Limits...
Biopsy Diagnosis of Oral Carcinoma by the Combination of Morphological and Spectral Methods Based on Embedded Relay Lens Microscopic Hyperspectral Imaging System.

Ou-Yang M, Hsieh YF, Lee CC - J Med Biol Eng (2015)

Bottom Line: Hence, the proposed cocktail approach is used to determine the effectiveness of the spectral methods in correlating with the sampling conditions.And then we use a combination of effective spectral methods according to the sample conditions for diagnosing a sample.The results demonstrate a sensitivity of 90 ± 4.53 % and a specificity of 87.8 ± 5.21 %.

View Article: PubMed Central - PubMed

Affiliation: Department of Electrical and Computer Engineering, National Chiao-Tung University, 1001 Ta-Hsueh Rd., Hsinchu, Taiwan.

ABSTRACT

Cytopathological examination through biopsy is very important for carcinoma detection. The embedded relay lens microscopic hyperspectral imaging system (ERL-MHIS) provides a morphological image of a biopsy sample and the spectrum of each pixel in the image simultaneously. Based on the ERL-MHIS, this work develops morphological and spectral methods to diagnose oral carcinoma biopsy. In morphological discrimination, the fractal dimension method is applied to differentiate between normal and abnormal tissues. In spectral identification, normal and cancerous cells are distinguished using five methods. However, the spectra of normal and cancerous cells vary with patient. The diagnostic performances of the five methods are thus not ideal. Hence, the proposed cocktail approach is used to determine the effectiveness of the spectral methods in correlating with the sampling conditions. And then we use a combination of effective spectral methods according to the sample conditions for diagnosing a sample. A total of 68 biopsies from 34 patients are analyzed using the ERL-MHIS. The results demonstrate a sensitivity of 90 ± 4.53 % and a specificity of 87.8 ± 5.21 %. Furthermore, in our survey, this system is the first time utilized to study oral carcinoma biopsies.

No MeSH data available.


Related in: MedlinePlus

Biopsy image of patient 7. a Transmission image of normal tissue. Fluorescence images of normal tissue under (b) F1 and (c) F2 excitation. d Binary image of normal tissue for calculating fraction dimension. e Transmission image of cancerous tissue. Fluorescence images of cancerous tissue under (f) F1 and (g) F2 excitation. h Binary image of cancerous tissue for calculating fraction dimension
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4551557&req=5

Fig2: Biopsy image of patient 7. a Transmission image of normal tissue. Fluorescence images of normal tissue under (b) F1 and (c) F2 excitation. d Binary image of normal tissue for calculating fraction dimension. e Transmission image of cancerous tissue. Fluorescence images of cancerous tissue under (f) F1 and (g) F2 excitation. h Binary image of cancerous tissue for calculating fraction dimension

Mentions: Table 1 shows information on the biopsies. Figure 2 displays the biopsy image of patient 7. Pathologists can obtain more information about the biopsy (e.g., the differentiation and the stage) from the transmission images (Fig. 2a, e).Table 1


Biopsy Diagnosis of Oral Carcinoma by the Combination of Morphological and Spectral Methods Based on Embedded Relay Lens Microscopic Hyperspectral Imaging System.

Ou-Yang M, Hsieh YF, Lee CC - J Med Biol Eng (2015)

Biopsy image of patient 7. a Transmission image of normal tissue. Fluorescence images of normal tissue under (b) F1 and (c) F2 excitation. d Binary image of normal tissue for calculating fraction dimension. e Transmission image of cancerous tissue. Fluorescence images of cancerous tissue under (f) F1 and (g) F2 excitation. h Binary image of cancerous tissue for calculating fraction dimension
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4551557&req=5

Fig2: Biopsy image of patient 7. a Transmission image of normal tissue. Fluorescence images of normal tissue under (b) F1 and (c) F2 excitation. d Binary image of normal tissue for calculating fraction dimension. e Transmission image of cancerous tissue. Fluorescence images of cancerous tissue under (f) F1 and (g) F2 excitation. h Binary image of cancerous tissue for calculating fraction dimension
Mentions: Table 1 shows information on the biopsies. Figure 2 displays the biopsy image of patient 7. Pathologists can obtain more information about the biopsy (e.g., the differentiation and the stage) from the transmission images (Fig. 2a, e).Table 1

Bottom Line: Hence, the proposed cocktail approach is used to determine the effectiveness of the spectral methods in correlating with the sampling conditions.And then we use a combination of effective spectral methods according to the sample conditions for diagnosing a sample.The results demonstrate a sensitivity of 90 ± 4.53 % and a specificity of 87.8 ± 5.21 %.

View Article: PubMed Central - PubMed

Affiliation: Department of Electrical and Computer Engineering, National Chiao-Tung University, 1001 Ta-Hsueh Rd., Hsinchu, Taiwan.

ABSTRACT

Cytopathological examination through biopsy is very important for carcinoma detection. The embedded relay lens microscopic hyperspectral imaging system (ERL-MHIS) provides a morphological image of a biopsy sample and the spectrum of each pixel in the image simultaneously. Based on the ERL-MHIS, this work develops morphological and spectral methods to diagnose oral carcinoma biopsy. In morphological discrimination, the fractal dimension method is applied to differentiate between normal and abnormal tissues. In spectral identification, normal and cancerous cells are distinguished using five methods. However, the spectra of normal and cancerous cells vary with patient. The diagnostic performances of the five methods are thus not ideal. Hence, the proposed cocktail approach is used to determine the effectiveness of the spectral methods in correlating with the sampling conditions. And then we use a combination of effective spectral methods according to the sample conditions for diagnosing a sample. A total of 68 biopsies from 34 patients are analyzed using the ERL-MHIS. The results demonstrate a sensitivity of 90 ± 4.53 % and a specificity of 87.8 ± 5.21 %. Furthermore, in our survey, this system is the first time utilized to study oral carcinoma biopsies.

No MeSH data available.


Related in: MedlinePlus