Automatic classification of takeaway food outlet cuisine type using machine (deep) learning.
Ontology highlight
ABSTRACT: Background and purpose
Researchers have not disaggregated neighbourhood exposure to takeaway ('fast-') food outlets by cuisine type sold, which would otherwise permit examination of differential impacts on diet, obesity and related disease. This is partly due to the substantial resource challenge of manual classification of unclassified takeaway outlets at scale. We describe the development of a new model to automatically classify takeaway food outlets, by 10 major cuisine types, based on business name alone.Material and methods
We used machine (deep) learning, and specifically a Long Short Term Memory variant of a Recurrent Neural Network, to develop a predictive model trained on labelled outlets (n = 14,145), from an online takeaway food ordering platform. We validated the accuracy of predictions on unseen labelled outlets (n = 4,000) from the same source.Results
Although accuracy of prediction varied by cuisine type, overall the model (or 'classifier') made a correct prediction approximately three out of four times. We demonstrated the potential of the classifier to public health researchers and for surveillance to support decision-making, through using it to characterise nearly 55,000 takeaway food outlets in England by cuisine type, for the first time.Conclusions
Although imperfect, we successfully developed a model to classify takeaway food outlets, by 10 major cuisine types, from business name alone, using innovative data science methods. We have made the model available for use elsewhere by others, including in other contexts and to characterise other types of food outlets, and for further development.
SUBMITTER: Bishop TRP
PROVIDER: S-EPMC8700226 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Consumer food environments have transformed dramatically in the last decade. Food outlet prevalence has increased, and people are eating food outside the home more than ever before. Despite these developments, national spending on food control has reduced. The National Audit Office report that only 14% of local authorities are up to date with food business inspections, exposing consumers to unknown levels of risk. Given the scarcity of local authority resources, this paper presents a data-driven approach to predict compliance for newly opened businesses and those awaiting repeat inspections. This work capitalizes on the theory that food outlet compliance is a function of its geographic context, namely the characteristics of the neighborhood within which it sits. We explore the utility of three machine learning approaches to predict non-compliant food outlets in England and Wales using openly accessible socio-demographic, business type, and urbanness features at the output area level. We find that the synthetic minority oversampling technique alongside a random forest algorithm with a 1:1 sampling strategy provides the best predictive power. Our final model retrieves and identifies 84% of total non-compliant outlets in a test set of 92,595 (sensitivity = 0.843, specificity = 0.745, precision = 0.274). The originality of this work lies in its unique and methodological approach which combines the use of machine learning with fine-grained neighborhood data to make robust predictions of compliance.
| S-EPMC8656817 | biostudies-literature
Project description:BackgroundPattern of emphysema at chest CT, scored visually by using the Fleischner Society system, is associated with physiologic impairment and mortality risk.PurposeTo determine whether participant-level emphysema pattern could predict impairment and mortality when classified by using a deep learning method.Materials and MethodsThis retrospective analysis of Genetic Epidemiology of COPD (COPDGene) study participants enrolled between 2007 and 2011 included those with baseline CT, visual emphysema scores, and survival data through 2018. Participants were partitioned into nonoverlapping sets of 2407 for algorithm training, 100 for validation and parameter tuning, and 7143 for testing. A deep learning algorithm using convolutional neural network and long short-term memory architectures was trained to classify pattern of emphysema according to Fleischner criteria. Deep learning scores were compared with visual scores and clinical parameters including pulmonary function tests. Cox proportional hazard models were used to evaluate relationships between emphysema scores and survival. The algorithm was also tested by using CT and clinical data in 1962 participants enrolled in the Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-points (ECLIPSE) study.ResultsA total of 7143 COPDGene participants (mean age ± standard deviation, 59.8 years ± 8.9; 3734 men and 3409 women) were evaluated. Deep learning emphysema classifications were associated with impaired pulmonary function tests, 6-minute walk distance, and St George's Respiratory Questionnaire at univariate analysis (P < .001 for each). Testing in the ECLIPSE cohort showed similar associations (P < .001). In the COPDGene test cohort, deep learning emphysema classification improved the fit of linear mixed models in the prediction of these clinical parameters compared with visual scoring (P < .001). Compared with participants without emphysema, mortality was greater in participants classified by the deep learning algorithm as having any grade of emphysema (adjusted hazard ratios were 1.5, 1.7, 2.9, 5.3, and 9.7, respectively, for trace, mild, moderate, confluent, and advanced destructive emphysema; P < .05).ConclusionDeep learning automation of the Fleischner grade of emphysema at chest CT is associated with clinical measures of pulmonary insufficiency and the risk of mortality.© RSNA, 2019Online supplemental material is available for this article.
| S-EPMC6996603 | biostudies-literature
Project description:Breast cancer is a heterogeneous disease defined by molecular types and subtypes. Advances in genomic research have enabled use of precision medicine in clinical management of breast cancer. A critical unmet medical need is distinguishing triple negative breast cancer, the most aggressive and lethal form of breast cancer, from non-triple negative breast cancer. Here we propose use of a machine learning (ML) approach for classification of triple negative breast cancer and non-triple negative breast cancer patients using gene expression data. We performed analysis of RNA-Sequence data from 110 triple negative and 992 non-triple negative breast cancer tumor samples from The Cancer Genome Atlas to select the features (genes) used in the development and validation of the classification models. We evaluated four different classification models including Support Vector Machines, K-nearest neighbor, Naïve Bayes and Decision tree using features selected at different threshold levels to train the models for classifying the two types of breast cancer. For performance evaluation and validation, the proposed methods were applied to independent gene expression datasets. Among the four ML algorithms evaluated, the Support Vector Machine algorithm was able to classify breast cancer more accurately into triple negative and non-triple negative breast cancer and had less misclassification errors than the other three algorithms evaluated. The prediction results show that ML algorithms are efficient and can be used for classification of breast cancer into triple negative and non-triple negative breast cancer types.
| S-EPMC7909418 | biostudies-literature
Project description:Automated assessment of histological features of non-alcoholic fatty liver disease (NAFLD) may reduce human variability and provide continuous rather than semiquantitative measurement of these features. As part of a larger effort, we perform automatic classification of steatosis, the cardinal feature of NAFLD, and other regions that manifest as white in images of hematoxylin and eosin-stained liver biopsy sections. These regions include macrosteatosis, central veins, portal veins, portal arteries, sinusoids and bile ducts. Digital images of hematoxylin and eosin-stained slides of 47 liver biopsies from patients with normal liver histology (n = 20) and NAFLD (n = 27) were obtained at 20× magnification. The images were analyzed using supervised machine learning classifiers created from annotations provided by two expert pathologists. The classification algorithm performs with 89% overall accuracy. It identified macrosteatosis, bile ducts, portal veins and sinusoids with high precision and recall (≥ 82%). Identification of central veins and portal arteries was less robust but still good. The accuracy of the classifier in identifying macrosteatosis is the best reported. The accurate automated identification of macrosteatosis achieved with this algorithm has useful clinical and research-related applications. The accurate detection of liver microscopic anatomical landmarks may facilitate important subsequent tasks, such as localization of other histological lesions according to liver microscopic anatomy.
| S-EPMC8374469 | biostudies-literature
Project description:Cell type classification is an important problem in cancer research, especially with the advent of single cell technologies. Correctly identifying cells within the tumor microenvironment can provide oncologists with a snapshot of how a patient's immune system reacts to the tumor. Wide and deep learning (WDL) is an approach to construct a cell-classification prediction model that can learn patterns within high-dimensional data (deep) and ensure that biologically relevant features (wide) remain in the final model. In this paper, we demonstrate that regularization can prevent overfitting and adding a wide component to a neural network can result in a model with better predictive performance. In particular, we observed that a combination of dropout and ℓ2 regularization can lead to a validation loss function that does not depend on the number of training iterations and does not experience a significant decrease in prediction accuracy compared to models with ℓ1 , dropout, or no regularization. Additionally, we show WDL can have superior classification accuracy when the training and testing of a model are completed data on that arise from the same cancer type but different platforms. More specifically, WDL compared to traditional deep learning models can substantially increase the overall cell type prediction accuracy (36.5 to 86.9%) and T cell subtypes (CD4: 2.4 to 59.1%, and CD8: 19.5 to 96.1%) when the models were trained using melanoma data obtained from the 10X platform and tested on basal cell carcinoma data obtained using SMART-seq. WDL obtains higher accuracy when compared to state-of-the-art cell classification algorithms CHETAH (70.36%) and SingleR (70.59%).
| S-EPMC7878986 | biostudies-literature
Project description:PurposeTo develop a software package for automated classification of anterior chamber angle of the eye by using ultrasound biomicroscopy.MethodsUltrasound biomicroscopy images were collected, and the trabecular-iris angle was manually measured and classified into three categories: open angle, narrow angle, and angle closure. Inception v3 was used as the classifying convolutional neural network and the algorithm was trained.ResultsWith a recall rate of 97% in the test set, the neural network's classification accuracy can reach 97.2% and the overall area under the curve was 0.988. The sensitivity and specificity were 98.04% and 99.09% for the open angle, 96.30% and 98.13% for the narrow angle, and 98.21% and 99.05% for the angle closure categories, respectively.ConclusionsPreliminary results show that an automated classification of the anterior chamber angle achieved satisfying sensitivity and specificity and could be helpful in clinical practice.Translational relevanceThe present work suggests that the algorithm described here could be useful in the categorizing of anterior chamber angle and screening for subjects who are at high risk of angle closure.
| S-EPMC6703191 | biostudies-literature
Project description:Linear biometric measurements on magnetic resonance images are important for the assessment of fetal brain development, which is expert knowledge dependent and laborious. This study aims to construct a segmentation-based method for automatic two-dimensional biometric measurements of fetal brain on magnetic resonance images that provides a fast and accurate measurement of fetal brain. A total of 268 volumes (5360 images) magnetic resonance images of normal fetuses were included. The automatic method involves two steps. First, the fetal brain was segmented into four parts with a deep segmentation network: cerebrum, cerebellum, and left and right lateral ventricles. Second, the measurement plane was determined, and the corresponding biometric parameters were calculated according to clinical guidelines, including cerebral biparietal diameter (CBPD), transverse cerebellar diameter (TCD), left and right atrial diameter (LAD/RAD). Pearson correlation coefficient and Bland-Altman plots were used to assess the correlation and agreement between computer-predicted values and manual measurements. Mean differences were used to evaluate the errors quantitatively. Analysis of fetal cerebral growth based on the automatic measurements was also displayed. The experiment results show that correlation coefficients for CBPD, TCD, LAD and RAD were as follows: 0.977, 0.990, 0.817, 0.719, mean differences were - 2.405 mm, - 0.008 mm, - 0.33 mm, - 0.213 mm, respectively. The correlation between the errors and gestational age was not statistically significant (p values were 0.2595, 0.0510, 0.1995, and 0.0609, respectively). The proposed automatic method for linear measurements on fetal brain MRI achieves excellent performance, which is expected to be applied in clinical practice and be helpful for prenatal diagnosis and clinical work efficiency improvement.
| S-EPMC10587162 | biostudies-literature
Project description:Ultrasonic vocalizations (USVs) analysis is a well-recognized tool to investigate animal communication. It can be used for behavioral phenotyping of murine models of different disorders. The USVs are usually recorded with a microphone sensitive to ultrasound frequencies and they are analyzed by specific software. Different calls typologies exist, and each ultrasonic call can be manually classified, but the qualitative analysis is highly time-consuming. Considering this framework, in this work we proposed and evaluated a set of supervised learning methods for automatic USVs classification. This could represent a sustainable procedure to deeply analyze the ultrasonic communication, other than a standardized analysis. We used manually built datasets obtained by segmenting the USVs audio tracks analyzed with the Avisoft software, and then by labelling each of them into 10 representative classes. For the automatic classification task, we designed a Convolutional Neural Network that was trained receiving as input the spectrogram images associated to the segmented audio files. In addition, we also tested some other supervised learning algorithms, such as Support Vector Machine, Random Forest and Multilayer Perceptrons, exploiting informative numerical features extracted from the spectrograms. The performance showed how considering the whole time/frequency information of the spectrogram leads to significantly higher performance than considering a subset of numerical features. In the authors' opinion, the experimental results may represent a valuable benchmark for future work in this research field.
| S-EPMC7815145 | biostudies-literature
Project description:Traces from single-molecule fluorescence microscopy (SMFM) experiments exhibit photophysical artifacts that typically necessitate human expert screening, which is time-consuming and introduces potential for user-dependent expectation bias. Here, we use deep learning to develop a rapid, automatic SMFM trace selector, termed AutoSiM, that improves the sensitivity and specificity of an assay for a DNA point mutation based on single-molecule recognition through equilibrium Poisson sampling (SiMREPS). The improved performance of AutoSiM is based on accepting both more true positives and fewer false positives than the conventional approach of hidden Markov modeling (HMM) followed by hard thresholding. As a second application, the selector is used for automated screening of single-molecule Förster resonance energy transfer (smFRET) data to identify high-quality traces for further analysis, and achieves ~90% concordance with manual selection while requiring less processing time. Finally, we show that AutoSiM can be adapted readily to novel datasets, requiring only modest Transfer Learning.
| S-EPMC7673028 | biostudies-literature
Project description:Reliable markers measuring disease progression in Huntington's disease (HD), before and after disease manifestation, may guide a therapy aimed at slowing or halting disease progression. Quantitative electroencephalography (qEEG) may provide a quantification method for possible (sub)cortical dysfunction occurring prior to or concomitant with motor or cognitive disturbances observed in HD. In this pilot study we construct an automatic classifier distinguishing healthy controls from HD gene carriers using qEEG and derive qEEG features that correlate with clinical markers known to change with disease progression in HD, with the aim of exploring biomarker potential. We included twenty-six HD gene carriers (49.7?±?8.5 years) and 25 healthy controls (52.7?±?8.7 years). EEG was recorded for three minutes with subjects at rest. An EEG index was created by applying statistical pattern recognition to a large set of EEG features, which was subsequently tested using 10-fold cross-validation. The index resulted in a continuous variable ranging from 0 to 1: a low value indicating a state close to normal and a high value pointing to HD. qEEG features that correlate specifically with commonly used clinical markers in HD research were derived. The classification index had a specificity of 83%, a sensitivity of 83% and an accuracy of 83%. The area under the curve of the receiver operator characteristic curve was 0.9. qEEG analysis on subsets of electrophysiological features resulted in two highly significant correlations with clinical scores. The results of this pilot study suggest that qEEG may serve as a biomarker in HD. The indices correlating with modalities changing with the progression of the disease may lead to tools based on qEEG that help monitor efficacy in intervention studies.
| S-EPMC6208376 | biostudies-literature