Presentation Authors: Hiroshi Sasaki*, Hiroyuki Inaba, Takahiro Kimura, Koki Morishita, Shin Egawa, Tokyo, Japan
Introduction: Hyperspectral imaging (HSI) is an emerging modality used to obtain spatially resolved spectral images that provides diagnostic information on tissue physiology and morphology. The hyperspectral image data of prostate cancer tissues from prostatectomy specimen were analyzed to determine its clinical utility.
Methods: After obtaining the institutional review board&[prime]s approval, 22 prostate cancer specimens obtained from radical prostatectomy were evaluated. A randomly selected square of 5×5 mm pixel was identified on each specimen slices. We evaluated 209 pixels from 132 cancer and 77 normal areas. The prostatic glands were sliced at 4mm intervals perpendicular to the urethra. The HSI system was used to obtain hyperspectral images of the specimens. This system is capable of obtaining reflectance images with a size of 1920Ã—1440 over 400 to 700nm (Î´Î»=15 nm). CIELAB values represents colors using the coordinates in a uniform color space consisting of lightness variable L* and chromaticity indices a* and b*. CIELAB values are calculated as ([L*]2+[a*]2+[b*]2)1/2. The CIELAB of malignant and normal areas were calculated, respectively, and plotted in the color space. The association between CIELAB and tumor characters such as Gleason score (GS) classification was determined.
Results: Higher GS of the cancer tissue was associated with a relatively high-intensity wavelength range. The mean CIELAB data were not significantly different between the normal areas and GS 3+3 tumors (p=0.41). However, the mean CIELAB data of GS 3+4â‰¥tumors were significantly higher than those of GS 3+3 tumors (meanÂ±SD, 67.5Â±1.1 [GS 3+4] vs 64.2Â±0.96 [GS 3+3], p=0.03). Regarding the Gleason pattern 4 morphology, the glomeruloid type CIELAB was significantly lower than those in other morphological patterns (p=0.01).
Conclusions: This study indicates that HSI evaluation using CIELAB values could be used to further stratify malignant potentials of prostate cancer. This pilot study may lead to advances in the optical diagnosis of pathological morphology for prostate cancer.
Source of Funding: This research was supported by Olympus Corporation.