scholarly journals Exploring Survival Models Associated with MCI to AD Conversion: A Machine Learning Approach

2019 ◽  
Author(s):  
Jorge Orozco-Sanchez ◽  
Victor Trevino ◽  
Emmanuel Martinez-Ledesma ◽  
Joshua Farber ◽  
Jose Tamez-Peña
Keyword(s):  
Machine Learning ◽  
Regression Models ◽  
Cox Regression ◽  
Early Stage ◽  
Final Analysis ◽  
Structural Mri ◽  
Survival Models ◽  
Mri Findings ◽  
Cox Models ◽  
Mri Features

AbstractSeveral studies have documented that structural MRI findings are associated with the presence of early-stage Alzheimer Disease (AD). However, the association of each MRI feature with the rate of conversion from mild cognitive impairment (MCI) to AD in a multivariate setting has not been studied fully. The objective of this work is the comprehensive exploration of four different machine learning (ML) strategies to build MRI-based multivariate Cox regression models. These models evaluated the association of MRI features with the time of MCI to clinical AD conversion. We used 442 MCI subjects from the Alzheimer’s disease Neuroimaging Initiative (ADNI) study. Each subject was described by 346 MRI features and time to AD conversion. Cox regression models then estimated the rate of conversion. Models were built using four ML methodologies in a cross-validation (CV) setting. All the ML methods returned successful Cox models with different CV performances. The best model exhibited a concordance index of 0.84 (95% CI: 0.82-0.86). The final analysis described the hazard ratios (HR) of the top ten MRI features associated with MCI to AD conversion. Our results suggest ML exploration is a viable strategy for building and analyzing survival models that predict subjects at risk of AD conversion.

scholarly journals A machine learning prediction model for waiting time to kidney transplant

PLoS ONE ◽  
2021 ◽  
Vol 16 (5) ◽  
pp. e0252069
Author(s):  
Juliana Feiman Sapiertein Silva ◽  
Gustavo Fernandes Ferreira ◽  
Marcelo Perosa ◽  
Hong Si Nga ◽  
Luis Gustavo Modelli de Andrade
Keyword(s):  
Machine Learning ◽  
Kidney Transplant ◽  
Waiting Time ◽  
Cox Regression ◽  
Final Analysis ◽  
Deceased Donor ◽  
Waiting List ◽  
Sensitivity Analyses ◽  
Cox Models ◽  
Use Of Resources

Background Predicting waiting time for a deceased donor kidney transplant can help patients and clinicians to discuss management and contribute to a more efficient use of resources. This study aimed at developing a predictor model to estimate time on a kidney transplant waiting list using a machine learning approach. Methods A retrospective cohort study including data of patients registered, between January 1, 2000 and December 31, 2017, in the waiting list of São Paulo State Organ Allocation System (SP-OAS) /Brazil. Data were randomly divided into two groups: 75% for training and 25% for testing. A Cox regression model was fitted with deceased donor transplant as the outcome. Sensitivity analyses were performed using different Cox models. Cox hazard ratios were used to develop the risk-prediction equations. Results Of 54,055 records retrieved, 48,153 registries were included in the final analysis. During the study period, approximately 1/3 of the patients were transplanted with a deceased donor. The major characteristics associated with changes in the likelihood of transplantation were age, subregion, cPRA, and frequency of HLA-DR, -B and -A. The model developed was able to predict waiting time with good agreement in internal validation (c-index = 0.70). Conclusion The kidney transplant waiting time calculator developed shows good predictive performance and provides information that may be valuable in assisting candidates and their providers. Moreover, it can significantly improve the use of economic resources and the management of patient care before transplant.

scholarly journals Machine learning and mechanistic modeling for prediction of metastatic relapse in early-stage breast cancer

2019 ◽  
Author(s):  
C. Nicolò ◽  
C. Périer ◽  
M. Prague ◽  
C. Bellera ◽  
G. MacGrogan ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Machine Learning ◽  
Cox Regression ◽  
Early Stage ◽  
Mechanistic Model ◽  
Early Stage Breast Cancer ◽  
Model Parameters ◽  
Machine Learning Classification ◽  
Metastatic Relapse ◽  
Random Survival Forest

AbstractPurposeFor patients with early-stage breast cancer, prediction of the risk of metastatic relapse is of crucial importance. Existing predictive models rely on agnostic survival analysis statistical tools (e.g. Cox regression). Here we define and evaluate the predictive ability of a mechanistic model for the time to metastatic relapse.MethodsThe data consisted of 642 patients with 21 clinicopathological variables. A mechanistic model was developed on the basis of two intrinsic mechanisms of metastatic progression: growth (parameterα) and dissemination (parameterμ). Population statistical distributions of the parameters were inferred using mixed-effects modeling. A random survival forest analysis was used to select a minimal set of 5 covariates with best predictive power. These were further considered to individually predict the model parameters, by using a backward selection approach. Predictive performances were compared to classical Cox regression and machine learning algorithms.ResultsThe mechanistic model was able to accurately fit the data. Covariate analysis revealed statistically significant association ofKi67expression withα(p=0.001) andEGFRwithμ(p=0.009). Achieving a c-index of 0.65 (0.60-0.71), the model had similar predictive performance as the random survival forest (c-index 0.66-0.69) and Cox regression (c-index 0.62 - 0.67), as well as machine learning classification algorithms.ConclusionBy providing informative estimates of the invisible metastatic burden at the time of diagnosis and forward simulations of metastatic growth, the proposed model could be used as a personalized prediction tool of help for routine management of breast cancer patients.

scholarly journals Early elevation of FIB-4 liver fibrosis score is associated with adverse outcomes among patients with COVID-19

2020 ◽  
Author(s):  
Fangfei Xiang ◽  
Jing Sun ◽  
Po-Hung Chen ◽  
Peijin Han ◽  
Haipeng Zheng ◽  
...  
Keyword(s):  
Regression Models ◽  
Cox Regression ◽  
Early Stage ◽  
Severe Disease ◽  
High Flow ◽  
Adverse Outcomes ◽  
Prolonged Hospitalization ◽  
Severe Stage ◽  
Increased Risk ◽  
High Flow Oxygen

Background Limited prior data suggest that pre-existing liver disease was associated with adverse outcomes among patients with COVID-19. FIB-4 is a noninvasive index of readily available laboratory measurements that represents hepatic fibrosis. The association of FIB-4 with COVID-19 outcomes has not been previously evaluated. Methods FIB-4 was evaluated at admission in a cohort of 267 patients admitted with early-stage COVID-19 confirmed through RT-PCR. Hazard of ventilator use and of high-flow oxygen was estimated using Cox regression models controlled for covariates. Risk of progress to severe cases and of death/prolonged hospitalization (>30 days) were estimated using logistic regression models controlled for same covariates. Results Forty-one (15%) patients progressed to severe cases, 36 (14%) required high-flow oxygen support, 10 (4%) required mechanical ventilator support, and 1 died. Patients with high FIB-4 score (>3.25) were more likely to be older with pre-existing conditions. FIB-4 between 1.45-3.25 was associated with over 5-fold (95% CI: 1.2-28) increased hazard of high-flow oxygen use, over 4-fold (95% CI: 1.5-14.6) increased odds of progress to severe stage, and over 3-fold (95% CI: 1.4-7.7) increased odds of death or prolonged hospitalization. FIB-4>3.25 was associated with over 12-fold (95% CI: 2.3-68. 7) increased hazard of high-flow oxygen use and over 11-fold (95% CI: 3.1-45) increased risk of progress to severe disease. All associations were independent of sex, number of comorbidities, and inflammatory markers (D-dimer, C-reactive protein). Conclusions FIB-4 at early-stage of COVID-19 disease had an independent and dose-dependent association with adverse outcomes during hospitalization. FIB-4 provided significant prognostic value to adverse outcomes among COVID-19 patients.

scholarly journals Radiological identification of temporal lobe epilepsy using artificial intelligence: a feasibility study

2021 ◽  
Author(s):  
Ezequiel Gleichgerrcht ◽  
Brent Munsell ◽  
Simon Keller ◽  
Daniel L Drane ◽  
Jens H Jensen ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Neural Networks ◽  
Temporal Lobe Epilepsy ◽  
Temporal Lobe ◽  
Neurological Diseases ◽  
Visual Assessment ◽  
Structural Mri ◽  
Mesial Temporal Sclerosis ◽  
Mri Findings ◽  
Mri Features

Abstract Temporal lobe epilepsy is associated with magnetic resonance imaging (MRI) findings reflecting underlying mesial temporal sclerosis. Identifying these MRI features is critical for the diagnosis and management of temporal lobe epilepsy. To date, this process relies on visual assessment by highly trained human experts (e.g. neuroradiologists, epileptologists). Artificial intelligence is increasingly recognized as a promising aid in the radiological evaluation of neurological diseases, yet its applications in temporal lobe epilepsy have been limited. Here, we applied a convolutional neural networks to assess the classification accuracy of temporal lobe epilepsy based on structural MRI. We demonstrate that convoluted neural networks can achieve high accuracy in the identification of unilateral temporal lobe epilepsy cases even when the MRI had been originally interpreted as normal by experts. We show that accuracy can be potentiated by employing smoothed gray matter maps and a direct acyclic graphs approach. We further discuss the foundations for the development of computer-aided tools to assist with the diagnosis of epilepsy.

scholarly journals Taking the Inner Route: Spatial and Demographic Factors Affecting Vulnerability to COVID-19 Among 604 Cities from Inner São Paulo State, Brazil

2020 ◽  
Author(s):  
Carlos Magno Castelo Branco Fortaleza ◽  
Raul Borges Guimarães ◽  
Gabriel Berg de Almeida ◽  
Micheli Pronunciate ◽  
Cláudia Pio Ferreira
Keyword(s):  
Population Density ◽  
Regression Models ◽  
Cox Regression ◽  
Positive Association ◽  
Urban Centers ◽  
Cox Models ◽  
Factors Affecting ◽  
Paulo State ◽  
The Impact ◽  
Road Distance

Objectives: The impact of COVID-19 in metropolitan areas has been extensively studied. The geographic spread to smaller cities is of great concern and may follow hierarchical influence of urban centers. With that in mind, we investigated factors that affect vulnerability of inner municipalities in São Paulo State, Brazil, an area with 24 million inhabitants. Methods: Surveillance data for confirmed COVID-19 cases and admissions for severe acute respiratory disease (SARD) up to April 18th were recorded for each of 604 municipalities that lay outside São Paulo metropolitan area. Vulnerability was assessed in Multivariable models, including sociodemographic indexes, road distance to the State Capital and the municipalities classification proposed by the Brazilian Institute of Geography and Statistics. Municipalities of great regional relevance were used as reference category for that classification. The outcome of interest for Cox regression was having COVID-cases, with time counting from the first report in São Paulo State. For binomial negative regression models, the outcomes of interest were rates of confirmed COVID-19 cases and admissions for SARD.Results: A total of 198 (32.8%) municipalities had autochthonous COVID-19 cases. In Cox models, affected municipalities were likely to have greater population density (Hazard Ratio[HR] for each 100 inhabitants per square kilometer, 1.07; 95% Confidence Interval [CI], (1.05-1.10)), proportion of inhabitants in urban area (HR, 1.02; 95%CI, 1.00-1.04), higher human development index (HDI, HR for 1%, 1.06; 95%CI, 1.00-1.13) and Gini Index for Inequality of income (HR for 1%, 1.04, 95% CI, 1.00-1.07). On the other hand, distance from the Capital was protective (HR for each 100Km, 0.82; 95%CI, 0.74-0.90). The HR95%[95%CI] also varied negatively according to the categories of influence of major centers (0.41 [0.22-0.77], 0.16 [0.09-0.32], 0.07 [0.03-0.15]). The binomial negative regression models for COVID-19 incidence also detected positive association with population density (Incidence Rate Ratio[IRR], 1.13; 95%CI, 1.07-1.18) and proportion of urban population (IRR, 1.04; 95%CI, 1.01-1.05), protection for cities distant to the Capital (IRR=0.73; 95%CI, 0.68-0.81) and increasing negative association for categories of influence (0.19 [0.09-0.42], 0.07 [0.03-0.15] and 0.03 [0.02-0.08]). Similar findings were detected when we used SARD incidence as outcome.Conclusion: Municipalities with greater population, density and regional influence were more likely to be affected earlier and more intensely by COVID-19. Non-pharmacological measures should be strengthened in those areas of greater risk.

A DNA-repair prognostic signature for early-stage NSCLC patients, in IFCT-0002 trial of neoadjuvant chemotherapy.

2013 ◽  
Vol 31 (15_suppl) ◽  
pp. 7515-7515
Author(s):  
Gerard Zalcman ◽  
Guenaelle Levallet ◽  
Pierre Fouret ◽  
Martine Antoine ◽  
Elisabeth Brambilla ◽  
...  
Keyword(s):  
Overall Survival ◽  
Dna Repair ◽  
Cox Regression ◽  
Early Stage ◽  
Prognostic Score ◽  
Univariate Analysis ◽  
Cox Models ◽  
Risk Of Death ◽  
Nsclc Patients

7515 Background: IFCT-0002 trial compared two perioperative CT regimens, CDDP-Gemcitabine vs.CBDCA-Paclitaxel in 528 stage I-II NSCLC patients. Paraffin-embedded post-chemo specimens were collected in the 490 non-complete responder patients for tissue expression studies of DNA-repair proteins. Methods: Surgical specimens were processed for immunohistochemistry as previously published. Variables were studied as continuous variables. Cut-off values were validated by bootstrap. Multivariate backward Cox regressions were used to adjust for patients’ characteristics associated with the corresponding outcome at p<0.20 in univariate analysis. Discrimination of the proposed Cox models was estimated using the c-indexes corrected for over-optimism by a resampling procedure. Median follow-up was 72.0 months, 95%CI [69.7-73.5]. Results: ERCC1, MSH2, XRCC5/Ku80 and BRCA1 immunostainings were available in 413, 356, 396 and 221 specimens. Expressed as a continuous variable, only MSH2 staining score correlated with overall survival. XRCC5 showed no influence on survival. When dichotomised, low BRCA1 (under median value) and ERCC1 (ERCC1=0), while high MSH2 protein expression (over median value), adversely affected overall survival with respective adj. HRs of 1.56, 95%CI [1.05-2.32], p=0.028 ; 1.37 95%CI [1.01-1.86], p=0.042 and 1.53, 95%CI [1.12-2.09], p=0.007. No interaction was found between the attributed treatment and any of the 4 markers. High MSH2 and low ERCC1 variables were tested in 200 bootstrap multivariate Cox models and correlated with OS in respectively 87% and 78.5% (c-index=0.570), whereas stage predicted survival in only 49% of those theoretical samples. A prognostic score led to the definition of three groups of high-, intermediate- and low-risk of death with respective HRs of 2.83, 1.60 and 1. Median OS were respectively 28.3 months, 71.5 and not reached, 5-y survival rates were 34.2%, 54.8% and 66.3% (Log-Rank p<0.0001). Conclusions: With a 6-year median follow-up, a prognostic score derived from multivariate Cox regression, validated by bootstraping, accurately discriminates a sub-group with high risk of death according to tumor expression of MSH2 and ERCC1.

Sign in / Sign up

Export Citation Format

Share Document