RESUMEN
Objective: Neoadjuvant chemotherapy (NAC) is a key element of treatment for locally advanced breast cancer (LABC). Predicting the response to NAC for patients with Locally Advanced Breast Cancer (LABC) before treatment initiation could be beneficial to optimize therapy, ensuring the administration of effective treatments. The objective of the work here was to develop a predictive model to predict tumor response to NAC for LABC using deep learning networks and computed tomography (CT). Materials and methods: Several deep learning approaches were investigated including ViT transformer and VGG16, VGG19, ResNet-50, Res-Net-101, Res-Net-152, InceptionV3 and Xception transfer learning networks. These deep learning networks were applied on CT images to assess the response to NAC. Performance was evaluated based on balanced_accuracy, accuracy, sensitivity and specificity classification metrics. A ViT transformer was applied to utilize the attention mechanism in order to increase the weight of important part image which leads to better discrimination between classes. Results: Amongst the 117 LABC patients studied, 82 (70%) had clinical-pathological response and 35 (30%) had no response to NAC. The ViT transformer obtained the best performance range (accuracy = 71 ± 3% to accuracy = 77 ± 4%, specificity = 86 ± 6% to specificity = 76 ± 3%, sensitivity = 56 ± 4% to sensitivity = 52 ± 4%, and balanced_accuracy=69 ± 3% to balanced_accuracy=69 ± 3%) depending on the split ratio of train-data and test-data. Xception network obtained the second best results (accuracy = 72 ± 4% to accuracy = 65 ± 4, specificity = 81 ± 6% to specificity = 73 ± 3%, sensitivity = 55 ± 4% to sensitivity = 52 ± 5%, and balanced_accuracy = 66 ± 5% to balanced_accuracy = 60 ± 4%). The worst results were obtained using VGG-16 transfer learning network. Conclusion: Deep learning networks in conjunction with CT imaging are able to predict the tumor response to NAC for patients with LABC prior to start. A ViT transformer could obtain the best performance, which demonstrated the importance of attention mechanism.
RESUMEN
RATIONALE: Although numerous candidate features exist for predicting risk of higher risk of healthcare utilization in patients with chronic obstructive pulmonary disease (COPD), the process for selecting the most discriminative features remains unclear. OBJECTIVE: The objective of this study was to develop a robust feature selection method to identify the most discriminative candidate features for predicting healthcare utilization in COPD, and compare the model performance with other common feature selection methods. MATERIALS AND METHODS: In this retrospective study, demographic, lung function measurements and CT images were collected from 454 COPD participants from the Canadian Cohort Obstructive Lung Disease study from 2010-2017. A follow-up visit was completed approximately 1.5 years later and participants reported healthcare utilization. CT analysis was performed for feature extraction. A two-step hybrid feature selection method was proposed that utilized: (1) sparse subspace learning with nonnegative matrix factorization, and, (2) genetic algorithm. Seven commonly used feature selection methods were also implemented that reported the top 10 or 20 features for comparison. Performance was evaluated using accuracy. RESULTS: Of the 454 COPD participants evaluated, 161 (35%) utilized healthcare services at follow-up. The accuracy for predicting subsequent healthcare utilization for the seven commonly used feature selection methods ranged from 72%-76% with the top 10 features, and 77%-80% with the top 20 features. Relative to these methods, hybrid feature selection obtained significantly higher accuracy for predicting subsequent healthcare utilization at 82% ± 3% (p < 0.05). Selected features with the proposed method included: DLCO, FEV1, RV, FVC, TAC, LAA950, Pi-10, LAA856, LAC total hole count, outer area RB1, wall area RB1, wall area and Jacobian. CONCLUSION: The hybrid feature selection method identified the most discriminative features for classifying individuals with and without future healthcare utilization, and increased the accuracy compared to other state-of-the-art approaches.
RESUMEN
High-dimensional problems have increasingly drawn attention in gene selection and analysis. To add insult to injury, usually the number of features is greater than number of samples in microarray gene dataset which leads to an ill-posed underdetermined equation system. Poor performance and high computational time for learning algorithms are consequences of redundant features in high-dimensional data. Feature selection is a noteworthy pre-processing method to ameliorate the curse of dimensionality with aim of maximum relevancy and minimum redundancy information preservation. Likewise, unsupervised feature selection has been important since collecting labels for data is expensive. In this paper, we develop a novel robust unsupervised feature selection to select discriminative subset of features for unlabeled data based on rank constrained and dual regularized nonnegative matrix factorization. The major focus of the proposed technique is to discard redundant features while keeping the informative features. Proposed feature selection technique consists of nonnegative matrix factorization to decompose the data into feature weight matrix and representation matrix, inner product norm as regularization for both feature weight matrix and representation matrix, adaptive structure learning to preserve local information and Schatten-p norm as rank constraint. To demonstrate the effectiveness of the proposed method, numerical studies are conducted on six benchmark microarray datasets. The results show that the proposed technique outperforms eight state-of-art unsupervised feature selection techniques in terms of clustering accuracy and normalized mutual information.
Asunto(s)
Algoritmos , Análisis por Micromatrices , Análisis por Conglomerados , Expresión GénicaRESUMEN
Gene selection as a problem with high dimensions has drawn considerable attention in machine learning and computational biology over the past decade. In the field of gene selection in cancer datasets, different types of feature selection techniques in terms of strategy (filter, wrapper and embedded) and label information (supervised, unsupervised, and semi-supervised) have been developed. However, using hybrid feature selection can still improve the performance. In this paper, we propose a hybrid feature selection based on filter and wrapper strategies. In the filter-phase, we develop an unsupervised features selection based on non-convex regularized non-negative matrix factorization and structure learning, which we deem NCNMFSL. In the wrapper-phase, for the first time, mushroom reproduction optimization (MRO) is leveraged to obtain the most informative features subset. In this hybrid feature selection method, irrelevant features are filtered-out through NCNMFSL, and most discriminative features are selected by MRO. To show the effectiveness and proficiency of the proposed method, numerical experiments are conducted on Breast, Heart, Colon, Leukemia, Prostate, Tox-171 and GLI-85 benchmark datasets. SVM and decision tree classifiers are leveraged to analyze proposed technique and top accuracy are 0.97, 0.84, 0.98, 0.95, 0.98, 0.87 and 0.85 for Breast, Heart, Colon, Leukemia, Prostate, Tox-171 and GLI-85, respectively. The computational results show the effectiveness of the proposed method in comparison with state-of-art feature selection techniques.
Asunto(s)
Agaricales , Leucemia , Neoplasias , Masculino , Humanos , Algoritmos , Aprendizaje Automático , Neoplasias/genéticaRESUMEN
RATIONALE: Predicting increased risk of future healthcare utilization in chronic obstructive pulmonary disease (COPD) patients is an important goal for improving patient management. OBJECTIVE: Our objective was to determine the importance of computed tomography (CT) lung imaging measurements relative to other demographic and clinical measurements for predicting future health services use with machine learning in COPD. MATERIALS AND METHODS: In this retrospective study, lung function measurements and chest CT images were acquired from Canadian Cohort of Obstructive Lung Disease study participants from 2010 to 2017 (https://clinicaltrials.gov, NCT00920348). Up to two follow-up visits (1.5- and 3-year follow-up) were performed and participants were asked for details related to healthcare utilization. Healthcare utilization was defined as any COPD hospitalization or emergency room visit due to respiratory problems in the 12 months prior to the follow-up visits. CT analysis was performed (VIDA Diagnostics Inc.); a total of 108 CT quantitative emphysema, airway and vascular measurements were investigated. A hybrid feature selection method with support vector machine classifier was used to predict healthcare utilization. Performance was determined using accuracy, F1-measure and area under the receiver operating characteristic curve (AUC) and Matthews's correlation coefficient (MC). RESULTS: Of the 527 COPD participants evaluated, 179 (35%) used healthcare services at follow-up. There were no significant differences between the participants with or without healthcare utilization at follow-up for age (p = 0.50), sex (p = 0.44), BMI (p = 0.05) or pack-years (p = 0.76). The accuracy for predicting subsequent healthcare utilization was 80% ± 3% (F1-measure = 74%, AUC = 0.80, MC = 0.6) when all measurements were considered, 76% ± 6% (F1-measure = 72%, AUC = 0.77, MC = 0.55) for CT measurements alone and 65% ± 5% (F1-measure = 60%, AUC = 0.67, MC = 0.34) for demographic and lung function measurements alone. CONCLUSION: The combination of CT lung imaging and conventional measurements leads to greater prediction accuracy of subsequent health services use than conventional measurements alone, and may provide needed prognostic information for patients suffering from COPD.
Asunto(s)
Enfermedad Pulmonar Obstructiva Crónica , Humanos , Estudios Retrospectivos , Canadá , Enfermedad Pulmonar Obstructiva Crónica/diagnóstico por imagen , Enfermedad Pulmonar Obstructiva Crónica/terapia , Pulmón/diagnóstico por imagen , Tomografía Computarizada por Rayos X/métodos , Aprendizaje Automático , Hospitalización , Servicio de Urgencia en HospitalRESUMEN
RATIONALE: Texture-based radiomics analysis of lung computed tomography (CT) images has been shown to predict chronic obstructive pulmonary disease (COPD) status using machine learning models. However, various approaches are used and it is unclear which provides the best performance. OBJECTIVES: To compare the most commonly used feature selection and classification methods and determine the optimal models for classifying COPD status in a mild, population-based COPD cohort. MATERIALS AND METHODS: CT images from the multi-center Canadian Cohort Obstructive Lung Disease (CanCOLD) study were pre-processed by resampling the image to a 1mm isotropic voxel volume, segmenting the lung and removing the airways (VIDA Diagnostics Inc.), and applying a threshold of -1000HU-to-0HU. A total of 95 texture features were then extracted from each CT image. Combinations of 17 feature selection methods and 9 classifiers were tested and evaluated. In addition, the role of data cleaning (outlier removal and highly correlated feature removal) was evaluated. The area under the curve (AUC) from the receiver operating characteristic curve was used to evaluate model performance. RESULTS: A total of 1204 participants were evaluated (n = 602 no COPD, n = 602 COPD). There were no significant differences between the groups for female sex (no COPD = 46.3%; COPD = 38.5%; p = 0.77), or body mass index (no COPD = 27.7 kg/m2; COPD = 27.4 kg/m2; p = 0.21). The highest AUC value for predicting COPD status (AUC = 0.78 [0.73, 0.84]) was obtained following data cleaning and feature selection using Elastic Net with the Linear-SVM classifier. CONCLUSION: In a population-based cohort, the optimal combination for radiomics-based prediction of COPD status was Elastic Net as the feature selection method and Linear-SVM as the classifier.
Asunto(s)
Pulmón , Enfermedad Pulmonar Obstructiva Crónica , Humanos , Femenino , Canadá , Pulmón/diagnóstico por imagen , Tomografía Computarizada por Rayos X/métodos , Enfermedad Pulmonar Obstructiva Crónica/diagnóstico por imagen , Aprendizaje AutomáticoRESUMEN
BACKGROUND: There are similarities and differences between chronic obstructive pulmonary disease (COPD) and asthma patients in terms of computed tomography (CT) disease-related features. Our objective was to determine the optimal subset of CT imaging features for differentiating COPD and asthma using machine learning. METHODS: COPD and asthma patients were recruited from Heidelberg University Hospital (Heidelberg, Germany). CT was acquired and 93 features were extracted: percentage of low-attenuating area below -950â HU (LAA950), low-attenuation cluster (LAC) total hole count, estimated airway wall thickness for an idealised airway with an internal perimeter of 10â mm (Pi10), total airway count (TAC), as well as airway inner/outer perimeters/areas and wall thickness for each of five segmental airways, and the average of those five airways. Hybrid feature selection was used to select the optimum number of features, and support vector machine learning was used to classify COPD and asthma. RESULTS: 95 participants were included (n=48 COPD and n=47 asthma); there were no differences between COPD and asthma for age (p=0.25) or forced expiratory volume in 1â s (p=0.31). In a model including all CT features, the accuracy and F1 score were 80% and 81%, respectively. The top features were: LAA950, outer airway perimeter, inner airway perimeter, TAC, outer airway area RB1, inner airway area RB1 and LAC total hole count. In the model with only CT airway features, the accuracy and F1 score were 66% and 68%, respectively. The top features were: inner airway area RB1, outer airway area LB1, outer airway perimeter, inner airway perimeter, Pi10, TAC, airway wall thickness RB1 and TAC LB10. CONCLUSION: COPD and asthma can be differentiated using machine learning with moderate-to-high accuracy by a subset of only seven CT features.