Biometric parameters and computer assisted analysis of pixel distribution on B-mode testicular ultrasonogram in young and sexually mature rams

This study was designed to investigate the relationships between basic biometric parameters and pixel distribution in predetermined regions of B-mode testicular ultrasonogram in young and sexually mature rams by computer assisted analysis. The animals were separated in two groups: group I (young rams, aged 5-8 months; n=6) and group II (sexually mature rams, aged 14-18 months; n=6). Biometric parameters as age, body weight and scrotal circumference were determined by routine methods. A trans-scrotal ultrasonography of both testes was performed with 7 MHz linear transrectal probe by the same operator with focus, gain and brightness values kept constant during the study. The ultrasound images were obtained in a longitudinal view plane and frozen when visualisation of the testicular mediastinum was clear and apparent. All images were transferred to a computer, converted to gray-scale (0-255 pixels) and submitted to pixel distribution analysis using Image ProPlus 7.0 analytical software and option range statistics. The pixel distribution analysis in the same region for left and right testes was based on the pixel ranges classification for different biological tissues (blood, lipids, muscle, fibrous and calcium) and expressed as colour area in percentages. The arithmetic mean of values for the left and right testis was accepted as a final value for each ram. The data for both groups were statistically processed, compared and the relationships between the biometric parameters and different pixel ranges were determined. Significant differences (P<0.05) among the groups were recorded for all biometric parameters and the lipid area only. The biometric parameters correlated positively (R≥0.92; P<0.05) each with the other, while their relationship (R≤ -0.94; P<0.05) with the lipid area and the correlation between muscle and fibrous area were negative (R= -0.92; P<0.05). In conclusion, the pixel distribution analysis of a predetermined region of B-mode testicular ultrasonogram based on pixel ranges classification for different biological tissues can be used for detection of testicular parenchyma changes in rams. Future comparative histologic and ultrasound investigations are needed to determine correct pixel values for different biological tissues in the testes of young and sexually mature rams.