Mean Absolute Prediction Error Research Articles

Accuracy of Intraocular lens power calculation in pediatric traumatic cataract.

To evaluate the prediction error (PE) in pediatric traumatic cataract surgery involving primary or secondary intraocular lens implantation (IOL) and the factors affecting it. Retrospective data of unilateral traumatic cataract eyes of children aged ≤16 years were collected between February 2019 and March 2022 at a tertiary eye care hospital. Absolute PE was calculated by deducting the target refraction from the observed refraction at 6 weeks postsurgery following suture removal. Simulated PE was calculated in eyes with corneal scar replacing the affected eye keratometry (K) with the K of the fellow eye and standard K (44D) and was then compared with absolute PE. Fifty children with a mean age of 9.5 years (2-16 years) were included. Mean absolute PE was 1.63 ± 1.8D (0 to 9D). Absolute PE was not affected by the age at surgery, method of biometry, duration of injury, type of cataract surgery, position of IOL, and preoperative keratometry. The absolute PE was affected by axial length (AXL) being <1.5D in AXL of 22.73 ± 0.84 mm, <1.5-2D in AXL of 22.07 ± 0.61 mm, and >2D in AXL of 22.01 ± 0.74 mm (P = 0.039) in univariate analysis. In multivariate analysis, none of the factors affected the absolute PE. In 34 eyes with corneal scar, higher variability in PE was observed. The standard K resulted in greater simulated PE as compared to the affected eye average K in eyes with scar involving the visual axis. Absolute PE following pediatric traumatic cataract was studied. It was higher in shorter AXLs. In corneal scar involving the visual axis, using the fellow eye K yielded lesser simulated PE as compared to standard K.

Comparison of IOL power formulas in eyes with a flat cornea

PurposeTo assess and compare accuracy of intraocular lens (IOL) power calculation performed with old generation, new generation and artificial intelligence (AI)-based formulas in eyes with flat corneas. DesignPatients with a tomography-derived mean keratometry <42 D were retrospectively recruited among those who underwent uncomplicated phacoemulsification with intracapsular IOL implantation in two different tertiary care centers. Mean prediction error (ME), mean absolute prediction error (MAE) and incidence of MAE>0.25D were calculated for Barrett Universal II (BUII),EVO 2.0, Hoffer QST, Kane, Olsen-C, Pearl-DGS and SRK/T formulas. Linear correlation between MAE and axial length was also calculated for each formula. ResultsEighty (80) eyes with a mean keratometry of 41.4 ± 0.6 D (range 40.6–41.9 D) were recruited. The SRK/T significantly differed from both the Olsen-C (p = 0.022) and the BUII (p = 0.048) in ME. The EVO 2.0, the Hoffer QST, the Kane and the PEARL-DGS showed a significantly lower MAE compared to all other formulas (p < 0.001) and a significant lower incidence of MAE>0.25D (p < 0.001), MAE>0.50 D (p < 0.001) and MAE>1.0 D (0.002). ConclusionFormulas based on AI and on the theory of vergence show superior accuracy in IOL power calculation in corneas with low mean keratometry; their MAE is not correlated to axial length.

Accuracy of Modern Intraocular Lens Formulas in Highly Myopic Eyes Implanted with Plate-Haptic Intraocular Lenses

PurposeTo evaluate the predictive accuracy of modern intraocular lens (IOL) formulas and axial length (AL) adjusted traditional IOL formulas, including Wang-Koch and Cooke-modified AL (CMAL) method, in long eyes with plate-haptic IOLs, and to compare refractive prediction error variances with C-loop IOLs. DesignRetrospective consecutive case series study. MethodsData from 391 eyes with Zeiss 509M and 302 eyes with Alcon SN6CWS implants in highly myopic patients, following cataract surgery from January 2019 to November 2023, were collected. One eye per patient was selected. Predictive outcomes of 15 modern formulas (Barrett Universal II (BU II), Cooke K6 (K6), Emmetropia Verifying Optical (EVO) 2.0, Hoffer-QST, Kane, Karmona, Ladas AI, Naeser 2, Olsen, Pearl-DGS, Radial Basis Function (RBF) 3.0, T2, VRF-G, Zhu-Lu, and Z-Calc) and 4 traditional IOL formulas (Haigis, Hoffer Q, Holladay 1, and SRK/T) with AL adjusted methods, were evaluated. The mean prediction error, mean absolute prediction error (MAE), root-mean-square absolute prediction error (RMSAE) and the proportions of eyes with PEs within ±0.25 Diopter (D), ±0.50 D, ±0.75 D and ±1.00 D were analyzed. Top 10 RMSAE-ranked formulas underwent further subgroup analysis based on AL, anterior chamber depth (ACD), and keratometry (K). ResultsFor the 509M group, RMSAE ranking for the top 10 IOL formulas were the RBF 3.0 (0.432), Zhu-Lu (0.436), Olsen (0.436), EVO 2.0 (0.437), Pearl-DGS (0.447), K6 (0.452), VRF-G (0.454), Naeser 2 (0.464), Haigis-CMAL (0.465) and Karmona (0.477). Karmona and Naeser 2 showed poorer performance in the extremely long AL and steep K subgroups, respectively (p≤0.042). Haigis-CMAL accuracy was significantly lower in shallow ACD and flat K subgroups (p≤0.045). The SN6CWS group showed significantly lower MAE and RMSAE compared to the 509M group for the BU II, EVO 2.0, Hoffer-QST, Kane, Pearl-DGS, and Zhu-Lu formulas (p≤0.024). ConclusionsIn long eyes with plate-haptic IOLs, RBF 3.0 performed best, closely followed by Zhu-Lu, Olsen, and EVO 2.0; Karmona and Naeser 2 are discouraged for extreme AL and steep K conditions, respectively; Haigis-CMAL is not suggested for shallow ACD and flat K cases. Refractive outcomes in eyes implanted with a C-loop design IOL were more accurate than for those implanted with a plate-haptic design, for most tested formulas.

An ensemble learning strategy for multi‐source hydrogen embrittlement data by introducing missing information

AbstractAccurately and quickly predicting hydrogen embrittlement performance is critical for the service of metal materials. However, due to multi‐source heterogeneity, existing hydrogen embrittlement data are missing, making it impractical to train reliable machine learning models. In this study, we proposed an ensemble learning training strategy for missing data based on the Adaboost algorithm. This method introduced a mask matrix with missing data and enabled each round of training to generate sub‐datasets, considering missing value information. The strategy first trained a subset of features based on the existing dataset and a selected method and continuously focused on the combination of features with the highest error for iterative training, where the mask matrix of the missing data was used as the input to fit the weights of each base learner using a neural network. Compared with directly modeling on highly sparse data, the predictive ability of this strategy was significantly improved by approximately 20%. In addition, in the testing of new samples, the predicted mean absolute error of the new model was successfully reduced from 0.2 to 0.09. This strategy offers good adaptability to the hydrogen embrittlement sensitivity of different sizes and can avoid interference from feature importance caused by filling data.

Daptomycin Exposure Prediction With a Limited Sampling Strategy.

Daptomycin is a cyclic lipopeptide antibiotic used to treat serious infectious endocarditis caused by Staphylococcus aureus. The pharmacodynamic parameter correlating best with efficacy is the ratio of the estimated area under the concentration (AUC0-24)-time curve to the minimum inhibitory concentration. The aim of the study is to develop a limited sampling strategy to estimate AUC0-24 using a reduced number of samples. Sixty-eight daptomycin AUC0-24 values were calculated for 50 White patients who underwent treatment for at least 5 consecutive days. Plasma concentrations were detected using a validated high-performance liquid chromatography-tandem mass spectrometry analytical method, with daptomycin-d5 as an internal standard. Multiple regression was used to evaluate the ability of 2 concentration-time points to predict the AUC0-24 calculated from the entire pharmacokinetic profile. Prediction bias was calculated as the mean prediction error, whereas prediction precision was estimated as the mean absolute prediction error. The development and validation datasets comprised 40 and 10 randomly selected patients, respectively. The AUC0-24 (mg*h/L) was best estimated using the daptomycin trough concentration and plasma concentrations detected 2 hours after dosing. We calculated a mean prediction error of 1.6 (95% confidence interval, -10.7 to 10.9) and a mean absolute prediction error of 11.8 (95% confidence interval, 5.3-18.3), with 73% of prediction errors within ±15%. An equation was developed to estimate daptomycin exposure (AUC0-24), offering clinical applicability and utility in generating personalized dosing regimens, especially for individuals at high risk of treatment failure or delayed response.

Exploring clozapine pharmacokinetics in Tunisian schizophrenic patients: A population-based modelling approach investigating the impact of genetic and non-genetic variables.

Clozapine is characterized by a large within- and between-patient variability in its pharmacokinetics, attributed to non-genetic and genetic factors. A cross-sectional analysis of clozapine trough concentration (Clz C0) issued from Tunisian schizophrenic patients was collected and analysed using a nonparametric modelling approach. We assessed the impact of demographic covariates (age, weight and sex), patient's habits (smoking status, alcohol and caffeine intake) and the genetic factors (CYP1A2*1C, CYP1A2*1F and CYP2C19*2 polymorphisms) on each pharmacokinetic parameter. An external validation of this pharmacokinetic model using an independent data set was performed. Fit goodness between observed- and individual-predicted data was evaluated using the mean prediction error (% MPE), the mean absolute prediction error (% MAPE) as a measure of bias, and the root mean squared error (% RMSE) as a measure of precision. Sixty-three CLz C0 values issued from 51 schizophrenic patients were assessed in this study and divided into building and validation groups. CYP1A2*1F polymorphism and smoking status were the only covariates significantly associated with clozapine clearance. Precision parameters were as follows: 1.02%, 0.95% and 22.4%, respectively, for % MPE, % MAPE and % RMSE. We developed and validated an accurate pharmacokinetic model able to predict Clz C0 in Tunisian schizophrenic patients using the two parameters CYP1A2*1F polymorphism and smoking.

Predictive models for short-term load forecasting in the UK's electrical grid.

There are global efforts to deploy Electric Vehicles (EVs) because of the role they promise to play in energy transition. These efforts underscore the e-mobility paradigm, representing an interplay between renewable energy resources, smart technologies, and networked transportation. However, there are concerns that these initiatives could burden the electricity grid due to increased demand. Hence, the need for accurate short-term load forecasting is pivotal for the efficient planning, operation, and control of the grid and associated power systems. This study presents robust models for forecasting half-hourly and hourly loads in the UK's power system. The work leverages machine learning techniques such as Support Vector Regression (SVR), Artificial Neural Networks (ANN), and Gaussian Process Regression (GPR) to develop robust prediction models using the net imports dataset from 2010 to 2020. The models were evaluated based on metrics like Root Mean Square Error (RMSE), Mean Absolute Prediction Error (MAPE), Mean Absolute Deviation (MAD), and the Correlation of Determination (R2). For half-hourly forecasts, SVR performed best with an R-value of 99.85%, followed closely by GPR and ANN. But, for hourly forecasts, ANN led with an R-value of 99.71%. The findings affirm the reliability and precision of machine learning methods in short-term load forecasting, particularly highlighting the superior accuracy of the SVR model for half-hourly forecasts and the ANN model for hourly forecasts.

Retrospective assessment of accuracy of nine intraocular lens power calculation formulae in eyes with axial myopia.

To compare the accuracy of nine conventional and newer-generation formulae in calculating intraocular lens power in eyes with axial myopia. Tertiary eye care center, Bengaluru, India. Retrospective cross-sectional, comparative study conducted in India. Patients undergoing uneventful phacoemulsification in eyes with axial length >26 mm were included. Preoperative biometry was done using Lenstar LS 900 (Haag-Streit AG, Switzerland). Single eye of patients undergoing bilateral implantation was randomly selected. Optimized lens constants were used to calculate the predicted postoperative refraction of each formula, which was then compared with the actual refractive outcomes to give the prediction errors, following which subgroup analysis was performed. The Kane formula, Barrett universal II, Emmetropia Verifying Optical (EVO) 2.0, Hill Radial Basis Function (Hill RBF) 3.0, Olsen formula, along with Wang Koch-adjusted four formulae, that is, Sanders Retzlaff Kraff/Theoretical (SRK/T), Holladay 1, Haigis, and Hoffer Q formula, were compared for intraocular lens power calculations. One hundred and sixty-five eyes that fulfilled all the inclusion criteria were studied. Hill RBF 3.0 had the lowest mean and median absolute prediction errors (0.355 and 0.275, respectively) compared to all formulas. In subgroup analysis (26-28, >28-30, and >30 mm), significant difference was seen only in extremely long eyes (>30 mm). The Hill RBF 3.0 formula generated the maximum percentage of eyes with refractive errors within ±0.25, ±0.5, ±0.75, and ±1 D (46%, 76.2%, 89.9%, and 95.8%, respectively). This is the first study evaluating all the formulas exclusively in the myopic eyes. Hill RBF 3 was found to be superior in accuracy to all other formulas.

The effect of corneal power on the accuracy of 14 IOL power formulas

BackgroundThis study evaluates the impact of corneal power on the accuracy of 14 newer intraocular lens (IOL) calculation formulas in cataract surgery. The aim is to assess how these formulas perform across different corneal curvature ranges, thereby guiding more precise IOL selection.MethodsIn this retrospective case series, 336 eyes from 336 patients who underwent cataract surgery were studied. The cohort was divided into three groups according to preoperative corneal power. Key metrics analyzed included mean prediction error (PE), standard deviation of PE (SD), mean absolute prediction error (MAE), median absolute error (MedAE), and the percentage of eyes with PE within ± 0.25 D, 0.50 D, ± 0.75 D, ± 1.00 D and ± 2.00 D.ResultsIn the flat K group (Km < 43 D), VRF-G, Emmetropia Verifying Optical Version 2.0 (EVO2.0), Kane, and Hoffer QST demonstrated lower SDs (± 0.373D, ± 0.379D, ± 0.380D, ± 0.418D, respectively) compared to the VRF formula (all P < 0.05). EVO2.0 and K6 showed significantly different SDs compared to Barrett Universal II (BUII) (all P < 0.02). In the medium K group (43 D ≤ Km < 46 D), VRF-G, BUII, Karmona, K6, EVO2.0, Kane, and Pearl-DGS recorded lower MAEs (0.307D to 0.320D) than Olsen (OLCR) and Castrop (all P < 0.03), with RBF3.0 having the second lowest MAE (0.309D), significantly lower than VRF and Olsen (OLCR) (all P < 0.05). In the steep K group (Km ≥ 46D), RBF3.0, K6, and Kane achieved significantly lower MAEs (0.279D, 0.290D, 0.291D, respectively) than Castrop (all P < 0.001).ConclusionsThe study highlights the varying accuracy of newer IOL formulas based on corneal power. VRF-G, EVO2.0, Kane, K6, and Hoffer QST are highly accurate for flat corneas, while VRF-G, RBF3.0, BUII, Karmona, K6, EVO2.0, Kane, and Pearl-DGS are recommended for medium K corneas. In steep corneas, RBF3.0, K6, and Kane show superior performance.

Short-term prediction of UT1-UTC and LOD via Dynamic Mode Decomposition and combination of least-squares and vector autoregressive model

Abstract This study presents a short-term forecast of UT1-UTC and LOD using two methods, i.e. Dynamic Mode Decomposition (DMD) and combination of Least-Squares and Vector Autoregression (LS+VAR). The prediction experiments were performed separately for yearly time spans, 2018-2022. The prediction procedure started on January 1 and ended on December 31, with 7-day shifts between subsequent 30-day forecasts. Atmospheric Angular Momentum data (AAM) were used as an auxiliary time series to potentially improve the prediction accuracy of UT1-UTC and LOD in LS+VAR procedure. An experiment was also conducted with and without elimination of effect of zonal tides from UT1-UTC and LOD time series. Two approaches to using the best steering parameters for the methods were applied:. First, an adaptive approach, which observes the rule that before every single forecast, a preliminary one must be performed on the pre-selected sets of parameters, and the one with the smallest prediction error is then used for the final prediction; and second, an averaged approach, whereby several forecasts are made with different sets of parameters (the same parameters as in adaptive approach) and the final values are calculated as the averages of these predictions. Depending on the method and data combination mean absolute prediction errors (MAPE) for UT1-UTC vary from 0.63 ms to 1.43ms for the 10th day and from 3.07 ms to 8.05ms for the 30th day of the forecast. Corresponding values for LOD vary from 0.110 ms to 0.245 ms for the 10th day and from 0.148 ms to 0.325 ms for the 30th day.

The LISA-PPV Formula: An Ensemble Artificial Intelligence-Based Thick Intraocular Lens Calculation Formula for Vitrectomized Eyes

PurposeTo investigate the relationship between effective lens position (ELP) and patient characteristics, and to further develop a new intraocular lens (IOL) calculation formula for cataract patients with previous pars plana vitrectomy (PPV). DesignCross-sectional study. MethodsA total of 2793 age-related cataract patients (group 1) and 915 post-PPV cataract patients (group 2) who underwent phacoemulsification with IOL implantation were included. The ELP of two groups was compared and the association between ELP and patient characteristics was further evaluated using standardized multivariate regression coefficients. An ensemble artificial intelligence-based ELP prediction model was developed using a training set of 810 vitrectomized eyes, and a thick-lens IOL formula (LISA-PPV) was constructed and compared with seven existing formulas on an external multi-center testing set of 105 eyes. ResultsCompared to eyes with age-related cataract, vitrectomized eyes showed a similar ELP distribution (P = 0.19), but different standardized coefficients of preoperative biometry for ELP. The standardized coefficients also varied with the type of vitreous tamponade, history of scleral buckling, and ciliary sulcus IOL implantation. The LISA-PPV formula showed the lowest mean and median absolute prediction error (MAE: 0.63 D; MedAE: 0.44 D), and the highest percentages of eyes within ± 0.5 D of prediction error (57.14%) in the testing dataset. ConclusionsThe ELP prediction required optimization specifically for vitrectomized eyes based on their biometric and surgical characteristics. The LISA-PPV formula is a useful and accurate tool for determining IOL power in cataract patients with previous PPV (available at http://ppv-iolcalculator.com/).

IOL Calculation using the ESCRS Online Calculator in Pediatric Eyes Undergoing Lens Extraction.

To evaluate the ESCRS online calculator for intraocular lens (IOL) calculation in children undergoing lens extraction and primary IOL implantation. Department of Ophthalmology, Goethe University Frankfurt, Germany. Retrospective, consecutive case series. We included eyes that received phacoemulsification and IOL implantation (Acrysof SN60AT, Alcon, Fort Worth, Tx, USA) due to congenital or juvenile cataract. We compared the mean prediction error (MPE), mean and median absolute prediction error (MAE, MedAE) of formulas provided by the recently introduced online calculator provided by the European Society of Cataract and Refractive Surgeons (ESCRS) to the SRK/T formula, as well as the number of eyes within ±0.5, ±1.0, ±2.0 diopters (D) of target refraction. Postoperative spherical equivalent was measured by retinoscopy 4 to 12 weeks after surgery. Sixty eyes from forty-seven patients with a mean age of 6.5 years ± 3.2 met the inclusion criteria. Mean axial length was 22.27mm ± 1.19. Mean preoperative spherical equivalent (SE) was -0.25 D ± 3.78 and mean postoperative SE was 0.69 D ± 1.53. The MedAE was lowest in the SRK/T formula (0.56 D, ± 1.03) performed significantly better (p = 0.037) than Hoffer QST and Kane, followed by BUII (0.64D, ± 0.92), Pearl DGS (0.65D, ± 0.94), EVO (0.69D, ± 0.94), Hoffer QST (0.75D, ± 0.99), and Kane (0.78D, ± 0.99). All of those were significantly above zero (p < 0.001). Forty-one eyes received an intraoperative optic capture (68%). When excluding eyes that did not receive intraoperative optic capture (n=19; 32%) the MedAE was shown to be lower. Using modern IOL calculation formulas provided by the ESCRS calculator provides good refractive predictability and compares for most of the formulas to the results with SRK/T. Additionally the formulas seem to anticipate the postoperative refraction better for eyes that receive a posterior optic capture.

Predictive accuracy of Barrett TK toric calculator compared to measured posterior corneal astigmatism using Pentacam in toric IOL power calculation.

To compare the predictive accuracy of Barrett total keratometry (Barrett TK) toric calculator with the measured posterior corneal astigmatism (PCA) by using Pentacam in toric intraocular lens (IOL) power calculation. A prospective analysis was done on 118 eyes requiring toric IOL power implantation. The absolute prediction error of the Barrett TK toric calculator and the measured PCA in the online toric calculator were assessed and compared to the standard Alcon toric calculator (with Barrett toric calculator incorporated). The mean absolute prediction error of the online toric calculator (0.32 D) (with Barrett toric calculator incorporated), Barrett TK Toric (0.34 D), and measured PCA of Pentacam in Barrett toric calculator (0.33 D) were found to be similar with no statistically significant difference. Subanalysis in eyes with with-the-rule astigmatism, against-the-rule astigmatism, and oblique astigmatism showed similar results. Alpins analysis showed that all three methods overcorrected corneal astigmatism. The Barrett TK toric calculator and the measured PCA of Pentacam in the Barrett toric calculator have similar predictive accuracy to the online toric calculator (with Barrett toric calculator incorporated).

Forecasting seasonal influenza activity in Canada-Comparing seasonal Auto-Regressive integrated moving average and artificial neural network approaches for public health preparedness.

Public health preparedness is based on timely and accurate information. Time series forecasting using disease surveillance data is an important aspect of preparedness. This study compared two approaches of time series forecasting: seasonal auto-regressive integrated moving average (SARIMA) modelling and the artificial neural network (ANN) algorithm. The goal was to model weekly seasonal influenza activity in Canada using SARIMA and compares its predictive accuracy, based on root mean square prediction error (RMSE) and mean absolute prediction error (MAE), to that of an ANN. An initial SARIMA model was fit using automated model selection by minimizing the Akaike information criterion (AIC). Further inspection of the autocorrelation function and partial autocorrelation function led to 'manual' model improvements. ANNs were trained iteratively, using an automated process to minimize the RMSE and MAE. A total of 378, 462 cases of influenza was reported in Canada from the 2010-2011 influenza season to the end of the 2019-2020 influenza season, with an average yearly incidence risk of 20.02 per 100,000 population. Automated SARIMA modelling was the better method in terms of forecasting accuracy (per RMSE and MAE). However, the ANN correctly predicted the peak week of disease incidence while the other models did not. Both the ANN and SARIMA models have shown to be capable tools in forecasting seasonal influenza activity in Canada. It was shown that applying both in tandem is beneficial, SARIMA better forecasted overall incidence while ANN correctly predicted the peak week.

Exhaust Gas Temperature Prediction of Aero-Engine via Enhanced Scale-Aware Efficient Transformer

This research introduces the Enhanced Scale-Aware efficient Transformer (ESAE-Transformer), a novel and advanced model dedicated to predicting Exhaust Gas Temperature (EGT). The ESAE-Transformer merges the Multi-Head ProbSparse Attention mechanism with the established Transformer architecture, significantly optimizing computational efficiency and effectively discerning key temporal patterns. The incorporation of the Multi-Scale Feature Aggregation Module (MSFAM) further refines 2 s input and output timeframe. A detailed investigation into the feature dimensionality was undertaken, leading to an optimized configuration of the model, thereby improving its overall performance. The efficacy of the ESAE-Transformer was rigorously evaluated through an exhaustive ablation study, focusing on the contribution of each constituent module. The findings showcase a mean absolute prediction error of 3.47∘R, demonstrating strong alignment with real-world environmental scenarios and confirming the model’s accuracy and relevance. The ESAE-Transformer not only excels in predictive accuracy but also sheds light on the underlying physical processes, thus enhancing its practical application in real-world settings. The model stands out as a robust tool for critical parameter prediction in aero-engine systems, paving the way for future advancements in engine prognostics and diagnostics.

Virtual Meter with Flow Pattern Recognition Using Deep Learning Neural Networks: Experiments and Analyses

Summary Operators often require real-time measurement of fluid flow rates in each well of their fields, which allows better control of production. However, petroleum is a complex multiphase mixture composed of water, gas, oil, and other sediments, which makes its flow challenging to measure and monitor. A critical issue is how the liquid component interacts with the gaseous phase, also known as the flow pattern. For example, sometimes liquids can accumulate in the lower part of the pipeline and block the flow completely, causing a gas pressure buildup that can lead to unstable flow regimes or even accidents (blowouts). On the other hand, some flow patterns can also facilitate sediment deposition, leading to obstructions and reduced production. Thus, this work aims to show that deep neural networks can act as a virtual flowmeter (VFM) using only a history of production, pressure, and temperature telemetry, accurately estimating the flow of all fluids in real time. In addition, these networks can also use the same input data to detect and recognize flow patterns that can harm the regular operation of the wells, allowing greater control without requiring additional costs or the installation of any new equipment. To demonstrate the feasibility of this approach and provide data to train the neural networks, a water-air loop was constructed to resemble an oil well. This setup featured inclined and vertical transparent pipes to generate and observe different flow patterns and sensors to record temperature, pressure, and volumetric flow rates. The results show that deep neural networks achieved up to 98% accuracy in flow pattern prediction and 1% mean absolute prediction error (MAPE) in flow rates, highlighting the capability of this technique to provide crucial insights into the behavior of multiphase flow in risers and pipelines.

Correction of axial length measurement error by IOLMaster 700 could improve refractive prediction accuracy in silicone oil-filled eyes.

To determine whether correcting the axial length (AL) measurement error of the IOLMaster 700 could improve the refractive prediction accuracy in silicone oil-filled eyes. This study included 265 cataract patients (265 eyes) with silicone oil tamponade who were scheduled for phacoemulsification with intraocular lens (IOL) implantation. The performances of various formulas, including Barrett Universal II, Emmetropia Verifying Optical, Hoffer-QST, Kane, Ladas Super Formula, Pearl-DGS, Radial Basis Function and traditional formulas (Haigis, Hoffer Q, Holladay 1 and SRK/T), were evaluated. The refractive prediction errors (PE) calculated with measured AL (ALmeas ) and corrected AL with silicone oil adjustment (SOAL ) were compared. Subgroup analysis was performed based on the ALmeas (<23 mm; 23-26 mm; ≥26 mm). Using SOAL significantly reduced the hyperopic PE of formulas when compared to ALmeas (-0.05 to 0.17 D vs 0.15 to 0.38 D, p < 0.001). After applying AL correction, all formulas showed a lower mean absolute PE (0.47-0.57 D vs 0.50-0.69 D). The percentage of eyes within ±1.0 D of PE increased from 84.91%-88.68% to 89.81%-91.32% for new formulas and from 78.11%-83.40% to 85.66%-88.68% for traditional formulas, with the use of SOAL . Subgroup analysis showed that the majority of formulas with SOAL in prediction accuracy for eyes with an AL ≥26 mm (p < 0.05). The refractive prediction accuracy in silicone oil-filled eyes was improved by correcting the AL measurement error of the IOLMaster 700, especially for long eyes.

Combination of Artificial Neural Networks and Principal Component Analysis for the Simultaneous Quantification of Dyes in Multi-Component Aqueous Mixtures

Dyes are organic compounds capable of transmitting their color to materials, which is why they are widely used, for example, in textile fibers, leather, paper, plastic, and the food industry. In the dying process, measuring the dye’s content is extremely important to evaluate the process efficiency and minimize the dye’s discharge in wastewater, but most of the time, dyes are present in multi-component mixtures; hence, quantification by spectrophotometric methods presents a great challenge because the signal obtained in the measurement overlaps the components in the mixture. In order to overcome this issue, the use of the high-performance liquid chromatography (HPLC) method is recommended; however, it has the disadvantage of being an expensive technique, complex, and requiring excessive sample preparation. In recent years, some direct spectrophotometric methods based on multivariate regression algorithms for the quantification of dyes in bicomponent mixtures have been reported. This study presents a new framework that uses a combined ANN and principal component analysis (PCA) model for the determination of the concentration of three dyes in aqueous mixtures: Tartrazine (TZ), Amaranth Red (AR), and Blue 1 CFC (B1) dyes. The PCA–ANN model was trained and validated with ternary mixture samples of TZ, AR, and B1, and with known different compositions, spectra absorbance samples were measured in a UV-Vis spectrophotometer at wavelengths between 350–700 nm with intervals of 1 nm. The PCA–ANN model showed a mean absolute prediction error and correlation coefficient (r2) of less than 1% and greater than 0.99, respectively. The results demonstrate that the PCA–ANN model is a quick and highly accurate alternative in the simultaneous determination of dyes in ternary aqueous mixtures.

Effect of Posterior Keratometry on the Accuracy of 10 Intraocular Lens Calculation Formulas: Standard Keratometry versus Total Keratometry.

To investigate the effect of posterior keratometry (PK) on the accuracy of 10 intraocular lens (IOL) power calculation formulas using standard keratometry (K) and total keratometry (TK). This is a retrospective consecutive case-series study. The IOL power was calculated using K and TK measured by IOLMaster 700 in 6 new-generation formulas (Barrett Universal II, Emmetropia Verifying Optical (EVO) 2.0, RBF Calculator 3.0, Hoffer QST, Kane, and Ladas Super Formula) and 4 traditional formulas (Haigis, Hoffer Q, Holladay 1, and SRK/T). The arithmetic prediction error (PE) and mean absolute PE (MAE) were evaluated. The locally-weighted scatterplot smoothing was performed to assess the relationship between PE and PK. A total of 576 patients (576 eyes) who underwent cataract surgery were included. Compared with using K, all formulas using TK showed a hyperopic shift in the whole group. Specifically, for eyes with PK exceeding -5.90 D, all formulas using TK exhibited a hyperopic shift (all P < 0.001), while eyes with PK less than -5.90 D showed a myopic shift (all P < 0.001). The MAE of new-generation formulas calculated with TK and K showed no statistical differences, while the MAE of traditional formulas with TK was larger (TK: 0.34 ~ 0.43 D; K: 0.33 ~ 0.42 D, all P < 0.05). The prediction bias of formulas with TK increased as PK deviated from -5.90 D. TK did not improve the prediction accuracy of new-generation formulas, and even performed worse in traditional formulas.

Non-destructive prediction of the total viable count (TVC) in Fujian oysters (Crassostrea angulata) based on the colorimetric sensor array

The total viable count (TVC) is a traditional method for evaluating seafood freshness. This study aimed to predict the TVC of Fujian oysters stored at 4℃ by using the colorimetric sensing array (CSA) technique combined with imaging and visible near-infrared (Vis-NIR) information channels. An optimal quantitative prediction model based on partial least squares (PLS) was established by combining feature variable selection with a data fusion strategy. All three data-fusion strategies showed better predictive performance than the individual datasets. The high-level fusion strategy performed the best, with a root mean square error of prediction (RMSEP) of 0.17, mean absolute error of prediction (MAEP) of 0.14, coefficient of determination of prediction (Rp2) of 0.9822, and residual prediction deviation (RPD) of 7.37. The experimental results showed that CSA combined with multiple detection channels could be effectively used to predict the TVC of Fujian oysters. In this study, a reliable and comprehensive evaluation model was established to achieve a rapid and accurate prediction of oyster freshness.