Tissue multifractality and hidden Markov model based integrated framework for optimum precancer detection
- 0 Collaborators
We report the application of a hidden Markov model (HMM) on multifractal tissue optical properties derived via the Born approximation-based inverse light scattering method for effective discrimination of precancerous human cervical tissue sites from the normal ones. Two global fractal parameters, generalized Hurst exponent and the corresponding singularity spectrum width, computed by multifractal detrended fluctuation analysis (MFDFA), are used here as potential biomarkers. We develop a methodology that makes use of these multifractal parameters by integrating with different statistical classifiers like the HMM and support vector machine (SVM). It is shown that the MFDFA-HMM integrated model achieves significantly better discrimination between normal and different grades of cancer as compared to the MFDFA-SVM integrated model. ...learn more
Project status: Published/In Market
Overview / Usage
We have explored an integrated framework of light scatteringderived multifractal tissue optical properties (generalized Hurst
exponent and width of singularity spectrum) along with a robust HMM for multiclass classification of different precancerous
grades of human uterine cervix. The results clearly demonstrate that the use of HMM on the multifractal properties leads to significantly improved classification as compared to MFDFA-SVM based integrated model for multiclass classification. These
MFDFA-HMM based classification results show considerable promise by exploring multifractal tissue optical properties as
a biomarker for precancer detection. We are currently expanding our investigations toward in-vivo deployment of this integrated
approach for precancer detection using tissue light scattering spectra. In general, the use of this MFDFA-HMM integrated
model on elastic scattering spectroscopic data may lead to a diagnostic modality for the detection of other types of cancer.
Methodology / Approach
Here, we demonstrate the efficacy of MFDFA-HMM integrated framework for optical diagnosis of cancer. More specifically, it is found that HMM on the multifractal light scattering properties of the tissues shows remarkable efficiency in differentiating normal and different stages of precancer. It is particularly effective when applied on global fractal parameters like generalized Hurst exponent and the corresponding singularity spectrum width/strength of multifractality, characterizing the global morphological conditions of the tissues for multiclass cancer classification, as compared to the MFDFA-SVM integrated framework under same application.
Technologies Used
The spatial distribution of tissue refractive index (RI) was recorded by a differential interference contrast (DIC) microscope (Olympus IX-81, United States). At a magnification of 60×, these DIC images were recorded by a CCD camera (ORCA-ERG, Hamamatsu, 1344 × 1024 pixel dimension 6.45 μm). The elastic scattering spectra from the multiple sites of the biopsied tissue sections were recorded by the angle resolved spectral light scattering measurements. In brief, light emitted from a Xe-lamp (HPX-2000, Ocean Optics, United States) was collimated by a combination of lenses and illuminated the tissue sample at the center of a goniometric arrangement (spot size ∼1-mm-diameter). The collimated scattered light from the sample was focused into a collecting fiber probe coupled to a spectrometer (USB4000FL, Ocean Optics) for wavelength resolved signal detection.The recordings of spectra were performed (360 to 800 nm) with a spectral resolution of 2.05 nm, where the angular range was kept at
10 deg to 150 deg with an interval of 10 deg multifractal study, the spectra were recorded at backscattering angle θ = 150 deg. During data analysis, codes were written in matlab.