A comparison of stereological and computer-assisted histomorphometric analysis as tools for histological...

This study was designed to compare computer-assisted histomorphometric analysis (CAHA) and stereology (STER) as measurement tools for evaluating the repair response during periodontal wound healing. Thirty-six histological sections derived from 4 surgically created defects in the furcation of mandibular second premolars of sheep were measured by each technique to determine the furcation area and volume, and the percentage of new bone formation at 7 wk postoperatively. Slides were viewed in random order with the source unknown to the examiner (JL). One section from each of the 4 specimens was flagged for triplicate measurement by each technique. Intraexaminer error was determined to be low as the coefficient of variation in each of the 2 techniques was between 1% and 4%. A consistently higher percentage of bone was identified using stereology. The coefficient of agreement was plotted to determine how closely these 2 techniques were matched in their respective estimations of bone fill in a furcation defect. This analysis revealed statistical bias between the 2-techniques and a low degree of agreement between them. This study demonstrates that the 2 techniques are not interchangeable. It also emphasizes that the reader must be cautious when comparing results from studies in which different systems of measurement and analysis have been used. Stereology was determined to be the measurement tool of choice due to its high degree of reproducibility, ease of use and efficient use of time.