A computer aided diagnostic method for the evaluation of type II diabetes mellitus in facial thermograms

A Review of Facial Thermography Assessment for Vital Signs Estimation

“…The same apparent for estimated BP, which may favor the need for an invasive PPG sensor [63] and a sphygmomanometer [64] to analyze the level of BP. In general, standard biochemical invasive methods such as glucometer can be used to collect blood glucose [44]. It differs from the research proposed in [64], which might include an Olympus auto analyzer and the research proposed in [56], which uses a hyperinsulinemia glucose clamping procedure and a Continuous Subcutaneous Glucose Monitor (CGM) in their biochemical analysis.…”

“…Based on research trends, one possible explanation is that most existing computational models, such as ML algorithms have been widely used in feature extraction. According to studies, Support Vector Machine (SVM) classifiers either in linear, radial basis function, quadratic or polynomial kernel have been frequently implemented in object detection due to their ability to manage data with high correlation, faster classification and produce highly accurate results in the range of 80-100% accuracy [28], [44], [45]. As confirmations of the studies, classifiers such as Random Forest (RF), Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN), Bayesian Generalized Gaussian Models (BGMM) and others have been employed to verify the effectiveness of extractor methods.…”

“…Normal blood oxygenation levels range from 95%-100%, but when oxygen saturation drops below 90%, it is considered low and may result in shortness of breath or hypoxemia [1]. Blood glucose is a biochemical analysis that measures the sugar concentration in a blood sample [44]. It may be affected by the pancreas, which can either produce insulin or not in the body which can be leading to hypoglycemia and hyperglycemia of blood glucose levels [56].…”

“…For the statistical analysis, the mean and standard deviation are two possible statistical measurements that had been used frequently in most published works specifically in calculating the vital signs estimations for summarizing data variability. Other measurements have also been considered, for example in the experiment described in [44] where statistical analysis was performed by including twenty two measurements to compute their extractor methods, in contrast to the proposed works in [6], which only comprised four types of measurements. Table VII shows the list of the statistical analysis that had been used in the prior research.…”

Exploration of Machine Learning and Statistical Techniques in Development of a Low-Cost Screening Method Featuring the Global Diet Quality Score for Detecting Prediabetes in Rural India

“…Machine learning (ML) and statistical methods may assist in discovering patterns present in data that are predictive of diabetes risk. ML techniques can be quite effective in identifying prediabetes, although many predictive algorithms currently in existence require expensive imaging technology or laboratory measurements ( 4 ). Some low-cost scoring methods using a combination of ML techniques and questionnaire data have been shown to be effective in screening for T2D ( 5 , 6 ).…”