MOLAR: MRI-based Opportunistic Localization and Recognition of teeth
- 1University of South Carolina
Abstract
Tooth loss is associated with cognitive decline, yet dental status is rarely available in large neuroimaging cohorts, although routine T1-weighted (T1w) brain MRI captures the dentition. We present MOLAR (MRI-based Opportunistic Localization And Recognition of teeth), an open pipeline that isolates the dental region (the "bite box") from T1w brain MRI and counts the teeth, without additional scanning or commercial software. Three raters annotated the usable OASIS-3 bite boxes (1,261 rated by all three; 63,829 point markers; count intraclass correlation coefficient, ICC(2,1) = 0.909). We reformulate counting from point labels as instance segmentation: each tooth becomes a compact capsule whose in-plane footprint adapts to voxel adjacency, so a standard three-dimensional nnU-Net counts teeth as connected components. On held-out scans the model detected teeth more consistently than raters agreed with one another (F1 0.873 against 0.787) and, after leave-one-fold-out threshold calibration, counted with mean absolute error 1.96 teeth, recovering 86% of the gap between an uninformed floor (3.57) and the inter-rater ceiling (1.70). A second network replaces registration by predicting the dental region, changing counts by 1.06 teeth, less than two raters disagree (1.58), whereas classical registration failed on 18% of scans. In an independent cohort (Aging Brain Cohort; n = 233; different scanner and raters), automated counts reproduced the manual association with the Montreal Cognitive Assessment (MoCA; Pearson r = +0.372 against +0.414, both p < 0.001), and after adjustment for age, sex and race the association was equivalent across counting methods. The software will be released open source.
Graphical Abstract
O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=155 SRC="FIGDIR/small/26363455v1_ufig1.gif" ALT="Figure 1"> View larger version (74K): [email protected]@e6361eorg.highwire.dtl.DTLVardef@1eb2473org.highwire.dtl.DTLVardef@a620f5_HPS_FORMAT_FIGEXP M_FIG C_FIG
Related
- Opportunistic computed tomography to identify osteoporosis in primary hyperparathyroidism using sestamibi SPECT/CT scans
- Opportunistic computed tomography assessment of bone, muscle, and adipose tissue: evidence, limitations, and clinical translation
- Federated learning for privacy-preserving opportunistic screening of lumbar-spine DXA-defined osteoporotic-level bone mineral density using chest radiographs: A multi-center independent validation study
- Moving Toward Universal Risk Stratification: Globally Standardized Lipoprotein(a) in Molar Units
- Histopathological Spectrum and Yield of Advanced Neoplasia in Opportunistic Colonoscopy Without Organised Screening: A Single-Centre Cross-Sectional Study from Western Romania
BibTeX
@misc{newmannorlund2026molar,
title = {{MOLAR: MRI-based Opportunistic Localization and Recognition of teeth}},
author = {R. Newman-Norlund and R. Pallapothu and S. Kudaravalli and J. Sridhar and P. Rajesh Kannan and S. Pallapothu},
howpublished = {medRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.18.26363455},
url = {https://www.medrxiv.org/content/10.64898/2026.09.18.26363455v1},
note = {CC BY-ND},
}
RIS
TY - UNPB
TI - MOLAR: MRI-based Opportunistic Localization and Recognition of teeth
AU - R. Newman-Norlund
AU - R. Pallapothu
AU - S. Kudaravalli
AU - J. Sridhar
AU - P. Rajesh Kannan
AU - S. Pallapothu
PY - 2026
DA - 2026/09/18
PB - medRxiv
DO - 10.64898/2026.09.18.26363455
UR - https://www.medrxiv.org/content/10.64898/2026.09.18.26363455v1
AB - Tooth loss is associated with cognitive decline, yet dental status is rarely available in large neuroimaging cohorts, although routine T1-weighted (T1w) brain MRI captures the dentition. We present MOLAR (MRI-based Opportunistic Localization And Recognition of teeth), an open pipeline that isolates the dental region (the "bite box") from T1w brain MRI and counts the teeth, without additional scanning or commercial software. Three raters annotated the usable OASIS-3 bite boxes (1,261 rated by all three; 63,829 point markers; count intraclass correlation coefficient, ICC(2,1) = 0.909). We reformulate counting from point labels as instance segmentation: each tooth becomes a compact capsule whose in-plane footprint adapts to voxel adjacency, so a standard three-dimensional nnU-Net counts teeth as connected components. On held-out scans the model detected teeth more consistently than raters agreed with one another (F1 0.873 against 0.787) and, after leave-one-fold-out threshold calibration, counted with mean absolute error 1.96 teeth, recovering 86% of the gap between an uninformed floor (3.57) and the inter-rater ceiling (1.70). A second network replaces registration by predicting the dental region, changing counts by 1.06 teeth, less than two raters disagree (1.58), whereas classical registration failed on 18% of scans. In an independent cohort (Aging Brain Cohort; n = 233; different scanner and raters), automated counts reproduced the manual association with the Montreal Cognitive Assessment (MoCA; Pearson r = +0.372 against +0.414, both p < 0.001), and after adjustment for age, sex and race the association was equivalent across counting methods. The software will be released open source. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=155 SRC="FIGDIR/small/26363455v1_ufig1.gif" ALT="Figure 1"> View larger version (74K): [email protected]@e6361eorg.highwire.dtl.DTLVardef@1eb2473org.highwire.dtl.DTLVardef@a620f5_HPS_FORMAT_FIGEXP M_FIG C_FIG
ER -