Evaluation of Diagnostic Performance of Machine Learning Algorithms to Classify the Fetal Heart Rate Baseline from Cardiotocograph

Cardiotocography (CTG) is the widely used cost-effective, non-invasive technique to monitor the fetal heart and mother’s uterine contraction pressure to assess the wellbeing of the fetus. The most important parameters of fetal heart is the baseline upon which the other parameters viz. acceleration, deceleration and variability depend. Accurate classification of the baseline into either normal, bradycardia or tachycardia is thus important to assess the fetal-health. Since visual estimation has its limitations, the authors use various Machine Learning Algorithms to classify the baseline. 110 CTG traces from CTU-UHB dataset, were divided into three subsets using stratified sampling to ensure that the sample is the accurate depiction of the population. The results were analyzed using various statistical methods and compared with the visual estimation by three obstetricians. FURIA provided greatest accuracy of 98.11%. From the analysis of Bland-Altman Plot FURIA was also found to have best agreement with physicians’ estimation.

Accuracy of an artificial intelligence-based model for estimating leftover liquid food in hospitals: validation study (Preprint)

BACKGROUND The accurate evaluation of the nutritional status of malnourished hospitalized patients at a higher risk of complications, such as frailty or disability, is crucial. Visual methods of estimating food intake are popular for evaluating the nutritional status in clinical environments. However, from the perspective of accurate measurement, such methods are unreliable. OBJECTIVE The accuracy of estimating leftover liquid food in hospitals using an artificial intelligence (AI)-based model was compared to that of visual estimation. METHODS The accuracy of obtained using the AI-based model (AI estimation) was compared to that of the visual estimation method for thin rice gruel as staple food, and fermented milk and peach juice as side dishes. A total of 576 images of liquid food (432 images of thin rice gruel, 72 of fermented milk, and 72 of peach juice) were used. Welch's t-test and the confusion matrix were used to examine the difference of mean absolute error between AI and visual estimation. RESULTS The mean absolute errors obtained through the AI estimation approach for side dishes were as follows: fermented milk: 0.63 and peach juice: 0.25. These were significantly smaller than those obtained using the visual estimation approach: fermented milk: 1.40 and peach juice: 0.90. Contrastingly, the mean absolute error for staple food obtained using the AI estimation method (0.99) did not differ significantly from that obtained using visual estimation (0.99). The confusion matrix for staple foods showed variation in the distribution of errors, indicating that the errors in the AI estimation were biased toward the case of many leftovers. CONCLUSIONS AI estimation measures liquid food intake in hospitals more precisely than visual estimation, but its accuracy in estimating staple food leftovers requires improvement.

Comparison of visual estimation and line point intercept vegetation survey methods on annual grass-invaded rangelands of Wyoming

Scientists and natural resource managers require suitable vegetation survey methods to assess the success of rangeland restoration projects. Visual estimation and point intercept methods are commonly used to evaluate vegetation cover. This study compared the performance of one visual (quadrat-based) and two line point intercept (LPI – canopy and basal) methods to assess biodiversity, cover, and to estimate biomass production on sites invaded by introduced annual grasses across Wyoming, USA. Greater species richness and higher Shannon index values were measured in quadrats, while introduced annual and native perennial graminoid cover values were higher in LPI canopy in general. Overall, these outcomes indicate quadrats as the most suitable survey method when biodiversity monitoring is the primary objective, while suggesting LPI canopy when monitoring vegetation cover is prioritized. Finally, our regression models indicated quadrat-based estimates as the most reliable to predict introduced annual and native perennial graminoid biomass.

Validation of visual estimation of neonatal jaundice in low-income and middle-income countries: a multicentre observational cohort study

ObjectiveDetermine the sensitivity and specificity of neonatal jaundice visual estimation by primary healthcare workers (PHWs) and physicians as predictors of hyperbilirubinaemia.DesignMulticentre observational cohort study.SettingHospitals in Chandigarh and Delhi, India; Dhaka, Bangladesh; Durban, South Africa; Kumasi, Ghana; La Paz, Bolivia.ParticipantsNeonates aged 1–20 days (n=2642) who presented to hospitals for evaluation of acute illness. Infants referred for any reason from another health facility or those needing immediate cardiopulmonary resuscitation were excluded.Outcome measuresInfants were evaluated for distribution (head, trunk, distal extremities) and degree (mild, moderate, severe) of jaundice by PHWs and physicians. Serum bilirubin level was determined for infants with jaundice, and analyses of sensitivity and specificity of visual estimations of jaundice used bilirubin thresholds of >260 µmol/L (need for phototherapy) and >340 µmol/L (need for emergency intervention in at-risk and preterm babies).Results1241 (47.0%) neonates had jaundice. High sensitivity for detecting neonates with serum bilirubin >340 µmol/L was found for ‘any jaundice of the distal extremities (palms or soles) OR deep jaundice of the trunk or head’ for both PHWs (89%–100%) and physicians (81%–100%) across study sites; specificity was more variable. ‘Any jaundice of the distal extremities’ identified by PHWs and physicians had sensitivity of 71%–100% and specificity of 55%–95%, excluding La Paz. For the bilirubin threshold >260 µmol/L, ‘any jaundice of the distal extremities OR deep jaundice of the trunk or head’ had the highest sensitivity across sites (PHWs: 58%–93%, physicians: 55%–98%).ConclusionsIn settings where serum bilirubin cannot be measured, neonates with any jaundice on the distal extremities should be referred to a hospital for evaluation and management, where delays in serum bilirubin measurement and appropriate treatment are anticipated following referral, the higher sensitivity sign, any jaundice on the distal extremities or deep jaundice of the trunk or head, may be preferred.

Importance of Visual Estimation of Coronary Artery Stenoses and Use of Functional Evaluation for Appropriate Guidance of Coronary Revascularization—Multiple Operator Evaluation

Background: Visual estimation (VE) of coronary stenoses is the first step during invasive coronary angiography. The aim of this study was to evaluate the accuracy of VE together with invasive functional assessment (IFA) in defining the functional significance (FS) of coronary stenoses based on the opinion of multiple operators. Methods: Fourteen independent operators visually evaluated 133 coronary lesions which had a previous FFR measurement, indicating the degree of stenosis (DS), FS and IFA intention. We determined the accuracy of FS prediction using several scenarios combining individual and group decision, considering IFA as deemed necessary by the operator or only in intermediate lesions. Results: The accuracy of VE in predicting FS was largely variable between operators (average 66.1%); it improved significantly when IFA was used either as per operator’s opinion (86.3%; p < 0.0001) or only in intermediate DS (82.9; p < 0.0001). There was no significant difference between using IFA per observer’s opinion or only in intermediate DS lesions (p = 0.166). The poorest accuracy of VE for FS was obtained in intermediate DS lesions (59.1%). Conclusions: There are significant inter-observer differences in reporting the degree of DS, while the accuracy of VE prediction of FS is also largely dependent on the operator, and the worst performance is obtained in the evaluation of intermediate DS.

Beneficial Insects for Biological Pest Control in Greenhouse Cannabis Production

A greenhouse cannabis cultivation took place in Agriculture university of Athens in order to quantify the efficiency of beneficial insects as a main method of pest management. Cannabis plants grown in two greenhouses and beneficial insects were released only in one greenhouse as a means to investigate the efficacy against pests by the comparison with the control greenhouse. Measurements included the visual estimation of infestation, the recording of pest species and populations, and the comparison of infestations and yields amongst greenhouses. Our results indicate that beneficial insects could control pest populations up to 100%. Even though the environmental conditions were not optimal and consecutive pest infestations were observed throughout the duration of our study, the beneficial insects successfully managed the pest populations. In conclusion, biological control with beneficial insects is a very effective method for pest management in greenhouse cannabis production.

Phytoliths from some grasses (Poaceae) in arid lands of Xinjiang, China

Opal phytoliths, as silicon dioxide inclusions, are abundant in different parts of a plant. It is known that grasses are the most representative in this respect. The research of phytoliths, removed from 25 most common grass species in the arid and semiarid lands of the Junggar Basin and adjacent areas, has been undertaken. The visual estimation of diversity and variability of silica cells and identification of their morphological types (patterns) were also the aim of our research. Since the work is preliminary, we have emphasized on the visual estimation of silica cell variability and involved only the leaf blades in the analysis. Drawings of the revealed silica cells, characteristic of 25 species, are provided. The sig-nificant morphological diversity of phytoliths has been revealed, as well as their taxonomic similarity at the level of subfamilies. These data can be used for the identification of phytoliths from sediments.

Visual estimation of the force applied by another person

As observers, we believe that we can visually estimate the force that another person applies to an object. However, it is unclear what kind of cues we use to do this. We have focused on two types of visual change that occur when a person pushes an elastic object from above with his or her finger: the first one relates to a finger/hand shaking, known as an ``induced tremor'' and the second one relates to object deformation due to the application of force. This study shows that human observers of videos combine these two visual cues to estimate the force being applied by another person in the videos. Overall, the apparent force was stronger when the shaking was larger and when the magnitude of the deformation was larger, although systematic individual differences existed. The estimation of force was likely made in separate cognitive functions from the estimation of object softness. Estimating the force that another person applies seems to be done by perceptually interpreting both the actions of others and their external outcomes.

Comparisons among measurement approaches to determine blood loss with different environmental sanitary effects

Labour and delivery are hazardous processes for bleeding. It will have complications, and even death for maternal because of overdue recognition and identification of the early signs of bleeding. Postpartum hemorrhage is still a problem and a major cause of maternal death. The definition, bleeding threshold, and method of assessing blood loss are still being debated today. This literature review aimed to determine the methods that had been used in measuring the amount of blood loss and risk to the mother. The review used Science Direct and Pubmed electronic online databases with related keyword searches. Measurement of blood loss was categorized into several methods, including visual estimation, direct measurement, gravimetry, and photometry. These methods were explained and then compared with various similar methods. Several studies indicated that measuring the amount of blood loss still often used conventional methods, even though this method was very improper and tends to underestimate blood loss. Several accurate and objective quantification methods had been introduced, but they were still very complicated, expensive, and time-consuming. Based on the review results, various types of blood loss measurement methods were displayed, but they were still not practical, accurate, and reliable. Researchers continue to carry out improvement research in finding methods so that postpartum hemorrhage can be prevented and treated.

The effect of patient positioning on the accuracy and reliability of assessment of knee range of motion over a telemedicine platform

Introduction Diagnostic accuracy is one of the key considerations of telemedicine usage in orthopedic surgery. The aim of this study was to determine the optimal patient positioning to achieve accurate and reliable visual estimation of the knee joint range of motion over a digital platform for telemedicine. Methods A single volunteer was recruited to perform a total of 120 discrete and random knee range of motion angles in three patient positions: sitting, standing and supine. The patient image was broadcast over a digital platform to six raters. The raters recorded their visual estimation of each discrete knee flexion angle independently. After each discrete knee flexion angle, a physical goniometer was used to obtain the actual flexion angle of the knee. Results A total of 120 discrete measurements (40 measurements in the sitting, standing, and supine positions each) were recorded by each of the six raters. The supine position resulted in the highest intraclass correlation of 0.97 (95% confidence interval: 0.98, 0.99). All three patient positions achieved low absolute difference between the goniometer and the raters with 5.6 degrees (95% limits of agreement: −21.0, 9.8) in sitting, 2.7 degrees (95% limits of agreement: −10.1, 15.4), and 1.2 degrees (95% limits of agreement: −9.8, 12.3) in the supine position. The supine position had the highest accuracy and reliability. Discussion Visual estimation of the knee joint range of motion over telemedicine is clinically accurate and reliable. Patients should be assessed in a supine position to obtain the highest accuracy and reliability for visual estimation of the knee joint range of motion during telemedicine.