Dataset Information

SEMPro: A Data-Driven Pipeline To Learn Structure-Property Insights from Scanning Electron Microscopy Images.

ABSTRACT: Analyzing hydrogel microstructure through scanning electron microscopy (SEM) images is crucial in understanding hydrogel properties. However, the analysis of SEM images in hydrogel research heavily relies on the intuition of individual researchers and is constrained by the limited size of the dataset. To address this, we propose SEMPro, a data-driven solution using web-scraping and deep learning (DL) to compile and analyze the structure-property relationships of hydrogels through SEM images. It accurately predicts the elastic modulus from SEM images within the same order of magnitude and displays a learned extraction of modulus-relevant features in SEM images as seen through the nontrivial activation mapping and transfer learning. By employing Explainable AI through activation map exposure, SEMPro validates the model predictions. SEMPro represents a closed-loop data collection and analysis pipeline, providing critical insights into hydrogels and soft materials. This innovative approach has the potential to revolutionize hydrogel research, offering high-dimensional insights for further advancements.

SUBMITTER: Ho B

PROVIDER: S-EPMC10630981 | biostudies-literature | 2023 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

SEMPro: A Data-Driven Pipeline To Learn Structure-Property Insights from Scanning Electron Microscopy Images.

Ho Brandon B Zhao Jiayu J Liu Joseph J Tang Lisa L Guan Zhecun Z Li Xiao X Li Minghao M Howard Elizabeth E Wheeler Rebecca R Bae Jinhye J

ACS materials letters 20231024 11

Analyzing hydrogel microstructure through scanning electron microscopy (SEM) images is crucial in understanding hydrogel properties. However, the analysis of SEM images in hydrogel research heavily relies on the intuition of individual researchers and is constrained by the limited size of the dataset. To address this, we propose SEMPro, a data-driven solution using web-scraping and deep learning (DL) to compile and analyze the structure-property relationships of hydrogels through SEM images. It ...[more]

PMID: 37969140

Similar Datasets

Project description:Pathological impairment of elastic fiber and other extracellular matrix (ECM) components are described for the aortic media of ascending thoracic aortic aneurysms (aTAA) but the exact pathological impairment of the structure and its degree still needs further investigations. To evaluate the quantity and quality of elastic fiber sheets and other ECM structures (e.g. collagen), cells were removed from different types of aneurysmal tissues (tricuspid aortic valve [TAV] associated-, bicuspid aortic valve [BAV] associated-aneurysmal tissue and acute aortic dissections [AAD]) using 2.5% sodium hydroxide (NaOH) and compared to decellularized control aortic tissue. Likewise, native tissue has been analysed. To evaluate the 2D- (histological evaluation, fluorescence- and auto-fluorescence based staining methods) and the 3D structure (scanning electron microscopic [SEM] examination) of the medial layer we first analysed for a successful decellularization. After proving for successful decellularization, we quantified the amount of elastic fiber sheets, elastin and other ECM components including collagen. Aside from clearly visible focal elastic fiber loss in TAV-aTAA tissue, decellularization resulted in reduction of elastic fiber auto-fluorescence properties, which is perhaps an indication from a disease-related qualitative impairment of elastic fibers, visible only after contact with the alkaline solution. Likewise, the loss of collagen amount in BAV-aTAA and TAV-aTAA tissue (compared to non-decellularized tissue) after contact with NaOH indicates a prior disease-associated impairment of collagen. Although the amount of ECM was not changed in type A dissection tissue, detailed electron microscopic evaluation revealed changes in ECM quality, which worsened after contact with alkaline solution but were not visible after histological analyses. Apart from the improved observation of the samples using electron microscopy, contact of aneurysmal and dissected tissue with the alkaline decellularization solution revealed potential disease related changes in ECM quality which can partly be connected to already published data, but have to be proven by further studies.

Dataset Information

SEMPro: A Data-Driven Pipeline To Learn Structure-Property Insights from Scanning Electron Microscopy Images.

Publications

SEMPro: A Data-Driven Pipeline To Learn Structure-Property Insights from Scanning Electron Microscopy Images.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets