Cell morphology measurement and statistical analysis using Python libraries. [Not invited]
村上龍太; 千葉陽一; 宮井由美; 松本晃一; 村山繁雄; 上野正樹
第114回日本病理学会総会 2025/04 仙台 古川徹
Hypertrophy of choroid plexus epithelial cells in diseases has been reported
via MRI, but histological studies are limited. In this study, we analyzed t
issue samples from 20 human cases to examine correlations between epithelial
cell morphology and age or disease. We measured the area, long axis, short
axis, and the epithelial cell ratio within the total choroid plexus, evaluat
ing 10 fields per case and a total of 4,846 cells. Python and its libraries
were used to efficiently process the data: OpenCV for automatic cell contour
detection and NumPy/Pandas for statistical analysis. Results showed a signi
ficant positive correlation between age and cell morphology, but no clear di
fferences among diseases. Our findings highlight Python’s capability for ef
ficient, reproducible morphology analysis, significantly reducing researcher
s' workload and opening new avenues in digital pathology. This study is unde
r submission, and this presentation will detail the methods used.