipate and intervene the occurrence of a hypoglycemic event often results in crisis. Hence, the prediction of hypoglycemia is a vital step in improving the life quality of a diabetic patient. The objective of this paper is to review work performed in the domain of hypoglycemia prediction by using machine learning and also to explore the latest trends and challenges that the researchers face in this area; (2) Methods: literature obtained from PubMed and Google Scholar was reviewed. Manuscripts from the last five years were searched for this purpose. A total of 903 papers were initially selected of which 57 papers were eventually shortlisted for detailed review; (3) Results: a thorough dissection of the shortlisted manuscripts provided an interesting split between the works based on two categories: hypoglycemia prediction and hypoglycemia detection. The entire review was carried out keeping this categorical distinction in perspective while providing a thorough overview of the machine learning approaches used to anticipate hypoglycemia, the type of training data, and the prediction horizon.