Back

Age-Adjusted Cerebellar Volumetry and Developmental Outcomes in Extremely Low Birth Weight Infants

Yao, L.; Brown, A. M.; Park, J.; Lazar, S.; Kralik, S.; Rhee, C. J.; Sillitoe, R. V.; Gill, J. S.

2026-10-01 neuroscience
10.64898/2026.09.29.755417 bioRxiv
Show abstract

Background: Infants with extremely low birth weight (ELBW) are at high risk for neurodevelopmental impairment, but the relationship between cerebellar structure and later developmental outcome remains incompletely defined. Methods: In this retrospective single-center study, we examined total cerebellar volume on clinically acquired brain magnetic resonance imaging (MRI) in ELBW infants while accounting for postmenstrual age (PMA) at imaging. Results: Of 65 infants screened, 42 were included. Mean gestational age at birth was 25.54 {+/-} 1.68 weeks, mean birth weight was 733.2 {+/-} 118.7 g, and mean PMA at MRI was 43.02 {+/-} 6.53 weeks (range, 31.57--68.00). Mean cerebellar volume was 22.17 {+/-} 9.63 cm. Independently adjudicated developmental outcomes comprised normal development in 16 infants, isolated or specific impairment in 12, and moderate/severe impairment in 14. Unadjusted cerebellar volumes did not differ across developmental groups (P = 0.490) or qualitative MRI categories (P = 0.875). However, the PMA-by-developmental-category interaction was significant (P = 0.002), with estimated cross-sectional slopes of 1.72, 1.56, and 0.39 cm per week, respectively. The interaction remained significant within 36--46 weeks PMA (n = 29; P = 0.011) and after adjustment for small-for-gestational-age (SGA) status. Sensitivity analyses indicated uncertainty in the stability of the interaction. Conclusions: These findings suggest that age-informed cerebellar volumetric patterns may provide developmental information not evident from average volumes alone and support prospective validation with standardized imaging and developmental assessments. As a result, cerebellar volumes obtained early in life may hold prognostic information for ELBW infants.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.