A Machine Learning Approach to Quantitative Analysis of Enamel Microstructure from Scanning Electron Microscopy Images DOI Creative Commons
Carli Marsico,

Cameron Renteria,

Jack Grimm

et al.

Small Structures, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 25, 2024

Dental enamel, the outermost tissue of mammalian teeth, must withstand a lifetime wear and cyclic contact. To meet this demand, enamel possesses combination high hardness resistance to fracture, properties that are typically mutually exclusive. The impressive damage tolerance has been attributed largely decussation rods, principal unit its microstructure. As such, is inspiring design next‐generation structural materials. However, quantitative descriptions decussated rod microstructure remain limited due challenges encountered in applying computed tomography acquiring quality images appropriate for traditional digital processing methods. Here, machine learning segmentation method applied obtained using scanning electron microscopy support analysis A pretrained convolutional neural network used expand input training image dataset allow random forest classifier, which ultimately segments with very small set ( n = 3 images). validation presented, addition application calculate relevant microstructural parameters tooth from selected species. methodology here equally applicable other hard tissues.

Language: Английский

Characterizing the microstructures of mammalian enamel by synchrotron phase contrast microCT DOI
Carli Marsico, Jack Grimm,

Cameron Renteria

et al.

Acta Biomaterialia, Journal Year: 2024, Volume and Issue: 178, P. 208 - 220

Published: Feb. 28, 2024

Language: Английский

Citations

4
Quantifying structural changes in organised biomineralized surfaces using synchrotron Polarisation-induced Contrast X-ray Fluorescence DOI Creative Commons

Hui Lynn Ooi,

Alexander P. Morrell, Aaron R. H. LeBlanc

et al.

Acta Biomaterialia, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 1, 2025

The quantitative characterization of the structure biomineral surfaces is needed for guiding regenerative strategies. Current techniques are compromised by a requirement extensive sample preparation, limited length-scales, or inability to repeatedly measure same surface over time and monitor structural changes. We aim address these deficiencies developing Calcium (Ca) K-edge Polarisation Induced Contrast X-ray Fluorescence (PIC-XRF) quantify hydroxyapatite (HAp) crystallite arrangements in high low textured surfaces. Minimally prepared human dental enamel was used as an exemplar initial structures, disruption caused short dietary acid exposures. By measuring at different rotational angles relative polarised focused (2x2µm) monochromatic source (at either 4049.2 4051.1 eV) it possible discriminate principal secondary orientations crystallites, along with their texture. It also organisation crystallites both (enamel cross-sections) highly (facial enamel) including identification aligned perpendicular surface-a challenge other synchrotron techniques. Surface modifications following term erosion (affecting <20µm depth) were detected significant shifts orientation (p<0.001) marked reduction texture (p<0.001). Findings suggest preferential dissolution HAp based on angular orientation. demonstrate that PIC-XRF powerful tool surfaces, minimal preparation enables monitoring changes through repeated measurements. STATEMENT OF SIGNIFICANCE: This study introduces method quantifying addressing limitations existing require cannot surface. using minimally enamel, successfully discriminated between end-on crystallites-a methods. Additionally, due short-term erosion. technique's potential non-invasively analyze offers new opportunities understanding dynamics treatments.

Language: Английский

Citations

0
Challenges of Studying Amelogenesis in Gene-Targeted Mouse Models DOI Open Access

Charles E. Smith,

John D. Bartlett, James P. Simmer

et al.

International Journal of Molecular Sciences, Journal Year: 2025, Volume and Issue: 26(10), P. 4905 - 4905

Published: May 20, 2025

Research on how a stratified oral epithelium gained the capability to create hardest hydroxyapatite-based mineralized tissue produced biologically protect surfaces of teeth has been ongoing for at least 175 years. Many advances have made in unraveling some key factors that allowed innermost undifferentiated epithelial cells sitting skin-type basement membrane transform into highly polarized capable forming and controlling mineralization extracellular organic matrix becomes enamel. Genetic manipulation mice proven be useful approach studying specific events amelogenesis developmental sequence but there pitfalls interpreting loss function data caused part by conflicting literature, technical problems preservation, total amount time spent tooth development between different species led equivocal conclusions. This critical review attempts discuss these issues highlight challenges characterizing gene-targeted mouse models.

Language: Английский

Citations

0
A Machine Learning Approach to Quantitative Analysis of Enamel Microstructure from Scanning Electron Microscopy Images DOI Creative Commons
Carli Marsico,

Cameron Renteria,

Jack Grimm

et al.

Small Structures, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 25, 2024

Dental enamel, the outermost tissue of mammalian teeth, must withstand a lifetime wear and cyclic contact. To meet this demand, enamel possesses combination high hardness resistance to fracture, properties that are typically mutually exclusive. The impressive damage tolerance has been attributed largely decussation rods, principal unit its microstructure. As such, is inspiring design next‐generation structural materials. However, quantitative descriptions decussated rod microstructure remain limited due challenges encountered in applying computed tomography acquiring quality images appropriate for traditional digital processing methods. Here, machine learning segmentation method applied obtained using scanning electron microscopy support analysis A pretrained convolutional neural network used expand input training image dataset allow random forest classifier, which ultimately segments with very small set ( n = 3 images). validation presented, addition application calculate relevant microstructural parameters tooth from selected species. methodology here equally applicable other hard tissues.

Language: Английский

Citations

0