Phase 1 Evidence
Biological Validation
Analytical validation with genetic disease models and endpoint mapping to published prenatal opioid exposure effects.
Key Conclusion
The observed clone-specific morphological differences are not random imaging noise, but reflect genotype-driven developmental trajectories that capture early neurodevelopmental impacts of disease-relevant mutations. These differences validate the use of automated morphological analysis to detect subtle but biologically meaningful phenotypes, demonstrating that ConductScreen can quantify real biological variation in organoid models.
Segmentation & Quality Control
Automated segmentation accuracy and image quality filtering ensure that downstream feature measurements are reliable.
Segmentation Performance
Median IoU
0.788
Intersection over Union
Median Dice
0.881
Dice similarity coefficient
vs. Otsu Baseline
6.5×
improvement over naive thresholding
Phase 2 target: IoU ≥ 0.85 with integration of U-Net-based deep learning segmentation.
Quality Control
Images Passing QC
1,361 / 1,407
96.7% pass rate — all 64 organoids retained
QC Thresholds
Sharpness: Laplacian variance ≥ 1.0
SNR: Signal-to-noise ratio ≥ 1.0
Batch Effects
Average ICC(1)
0.012
Negligible inter-lab batch effects
Clone vs. Lab Effects
7–14×
Clone effects larger than lab effects
Inter-lab variability is negligible relative to genotype-driven morphological differences, confirming that the pipeline measures biology rather than technical artifacts.
Preliminary Ethanol Signal
Re-analysis of Adams et al. (2023) ethanol exposure data demonstrates the pipeline's sensitivity to substance-induced morphological changes in an independent dataset.
Adams et al. 2023 Re-analysis
N = 4 control vs. N = 5 ethanol-exposed organoids
Perimeter
−20.8%
p = 0.049
Area
−28.9%
p = 0.085 (trend)
These preliminary results suggest that prenatal ethanol exposure produces detectable reductions in organoid size and boundary complexity — consistent with known teratogenic effects on cortical development.
Disease Model Biology
Each disease clone carries a mutation affecting a specific cellular pathway with predictable consequences for organoid morphology.
Alpha-Tubulin Mutant
Mutation
TUBA1A missense mutation (alpha-tubulin)
Biological Mechanism
Disrupted alpha-tubulin polymerization impairs cytoskeletal dynamics, leading to irregular cell packing and altered tissue architecture.
Expected Phenotype
Larger area, altered minor axis — consistent with disrupted cytoskeletal organization producing irregular, expanded tissue.
Observed Result
Larger area (|r| = 0.95), altered minor axis, perimeter, and solidity — all consistent with disrupted cytoskeletal organization.
Beta-Tubulin Mutant
Mutation
TUBB2A missense mutation (beta-tubulin)
Biological Mechanism
Reduced beta-tubulin function impairs microtubule dynamics, reducing proliferation and altering tissue organization.
Expected Phenotype
Smaller area, delayed growth — consistent with reduced proliferative capacity from beta-tubulin dysfunction.
Observed Result
Smaller area (|r| = −0.52), elevated texture contrast (|r| = 0.73), reduced circularity (|r| = −0.66) — matching predicted reduced proliferation and surface irregularity.
Tyrosine Hydroxylase Deficient
Mutation
TH deficiency (tyrosine hydroxylase)
Biological Mechanism
Impaired dopamine biosynthesis disrupts neuronal differentiation, leading to compensatory proliferation with heterogeneous cellular maturation.
Expected Phenotype
Largest area, high intensity variation — dysregulated neurogenesis produces rapid expansion with heterogeneous differentiation.
Observed Result
Largest area of all clones (|r| = 1.0), highest intensity variation (|r| = 0.94), fastest growth (14.4× over 30 days) — consistent with dysregulated neurogenesis and heterogeneous differentiation.
Feature-to-Biology Mapping
The measured features are quantitative proxies for underlying biological processes. Each feature category captures a distinct aspect of organoid biology.
Morphology
Size and shape features (area, perimeter, circularity, axes) reflect tissue growth rates, proliferative capacity, and cytoskeletal integrity. Mutations affecting microtubule components directly alter these measurements by disrupting cell packing and tissue self-organization.
Intensity
Brightness statistics (mean, standard deviation, range) capture cellular differentiation heterogeneity and tissue thickness. High intensity variation indicates mixed cell populations at different maturation stages — a hallmark of impaired neuronal differentiation.
Texture
Gradient-based texture descriptors (contrast, energy, homogeneity, correlation) measure local structural patterns within the organoid. These reflect internal tissue architecture, including boundaries between cell layers, structural complexity, and long-range organization.
Summary: Area/size → tissue growth and proliferation rates. Shape (axes, eccentricity) → tissue organization and cytoskeletal integrity. Intensity variation → cellular differentiation heterogeneity and structural complexity.
Developmental Impact
These genetic differences translate into developmental consequences observable in vitro through automated growth tracking.
Timing of Growth
Clone-specific growth trajectories
wt2D: Normal proliferation kinetics with steady expansion (4.7× over 30 days), establishing the baseline growth trajectory for comparison.
A1A: Alpha-tubulin disruption produces intermediate expansion (7.5×), suggesting partially impaired but still active proliferation from cytoskeletal dysfunction.
B2A: Beta-tubulin dysfunction results in the smallest organoids at every timepoint despite 10.9× fold-change, consistent with reduced proliferative capacity.
TH2: Tyrosine hydroxylase deficiency drives the fastest expansion (14.4×, 3× faster than wildtype), suggesting dysregulated neurogenesis with compensatory proliferation.
Morphological Heterogeneity
Shape and intensity variation
Variations in area, shape (minor/major axes), and intensity reflect differences in tissue architecture. In TUBA1A and TUBB2A clones, these correspond to microtubule-guided neuronal migration and packing density. For TH2, intensity heterogeneity reflects uneven differentiation into catecholaminergic neurons.
While effect sizes are small-to-medium for some features, they are statistically significant and biologically meaningful, representing the cumulative impact of these mutations on progenitor proliferation, cortical layering, and cytoskeletal organization.
Prenatal Opioid Exposure: Endpoint Mapping
Published organoid studies demonstrate that prenatal opioid exposure produces measurable morphological changes in human neural organoids — changes that map directly to the features the morphology module quantifies.
Methadone
Primary MOUD compound
Yao et al. (2020) reported dose-dependent growth arrest, tissue disintegration, and suppressed neural network activity. Kallupi et al. (2023) found altered synaptogenesis gene programs. These effects correspond to morphology module features: area (growth arrest), circularity and solidity (tissue disintegration), and texture homogeneity (structural disruption).
Buprenorphine
Alternative MOUD; widely prescribed in pregnancy
Nieto Estévez et al. (2022) reported altered interneuron migration into cortical spheroids via the nociceptin receptor. Ho et al. (2024) found cell-type-specific transcriptional responses. Migration changes measurable as changes in texture features (contrast, homogeneity) and intensity distribution (std_intensity).
Fentanyl
High-potency opioid; chronic exposure and withdrawal
Lin et al. (2024) modeled acute and chronic fentanyl exposure in iPSC-derived midbrain organoids with dopaminergic neurons. Impaired neuronal maturation corresponds to morphology module features: area (growth trajectory changes), solidity (tissue compactness), and std_intensity (differentiation heterogeneity).
Convergence of Endpoints
The features validated in our disease-model analysis — area, boundary metrics, solidity, texture, and intensity variability — are the same endpoints that published opioid-organoid studies report as affected. This convergence establishes that ConductScreen's validated analytical pipeline measures endpoints directly relevant to prenatal opioid exposure biology.
| Published Opioid Effect | Morphology Feature | Validated? | Reference |
|---|---|---|---|
| Growth arrest / halted proliferation | Area, Equiv. Diameter | Yes (all 3 clones) | Yao et al., 2020 |
| Tissue disintegration / boundary collapse | Circularity, Solidity | Yes (Solidity: all 3) | Yao et al., 2020 |
| Disrupted tissue organization | Contrast, Homogeneity | Yes (A1A, B2A) | Kallupi et al., 2023 |
| Differentiation heterogeneity | Std Intensity | Yes (TH2) | Ho et al., 2024 |
| Impaired neuronal maturation | Area trajectory shift | Yes (growth curves) | Lin et al., 2024 |
| Altered interneuron migration | Texture pattern changes | Yes (B2A texture) | Nieto Estévez et al., 2022 |