Analysis of Primary Care Computerized Medical Records (CMR) Data With Deep Autoencoders (DAE)

Thomas, Spencer A.; Smith, Nadia; Livina, Valerie N.; Yonova, Ivelina; Webb, Rebecca; Lusignan, Simon de

doi:10.3389/fams.2019.00042

Cited by 5 publications

(8 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to utilise metadata associated with images, we use the framework outlined in [22] to combine the image features obtained from a deep neural network and the associated metadata. We use a method inspired by [3] to obtain representations of the metadata that are compatible with the image features obtained in the neural network. This maps the metadata M to a numerical vector G (M )…”

Section: Methodsmentioning

confidence: 99%

“…Deep learning has emerged as a powerful suite of tools for image classification [1], and has a huge potential to solve challenges in healthcare settings. The use of deep neural networks is successful at tasks such as classification of medical images [2], analysis of electronic health records [3]- [5] and segmenting data from emerging medical technologies [6], [7]. This enormous potential comes with the caveat that very large amounts of data are required to train robust models that generalise beyond the training set.…”

Section: Introductionmentioning

confidence: 99%

“…scanner parameters, or relevant extracts from computerised medical records (CMR). These resources contain rich information relating to diseases [18], [19], and data driven methods can identify patterns of patients [3], [4].…”

Section: Introductionmentioning

confidence: 99%

“…Computerised medical records contain a patient (meta) data and are often complex high dimensional data sources. Methods for analysing CMRs have uncovered low dimensional patterns relating to sub-groups of patients [3], [4] and have enormous potential for insight into biomedical sciences. This paper focuses on the potential of integrating imaging data with non-imaging data (e.g.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Enhanced Transfer Learning Through Medical Imaging and Patient Demographic Data Fusion

Thomas¹

2021

Preprint

Self Cite

View full text Add to dashboard Cite

In this work we examine the performance enhancement in classification of medical imaging data when image features are combined with associated non-image data. We compare the performance of eight state-of-the-art deep neural networks in classification tasks when using only image features, compared to when these are combined with patient metadata. We utilise transfer learning with networks pretrained on ImageNet used directly as feature extractors and fine tuned on the target domain. Our experiments show that performance can be significantly enhanced with the inclusion of metadata and use interpretability methods to identify which features lead to these enhancements. Furthermore, our results indicate that the performance enhancement for natural medical imaging (e.g. optical images) benefit most from direct use of pre-trained models, whereas non natural images (e.g. representations of non imaging data) benefit most from fine tuning pre-trained networks. These enhancements come at a negligible additional cost in computation time, and therefore is a practical method for other applications.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Enhanced Transfer Learning Through Medical Imaging and Patient Demographic Data Fusion

Thomas¹

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Specifically ′ We use a random 90:10 split of the data for training and testing and use the subset of A' to train the models to predict the corresponding subset of L. When using A' directly we obtain the results seen in Figure 4.2 where no method is able to give robust performance (determined via the metrics) for all episode types. Due to the high dimensionality of the data and the ability of deep autoencoders to identify patterns in data 126 , we consider classification of the data as in Table 4.2 and data reduced using an autoencoder 127 . These results are given in Both Figure 4.…”

Section: Prediction Of Episode Typementioning

confidence: 99%

NMS 2018-2021 Life-sciences and healthcare project "Digital health: curation of healthcare data" - final report

Smith¹,

Romanchikova²,

Partarrieu³

et al. 2021

View full text Add to dashboard Cite

NPL Report MS 31 I TERMINOLOGY Data CurationOrganization and integration of data collected from various sources, annotation of the data, and publication and presentation of the data such that the value of the data is maintained over time, and the data remains available for reuse and preservation. Data Interoperability Addresses the ability of systems and services that create, exchange and consume data to have clear, shared expectations for the contents, context and meaning of that data. Data Integration Aggregation of datasets from heterogeneous sources by linking, combination or fusion. Data Provenance A history of the dataset that captures its origin, purpose and all modifications it underwent since its creation. FAIR principles A set of guiding principles to make data Findable, Accessible, Interoperable and Re-usable. MeasurementThe process of experimentally obtaining one or more quantity values that can reasonably be attributed to a quantity. MetadataA set of descriptors of the dataset that make the dataset easier to locate, understand and use. Metadata can include information about the purpose of the dataset creation, the experimental setup or the data formats. Depending on the nature of the investigation, metadata can be data in its own right. OntologyA formal representation of a knowledge domain that comprises a vocabulary of terms, as well as relations between them. An ontology can be presented as a knowledge graph where terms are depicted as vertices and relationships as edges. ReproducibilityCondition of measurement, out of a set of conditions that includes different locations, operators, measuring systems, and replicate measurements on the same or similar objects. TraceabilityProperty of a measurement result whereby the result can be related to a reference through a documented unbroken chain of calibrations, each contributing to the measurement uncertainty. Uncertainty of MeasurementThe parameter characterizing the dispersion of the quantity values being attributed to a measurand

show abstract

Fast Data Driven Estimation of Cluster Number in Multiplex Images using Embedded Density Outliers

Thomas

2022

2022 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB)

View full text Add to dashboard Cite

The usage of chemical imaging technologies is becoming a routine accompaniment to traditional methods in pathology. Significant technological advances have developed these next generation techniques to provide rich, spatially resolved, multidimensional chemical images. The rise of digital pathology has significantly enhanced the synergy of these imaging modalities with optical microscopy and immunohistochemistry, enhancing our understanding of the biological mechanisms and progression of diseases. Techniques such as imaging mass cytometry provide labelled multidimensional (multiplex) images of specific components used in conjunction with digital pathology techniques. These powerful techniques generate a wealth of high dimensional data that create significant challenges in data analysis. Unsupervised methods such as clustering are an attractive way to analyse these data, however, they require the selection of parameters such as the number of clusters. Here we propose a methodology to estimate the number of clusters in an automatic data-driven manner using a deep sparse autoencoder to embed the data into a lower dimensional space. We compute the density of regions in the embedded space, the majority of which are empty, enabling the high density regions (i.e. clusters) to be detected as outliers and provide an estimate for the number of clusters. This framework provides a fully unsupervised and data-driven method to analyse multidimensional data. In this work we demonstrate our method using 45 multiplex imaging mass cytometry datasets. Moreover, our model is trained using only one of the datasets and the learned embedding is applied to the remaining 44 images providing an efficient process for data analysis. Finally, we demonstrate the high computational efficiency of our method which is two orders of magnitude faster than estimating via computing the sum squared distances as a function of cluster number.

show abstract

Analysis of Primary Care Computerized Medical Records (CMR) Data With Deep Autoencoders (DAE)

Cited by 5 publications

References 28 publications

Enhanced Transfer Learning Through Medical Imaging and Patient Demographic Data Fusion

Enhanced Transfer Learning Through Medical Imaging and Patient Demographic Data Fusion

NMS 2018-2021 Life-sciences and healthcare project "Digital health: curation of healthcare data" - final report

Fast Data Driven Estimation of Cluster Number in Multiplex Images using Embedded Density Outliers

Contact Info

Product

Resources

About