Characterisation and differentiation of dentine caries using optical coherence tomography / Saad Ahmed Khan

Saad Ahmed , Khan (2016) Characterisation and differentiation of dentine caries using optical coherence tomography / Saad Ahmed Khan. PhD thesis, University of Malaya.

[img] PDF (The Candidate's Agreement)
Restricted to Repository staff only

Download (1533Kb)
    PDF (Thesis PhD)
    Download (8Mb) | Preview


      Introduction: It is clinically challenging to differentiate caries-infected dentine (ID), caries-affected dentine (AD) and healthy dentine (HD), especially during caries removal. The main aim of this study is to explore the potential of an optical method, Optical Coherence Tomography (OCT), in assisting caries removal decision making in a clinical setting. The initial step towards this aim is to characterise OCT signals in the carious dentine and identify OCT outcome measures that can discriminate these layers with high accuracy. Objectives: Firstly, to optically characterise these layers based on their microstructure, to develop and determine the OCT outcome measures that can differentiate the dentine layers and assess their accuracy of prediction. Secondly, to determine the relationship of OCT outcome measures with the quantity of major elements in carious and healthy dentine. Finally, to compare and explore the relationship of the mineral density (MD) and OCT outcome measures. Methods: Swept Source OCT (SS-OCT) was used to scan cross-sectioned natural dentine caries surfaces and validated against other reference methods. The microstructure, composition and MD characterisation were performed using field emission scanning electron microscopy (FESEM), energy dispersive X-ray spectroscopy (EDX) and micro computed tomography (micro-CT). A bespoke software developed with MATLAB (Mathworks Inc) was used for OCT data extraction and processing. Results: Characterisation of OCT signals based on the microstructure of the carious and healthy dentine showed that the OCT backscattered intensity, I, attenuated in two phases where it attenuated faster in the first phase relative to the second. When I of the top 100 μm physical depth was considered as a whole, it was observed to attenuate following a polynomial pattern in ID (R2 = 0.794), while it assumed an exponential attenuation in AD and HD (R2 = 0.783 and 0.998) respectively. Maximum Intensity (Imax), and area under the curve (AUC), AUCT were the two OCT outcome measures that was able to differentiate the three dentine layers. Artificial neural network (ANN) showed higher accuracy for classifying dentine caries when using Imax and AUC combined, ID and AD, AD and HD and ID and HD were correctly classified with 84%, 88% and 94% accuracy respectively. Ca and P wt% were significantly different between all three dentine layers (p˂0.05), except the Ca:P ratio between the AD and HD (p= 0.460). Imax demonstrated moderate positive relationship with Ca (r= 0.448) and P (r= 0.479), and moderate negative relationship with Ca:P ratio (r= -0.431). Mean MD of ID, AD and HD in this study was found to be 0.87 (± 0.10), 1.16 (± 0.82) and 1.52 (± 0.13) gm/cm3 respectively. As a whole, there was a moderate but significant inverse correlation between MD and ΔI (r=-0.531). The mean ΔI for ID, AD and HD were significantly different between ID and AD, and ID and HD (p˂ 0.05), except AD and HD (p=0.851). ΔI discriminated ID and AD with 90.5% sensitivity and 81.0% specificity. Conclusions: The attenuation characteristics of OCT backscattered intensity characterisation and the derived outcome measure can significantly differentiate and be used for predicting dentine layers with high accuracy.

      Item Type: Thesis (PhD)
      Additional Information: Thesis (PhD) – Faculty of Dentistry, University of Malaya, 2016.
      Uncontrolled Keywords: Optical coherence tomography; Caries-infected dentine (ID); Caries-affected dentine (AD); Swept Source OCT (SS-OCT)
      Subjects: R Medicine > RK Dentistry
      Divisions: Faculty of Dentistry
      Depositing User: Mr Mohd Safri Tahir
      Date Deposited: 26 Aug 2019 02:11
      Last Modified: 26 Aug 2019 02:12

      Actions (For repository staff only : Login required)

      View Item