Feature Detection and Hypothesis Testing for Extremely Noisy Nanoparticle Images using Topological Data Analysis

Thomas, Andrew M.; Crozier, Peter A.; Xu, Yuchen; Matteson, David S.

doi:10.1080/00401706.2023.2203744

Cited by 2 publications

(2 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent work has established the utility of incorporating multiple types of topological and geometric information derived from persistence diagrams of various filtrations of images to improve predictive capabilities of machine learning models. 19 Two such filtrations are the height 19,20 and greyscale 21-24 filtrations. In the case of the height filtration an image is first binarized at some fixed threshold and the evolution of shape is examined as the black pixels are unveiled along on a given direction in the image— see Figure 1 .…”

Section: Methodsmentioning

confidence: 99%

“…29 Finally, the trimmed mean ‘ tmean ’ and trimmed variance ‘ tvar ’ of pixel intensities, the aspect ratio ‘ ratio ’ of images and two additional persistence statistics ‘ alps_grey ’ and ‘ alps_bd ’ were derived from each image. The ALPS statistic 24 ‘ alps_grey ’ was calculated for the greyscale filtration after smoothing the image with σ = 2. To reduce the risk of a saturated model, the 910 radiomics features and 180 topological features were projected onto their first two principal components, denoted as ‘ rad_pca1 ’, ‘ rad_pca2 ’, ‘ tda_pca1 ’, and ‘ tda_pca2 ’ respectively.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A proof-of-concept investigation into predicting follicular carcinoma on ultrasound using topological data analysis and radiomics

Thomas,

Lin,

Deng

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Background: Multiple ultrasound (US) risk stratifications systems (RSS) have been developed to estimate malignancy risk in thyroid nodules and recommend the need for fine needle aspiration biopsy. However, sonographic risk patterns identified in established RSSs may not accurately stratify follicular carcinoma from adenoma, which share many similar US characteristics. Quantitative medical imaging analysis aims to extract high-dimensional textural features from tumor phenotypes imperceptible to the human eye. The purpose of this study is to investigate the performance of a multimodal machine learning model utilizing radiomics and topological data analysis (TDA) to predict malignancy in follicular thyroid neoplasms on ultrasonography. Methods: This is a retrospective study of patients who underwent thyroidectomy with pathology confirmed follicular adenoma or carcinoma at a single academic medical center between 2010-2022. The nodule of interest on pre-operative ultrasound was annotated and masked to only include pixels. Images were scaled to maintain aspect ratio and ensure similar image resolution. Features derived from radiomics and TDA were calculated. High-dimensional radiomics and TDA features were projected onto their first two principal components respectively. Logistic regression with L2 penalty was used to predict malignancy. Classifier performance was evaluated using leave-one-out cross-validation and area under the curve (AUC). Results: Patients with follicular adenomas (n=7) and follicular carcinomas (n=11) with available imaging were included. 910 radiomics features were extracted for each image. 180 topological features from the height filtration, mean and variance of pixel intensities, the aspect ratio of images and two additional persistence statistics were derived from each image. The models achieved an AUC of 0.68 (radiomics only) and 0.88 (radiomics and TDA). The subsampling bootstrap was used to assess the confidence of these estimates. Conclusions: We demonstrate that inclusion of topological features yields strong improvement over radiomics-based features alone in the prediction of follicular carcinoma on ultrasound. Despite low volume data, the TDA features of the height filtration with its connection to the persistent homology transform (PHT), explicitly capture shape information that likely augments performance of the multimodal machine learning model. This approach suggests that a quantitative based US RSS may contribute to the preoperative prediction of follicular carcinoma.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

A proof-of-concept investigation into predicting follicular carcinoma on ultrasound using topological data analysis and radiomics

Thomas,

Lin,

Deng

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Recent progress of operando transmission electron microscopy in heterogeneous catalysis

Zhang,

Liu

2024

Microstructures

View full text Add to dashboard Cite

As a highly intricate process encompassing multiple length scales, catalysis research evolves into a comprehensive understanding of reaction kinetics across microscopic to atomic dimensions when electron microscopy, particularly the in situ transmission electron microscopy (TEM), emerges to be increasingly relevant. Meanwhile, the absence of effective methodologies for measuring reaction products during catalysis complicates efforts to elucidate the operational state and catalytic activity of the catalyst. With ongoing advancements of refined gas-cell design within TEM and other in situ accessories, diverse methodologies have emerged to ascertain the occurrence of chemical reactions. In this review, we summarized the recent progress of operando TEM while further extending its conceptual boundaries by including newly emerged reaction-detecting approaches capable of bridging microstructure to the reaction process. These methods involve not only traditional ones of product detection, e.g., in situ mass spectrometry and electron energy loss spectroscopy, but also other reaction-correlative characterizations, such as directly imaging reactant molecule, modified in situ reactor for thermogravimetry and temperature-programmed reaction, and TEM image-based microstructure quantification and activity correlation. Applications, inherent challenges, and our perspectives within these operando TEM techniques are deliberated.

show abstract

Feature Detection and Hypothesis Testing for Extremely Noisy Nanoparticle Images using Topological Data Analysis

Cited by 2 publications

References 54 publications

A proof-of-concept investigation into predicting follicular carcinoma on ultrasound using topological data analysis and radiomics

A proof-of-concept investigation into predicting follicular carcinoma on ultrasound using topological data analysis and radiomics

Recent progress of operando transmission electron microscopy in heterogeneous catalysis

Contact Info

Product

Resources

About