Computer-based Predictive Model Research Articles

Explainable prediction model for the human papillomavirus status in patients with oropharyngeal squamous cell carcinoma using CNN on CT images

Several studies have emphasised how positive and negative human papillomavirus (HPV+ and HPV−, respectively) oropharyngeal squamous cell carcinoma (OPSCC) has distinct molecular profiles, tumor characteristics, and disease outcomes. Different radiomics-based prediction models have been proposed, by also using innovative techniques such as Convolutional Neural Networks (CNNs). Although some of these models reached encouraging predictive performances, there evidence explaining the role of radiomic features in achieving a specific outcome is scarce. In this paper, we propose some preliminary results related to an explainable CNN-based model to predict HPV status in OPSCC patients. We extracted the Gross Tumor Volume (GTV) of pre-treatment CT images related to 499 patients (356 HPV+ and 143 HPV−) included into the OPC-Radiomics public dataset to train an end-to-end Inception-V3 CNN architecture. We also collected a multicentric dataset consisting of 92 patients (43 HPV+ , 49 HPV−), which was employed as an independent test set. Finally, we applied Gradient-weighted Class Activation Mapping (Grad-CAM) technique to highlight the most informative areas with respect to the predicted outcome. The proposed model reached an AUC value of 73.50% on the independent test. As a result of the Grad-CAM algorithm, the most informative areas related to the correctly classified HPV+ patients were located into the intratumoral area. Conversely, the most important areas referred to the tumor edges. Finally, since the proposed model provided additional information with respect to the accuracy of the classification given by the visualization of the areas of greatest interest for predictive purposes for each case examined, it could contribute to increase confidence in using computer-based predictive models in the actual clinical practice.

Feasibility analysis of wind energy potential along the coastline of Pakistan

Wind energy is gaining popularity, especially across the countries with strong and continuous wind currents. Pakistan is bestowed with multiple wind corridors with prospective energy extraction, mainly concentrated along the coastline. This work aims to analyze the effectiveness and feasibility of wind energy potential in Pakistan coastline by choosing four unique zones (Karachi, Ormara, Pasni, and Gawadar) through analysis and demonstration for a 50 MW wind farm. Four years of seasonal data is used for the selected zones, reflecting the wind effectiveness in wind speed, wind density, and wind directions. Computer-based predictive models and artificial neural networks are also implemented to forecast the wind effectiveness, primarily based on the observed wind seasonal data. The wind farm's feasibility is proposed through a comparative evaluation for various turbine designs considering both geographical, installation, operational and financial factors. The factors of energy output, economic feasibility, environmental impact, and fuel-saving are analyzed for multiple locations and turbine designs. Critical observations, data, and findings are presented and discussed in detail for all zones. Zone 1 has a highest capacity factor of 46% followed by 40.4% (Zone 2), 29.3% (Zone 3), 37.5% (Zone 4). The study revealed that Karachi is the best-suited location for a 50 MW wind farm with the highest wind speed and the lowest variation in wind direction annually. It also represented the highest internal rate of return, benefit-to-cost ratio, and annual saving with a minimum payback period (4.5–7.2 years) for various commercially available turbine designs.

Predicting survival of patients with spinal metastatic disease using PathFx 3.0 – A validation study of 668 patients in Sweden

IntroductionPathFx is a computer-based prediction model for estimating survival of patients with bone metastasis. The model has been validated in several studies, but this is the first validation using exclusively patients with spinal metastases. Research questionIs PathFx 3.0 a tool useful for predicting survival for patients with spinal metastatic disease? Material and methods668 patients (67% male, median age 67 years) presenting with spinal metastases at two university hospitals in Sweden 1991–2014 were included. Of those, the majority (82%, n ​= ​551) underwent surgery. Data on all patients was analyzed with PathFx version 3.0, generating a probability of survival at 1, 3, 6, 12, 18 and 24 months. The predictions were compared to real survival data and the precision in estimation was evaluated with Receiver-Operating Characteristic curve (ROC) analysis where the Area Under Curve (AUC) was calculated. Brier score and decision curve analyses were also assessed. ResultsThe AUC for 1-, 3-, 6- and 12 months survival predictions were 0.64 (95% CI 0.5–0.71), 0.71 (95% CI 0.67–0.75), 0.70 (95% CI 0.66–0.77) and 0.74 (95% CI 0.70–0.78). For 18- and 24 months survival the AUC were 0.74 (95% CI 0.69–0.78) and 0.76 (95% CI 0.72–0.81). The Brier scores were all 0.23 or lower depending on the estimated survival time. Discussion and conclusionPathFx 3.0 is a reasonably reliable tool for predicting survival in patients with spinal metastatic disease. As the PathFx computer model can be updated to reflect advancements in oncology, we suggest this type of model, rather than rigid point-based scoring systems, to be used for estimating survival in patients with metastatic spinal disease in the future.

Computational modeling – an approach to the development of blood grouping reagents

ABSTRACT Introduction: Blood group antigens are defined by an immune response that generates antibodies against a red blood cell molecule. Antibodies against these antigens can be associated with hemolytic transfusion reactions. However, difficulties can arise when developing antibodies against antigens through the use of peptide sequences alone. Three-dimensional representations (models) of the molecular structure of antigen-bearing proteins can provide valuable insights into tertiary structures and their consequent antigenicity. This can be achieved through predictive computational modeling to produce both structural and molecular dynamics models of blood group proteins. Areas covered: Authors discuss the use of molecular dynamic simulations on existing structures, as well as the use of computational modeling techniques in the development of protein models lacking preexisting data. Finally, the authors discuss specific examples of the use of computationally derived models of the MNS blood group system and its use in attempts to produce antibodies against MNS proteins. Expert opinion: Although in silico techniques have limitations, computer-based predictive models can inform the direction of research into blood group proteins. It is to be expected that as computer-based techniques grow more powerful these contributions will be even more significant.

Ultrasound echogenicity is associated with fatigue-induced failure in a cadaveric Achilles tendon model

Achilles tendon disorders are among the most difficult sports-related injuries to predict with current diagnostic tools. The purpose of this study was to identify a clinically useful marker for early tendon damage. We hypothesized that alterations in mean echogenicity are linked with changes in vitro tendon mechanics. To test our hypothesis, we harvested Achilles tendons from 10 fresh-frozen cadaveric feet and cyclically fatigued them using a universal test frame while we continuously acquired ultrasound images. Throughout this fatigue protocol, we applied 2 stress tests every 500 loading cycles to quantify changes in ultrasound imaging echogenicity. We continued this fatigue protocol until each tendon either failed completely or survived 150,000 cycles. Tendons that failed during the fatigue loading (6/10) underwent greater changes in mean echogenicity compared to tendons that did not fail (P = 0.031). These tendons that failed during fatigue loading demonstrated greater changes in mean echogenicity that surpassed 1.0%; whereas survivor tendons exhibited less than 0.5% changes in mean echogenicity. We found that changes in mean echogenicity measured with ultrasound increased proportionally with increased tendon damage. The magnitude of these changes was relatively small (<1.5% change in mean echogenicity) but may be an effective predictor of tendon failure. Mean echogenicity is a promising marker for quantifying fatigue damage in cadaveric Achilles tendons during a stress test. Although these changes cannot be detected with the naked eye, computer-based predictive models may effectively assess risk of tendon damage in physically active adults.

Development of a validated computer-based preoperative predictive model for pseudarthrosis with 91% accuracy in 336 adult spinal deformity patients.

OBJECTIVEPseudarthrosis can occur following adult spinal deformity (ASD) surgery and can lead to instrumentation failure, recurrent pain, and ultimately revision surgery. In addition, it is one of the most expensive complications of ASD surgery. Risk factors contributing to pseudarthrosis in ASD have been described; however, a preoperative model predicting the development of pseudarthrosis does not exist. The goal of this study was to create a preoperative predictive model for pseudarthrosis based on demographic, radiographic, and surgical factors.METHODSA retrospective review of a prospectively maintained, multicenter ASD database was conducted. Study inclusion criteria consisted of adult patients (age ≥ 18 years) with spinal deformity and surgery for the ASD. From among 82 variables assessed, 21 were used for model building after applying collinearity testing, redundancy, and univariable predictor importance ≥ 0.90. Variables included demographic data along with comorbidities, modifiable surgical variables, baseline coronal and sagittal radiographic parameters, and baseline scores for health-related quality of life measures. Patients groups were determined according to their Lenke radiographic fusion type at the 2-year follow-up: bilateral or unilateral fusion (union) or pseudarthrosis (nonunion). A decision tree was constructed, and internal validation was accomplished via bootstrapped training and testing data sets. Accuracy and the area under the receiver operating characteristic curve (AUC) were calculated to evaluate the model.RESULTSA total of 336 patients were included in the study (nonunion: 105, union: 231). The model was 91.3% accurate with an AUC of 0.94. From 82 initial variables, the top 21 covered a wide range of areas including preoperative alignment, comorbidities, patient demographics, and surgical use of graft material.CONCLUSIONSA model for predicting the development of pseudarthrosis at the 2-year follow-up was successfully created. This model is the first of its kind for complex predictive analytics in the development of pseudarthrosis for patients with ASD undergoing surgical correction and can aid in clinical decision-making for potential preventative strategies.

72 - Preoperative Use of a Validated Computer-Based Predictive Model for Patient Selection for Adult Spinal Deformity (ASD) Surgery has the Potential to Significantly Enhance QALYs Gained at Two Years Postop: Simulation in 234 ASD Patients

72 - Preoperative Use of a Validated Computer-Based Predictive Model for Patient Selection for Adult Spinal Deformity (ASD) Surgery has the Potential to Significantly Enhance QALYs Gained at Two Years Postop: Simulation in 234 ASD Patients

Predicting postoperative vomiting among orthopedic patients receiving patient-controlled epidural analgesia using SVM and LR.

Patient-controlled epidural analgesia (PCEA) has been applied to reduce postoperative pain in orthopedic surgical patients. Unfortunately, PCEA is occasionally accompanied by nausea and vomiting. The logistic regression (LR) model is widely used to predict vomiting, and recently support vector machines (SVM), a supervised machine learning method, has been used for classification and prediction. Unlike our previous work which compared Artificial Neural Networks (ANNs) with LR, this study uses a SVM-based predictive model to identify patients with high risk of vomiting during PCEA and comparing results with those derived from the LR-based model. From January to March 2007, data from 195 patients undergoing PCEA following orthopedic surgery were applied to develop two predictive models. 75% of the data were randomly selected for training, while the remainder was used for testing to validate predictive performance. The area under curve (AUC) was measured using the Receiver Operating Characteristic curve (ROC). The area under ROC curves of LR and SVM models were 0.734 and 0.929, respectively. A computer-based predictive model can be used to identify those who are at high risk for vomiting after PCEA, allowing for patient-specific therapeutic intervention or the use of alternative analgesic methods.

Phosphoproteomic Analyses of NRAS(G12) and NRAS(Q61) Mutant Melanocytes Reveal Increased CK2α Kinase Levels in NRAS(Q61) Mutant Cells

In melanoma, mutant and thereby constantly active neuroblastoma rat sarcoma (NRAS) affects 15-20% of tumors, contributing to tumor initiation, growth, invasion, and metastasis. Recent therapeutic approaches aim to mimic RAS extinction by interfering with critical signaling pathways downstream of the mutant protein. This study investigates the phosphoproteome of primary human melanocytes bearing mutations in the two hot spots of NRAS, NRAS(G12) and NRAS(Q61). Stable isotope labeling by amino acids in cell culture followed by mass spectrometry identified 14,155 spectra of 3,371 unique phosphopeptides mapping to 1,159 proteins (false discovery rate < 2%). Data revealed pronounced PI3K/AKT signaling in NRAS(G12V) mutant cells and pronounced mitogen-activated protein kinase (MAPK) signaling in NRAS(Q61L) variants. Computer-based prediction models for kinases involved, revealed that CK2α is significantly overrepresented in primary human melanocytes bearing NRAS(Q61L) mutations. Similar differences were found in human NRAS(Q61) mutant melanoma cell lines that were also more sensitive to pharmacologic CK2α inhibition compared with NRAS(G12) mutant cells. Furthermore, CK2α levels were pronounced in patient samples of NRAS(Q61) mutant melanoma at the mRNA and protein level. The preclinical findings of this study reveal that codon 12 and 61 mutant NRAS cells have distinct signaling characteristics that could allow for the development of more effective, mutation-specific treatment modalities.

Development of Validated Computer-based Preoperative Predictive Model for Proximal Junction Failure (PJF) or Clinically Significant PJK With 86% Accuracy Based on 510 ASD Patients With 2-year Follow-up.

A retrospective review of large, multicenter adult spinal deformity (ASD) database. The aim of this study was to build a model based on baseline demographic, radiographic, and surgical factors that can predict clinically significant proximal junctional kyphosis (PJK) and proximal junctional failure (PJF). PJF and PJK are significant complications and it remains unclear what are the specific drivers behind the development of either. There exists no predictive model that could potentially aid in the clinical decision making for adult patients undergoing deformity correction. Inclusion criteria: age ≥18 years, ASD, at least four levels fused. Variables included in the model were demographics, primary/revision, use of three-column osteotomy, upper-most instrumented vertebra (UIV)/lower-most instrumented vertebra (LIV) levels and UIV implant type (screw, hooks), number of levels fused, and baseline sagittal radiographs [pelvic tilt (PT), pelvic incidence and lumbar lordosis (PI-LL), thoracic kyphosis (TK), and sagittal vertical axis (SVA)]. PJK was defined as an increase from baseline of proximal junctional angle ≥20° with concomitant deterioration of at least one SRS-Schwab sagittal modifier grade from 6 weeks postop. PJF was defined as requiring revision for PJK. An ensemble of decision trees were constructed using the C5.0 algorithm with five different bootstrapped models, and internally validated via a 70 : 30 data split for training and testing. Accuracy and the area under a receiver operator characteristic curve (AUC) were calculated. Five hundred ten patients were included, with 357 for model training and 153 as testing targets (PJF: 37, PJK: 102). The overall model accuracy was 86.3% with an AUC of 0.89 indicating a good model fit. The seven strongest (importance ≥0.95) predictors were age, LIV, pre-operative SVA, UIV implant type, UIV, pre-operative PT, and pre-operative PI-LL. A successful model (86% accuracy, 0.89 AUC) was built predicting either PJF or clinically significant PJK. This model can set the groundwork for preop point of care decision making, risk stratification, and need for prophylactic strategies for patients undergoing ASD surgery. 3.

Development of Validated Computer-Based Preoperative Predictive Model for Proximal Junction Failure (PJF) or Clinically Significant PJK with 86% Accuracy based on 510 ASD Patients with 2-year Follow-up

IntroductionProximal Junction Failure (PJF) and Proximal Junction Kyphosis (PJK) are significant complications. It remains unclear what are the specific drivers behind the development of either. Th...

An Intelligent Knowledge-based Approach for the Management of Human Immunodeficiency Virus and Tuberculosis

Dealing with Human Immunodeficiency Virus (HIV) and Tuberculosis (TB) as co-infection is not easy. Stigmatization, rejection and discrimination of TB and HIV patients have become the major challenges in medical field which makes it very difficult for the health care professionals to communicate with their patients. The inherent complexities of medical practices make traditional approaches of diagnosis and predicting treatment outcome inappropriate. There is the need for the provision of a system to mitigate these effects. Till date, the research done on Fuzzy Cognitive Maps have not been able to effectively handle missing data in order to predict the patient health status in spite of many existing methods of dealing with missing data. In this paper, a new diagnostics predictive algorithm of fuzzy cognitive map was developed to improve the efficiency of the prediction of patient health status in case of HIV or TB. The system is an advancement of that proposed by Djam and Kimbi in 2011, with an additional factors on effective prediction measures. The diagnosis and treatment processes were modeled based on Domain Experts’ judgments. The approach proposed in this paper was validated using real patient data and the experimental results of the research show that the new proposed algorithm can effectively simulate the behavior of a medical expert and was successfully employed in designing computer-based diagnostic predictive models for effective patient management.

Radiation protection issues on preparedness and response for a severe nuclear accident: experiences of the Fukushima accident.

Radiation protection issues on preparedness and response for a severe nuclear accident are discussed in this paper based on the experiences following the accident at Fukushima Daiichi nuclear power plant. The criteria for use in nuclear emergencies in the Japanese emergency preparedness guide were based on the recommendations of International Commission of Radiological Protection (ICRP) Publications 60 and 63. Although the decision-making process for implementing protective actions relied heavily on computer-based predictive models prior to the accident, urgent protective actions, such as evacuation and sheltering, were implemented effectively based on the plant conditions. As there were no recommendations and criteria for long-term protective actions in the emergency preparedness guide, the recommendations of ICRP Publications 103, 109, and 111 were taken into consideration in determining the temporary relocation of inhabitants of heavily contaminated areas. These recommendations were very useful in deciding the emergency protective actions to take in the early stages of the Fukushima accident. However, some suggestions have been made for improving emergency preparedness and response in the early stages of a severe nuclear accident.

A computer-based cost prediction model for institutional building projects in Nigeria

Purpose – The purpose of this study was develop a computer-based cost prediction model for institutional building projects in Nigeria through the use of artificial neural network (ANN) technique. The back-propagation network learns by example and provides good prediction to novel cases. Design/methodology/approach – The input variables were derived from related works with modification and advices from professionals through a field survey. Two hundred and sixty completed project data were used for training and development of the ANN model. Back-propagation algorithm using the gradient descent delta learning rule with a learning coefficient of 0.4 was used. The input layer of the model comprised of nine variables; building height, compactness of building, construction duration, external wall area, gross floor area, number of floors, proportion of opening on external walls, location index and time index. Findings – Several multi-layer perceptron networks were developed with varying architecture from which the network 9-7-5-1 was selected. The performance of the model over the validation sample revealed that the model has a mean absolute per cent error of 5.4 per cent and average error of prediction of −2.5 per cent over the sample. The ANN model was considered to be effective for construction cost prediction. Research limitations/implications – The model may not be suitable for other building types because of the uniqueness of such facility even though significant difference is not anticipated for buildings such as commercial and residential. The models were evaluated based on the prediction errors; other means of evaluation were not used. Originality/value – The study thus provides a simple, yet effective means of predicting construction costs of institutional building projects in Nigeria using an ANN model.

Predicting postoperative vomiting for orthopedic patients receiving patient-controlled epidural analgesia with the application of an artificial neural network.

Patient-controlled epidural analgesia (PCEA) was used in many patients receiving orthopedic surgery to reduce postoperative pain but is accompanied with certain incidence of vomiting. Predictions of the vomiting event, however, were addressed by only a few authors using logistic regression (LR) models. Artificial neural networks (ANN) are pattern-recognition tools that can be used to detect complex patterns within data sets. The purpose of this study was to develop the ANN based predictive model to identify patients with high risk of vomiting during PCEA used. From January to March 2007, the PCEA records of 195 patients receiving PCEA after orthopedic surgery were used to develop the two predicting models. The ANN model had a largest area under curve (AUC) in receiver operating characteristic (ROC) curve. The areas under ROC curves of ANN and LR models were 0.900 and 0.761, respectively. The computer-based predictive model should be useful in increasing vigilance in those patients most at risk for vomiting while PCEA is used, allowing for patient-specific therapeutic intervention, or even in suggesting the use of alternative methods of analgesia.

Forecasting Construction and Demolition Waste Using Gene Expression Programming

AbstractAccurate forecasting of construction and demolition waste (CDW) generation could provide valuable information for the planning, design, and management of CDW at municipal levels. However, the lack of reliable forecasting approaches and historical records makes it difficult to predict the amount of CDW for a long- or short-term plan. To effectively tackle the CDW forecasting problem, a novel computer-based prediction model, gene expression programming (GEP), is introduced and tested. With the CDW and other data on predictor variables from the last two decades, the amount of CDW is forecasted in this study. Results and findings obtained from this research show that GEP is an effective model for predicting waste generation, with lower average forecasting error than the multiple linear model and the artificial neural network. Research issues related to model selection, training, and validation are also discussed in the paper.

Toward a predictive model for estimating dew point pressure in gas condensate systems

Dew-point pressure is one of the most important quantities for characterizing and successful prediction of the future performance of gas condensate reservoirs. The objective of this study is to present a reliable, computer-based predictive model for prediction of dew-point pressure in gas condensate reservoirs. An intelligent approach based on least square support vector machine (LSSVM) modeling was developed for this purpose. To this end, the model was developed and tested using a total set of 562 experimental data points from different retrograde gas condensate fluids covering a wide range of variables. Coupled simulated annealing (CSA) was employed for optimization of hyper-parameters of the model. The results showed that the developed model significantly outperforms all the existing methods and provide predictions in acceptable agreement with experimental data. In addition, it is shown that the proposed model is capable of simulating the actual physical trend of the dew-point pressure versus temperature for a constant composition fluid on the phase envelope.

Report From the Jerusalem Workshop on Lynch Syndrome-Hereditary Nonpolyposis Colorectal Cancer

A Workshop was held in Jerusalem, Israel, on October 26 and 27, 2009 to discuss the management of Lynch syndrome-hereditary nonpolyposis colorectal cancer (CRC), with the primary goal to develop consensus for the optimal management of this disease. A second goal was to identify areas of research with the potential to advance the clinical management of Lynch syndrome. The perspectives and recommendations from the workshop are meant to be a platform for discussion and deliberation. The Workshop was organized by Moshe Shike (Memorial Sloan Kettering Cancer Center, New York) and sponsored by The Colon Cancer Foundation. More details of each presentation are available in an on-line supplement.

Predicting hypotensive episodes during spinal anesthesia with the application of artificial neural networks

Hypotension is one of the most frequent adverse effects of spinal anesthesia. Several factors might be related to the occurrence of hypotension. Predictions of the hypotensive event, however, had been addressed by only a few authors using logistic regression (LR) models. Artificial neural networks (ANN) are pattern-recognition tools that can be used to detect complex patterns within data sets. The purpose of this study was to develop the ANN-based predictive model to identify patients with high risk of hypotension during spinal anesthesia. From September 2004 to December 2006, the anesthesia records of 1501 patients receiving surgery under spinal anesthesia were used to develop the ANN and LR models. By random selection 75% of data were used for training and the remaining 25% of data were used as test set for validating the predictive performance. Five senior anesthesiologists were asked to review the data of test set and to make predictions of hypotensive event during spinal anesthesia by clinical experience. The ANN model had a sensitivity of 75.9% and specificity of 76.0%. The LR model had a sensitivity of 68.1% and specificity of 73.5%. The area under receiver operating characteristic curves were 0.796 and 0.748. The ANN model performed significantly better than the LR model. The prediction of clinicians had the lowest sensitivity of 28.7%, 22.2%, 21.3%, 16.1%, and 36.1%, and specificity of 76.8%, 84.3%, 83.1%, 87.0%, and 64.0%. The computer-based predictive model should be useful in increasing vigilance in those patients most at risk for hypotension during spinal anesthesia, in allowing for patient-specific therapeutic intervention, or even in suggesting the use of alternative methods of anesthesia.

The Integrated Acute Systemic Toxicity Project (ACuteTox) for the Optimisation and Validation of Alternative In Vitro Tests

The ACuteTox project is designed to replace animal testing for acute systemic toxicity, as is widely used today for regulatory purposes, by using in vitro and in silico alternatives. In spite of the fact that earlier studies on acute systemic toxicity demonstrated a good correlation between in vitro basal cytotoxicity data (the 50% inhibitory concentration [IC50]) in human cell lines and rodent LD50 values, and an even better correlation between IC50 values and human lethal blood concentrations, very few non-animal tests have been accepted for general use. Therefore, the aim of the ACuteTox project is to adapt new testing strategies, for example, the implementation of new endpoints and new cell systems for toxicity screening, organ-specific models, metabolism-dependent toxicity, tissue absorption, distribution and excretion, and computer-based prediction models. A new database, AcuBase, containing descriptions and results of in vitro tests of the 97 reference chemicals, as well as the results of animal experimentation, and human acute toxicity data, will be generated within the framework of ACuteTox. Scientists from 13 European countries are working together and making efforts to find the most appropriate testing strategies for the prediction of human acute systemic toxicity, and also to select a robust in vitro test battery for cytotoxicity testing of chemicals.