Virtual Spectral Decomposition with Dendritic Binary Gating Detects Pancreatic Cancer Tissue Transformation on Standard CT: Multi-Institutional Validation Across Three Independent Datasets with Evidence of Temporally Stable Pre-Pathological Signal

ABSTRACT Background Pancreatic ductal adenocarcinoma (PDAC) has a five-year survival rate of approximately 12%, largely because it is typically diagnosed at an advanced stage. CT-based computational methods for early detection exist but rely on black-box deep learning or large texture feature sets without tissue-specific interpretability. A method that decomposes standard CT into named tissue-component channels with full explainability would bridge the gap between computational detection and clinical understanding. Methods We developed Virtual Spectral Decomposition (VSD), which applies six parameterized sigmoid functions S(HU) = 1/(1+exp(−α(HU−μ))) to standard portal-venous CT, decomposing each pixel into tissue-specific response channels for fat (μ=−60), fluid (μ=10), parenchyma (μ=45), stroma (μ=75), vascular (μ=130), and calcification (μ=250). Dendritic Binary Gating identifies structural content per channel using morphological filtering, enabling co-firing analysis and lone firer identification. A 25-feature signature was extracted per patient. Three independent datasets were analyzed: NIH Pancreas-CT (n=78 healthy), Medical Segmentation Decathlon Task07 (n=281 PDAC, paired tumor/adjacent tissue), and CPTAC-PDA from The Cancer Imaging Archive (n=82, multi-institutional, with DICOM time point tags encoding days relative to pathological diagnosis). The same six sigmoid parameters were used across all datasets without retraining. Results VSD achieved AUC 0.943 for field effect detection (healthy vs. cancer-adjacent parenchyma) and AUC 0.931 for patient-stratified tumor specification on MSD. On CPTAC-PDA, VSD achieved AUC 0.961 (6 features) and 0.979 (25 features) for distinguishing healthy from cancer-bearing pancreas. All significant features replicated across datasets in the same direction: z_fat (d=−2.10, p=3.5×10 −27 ), z_fluid (d=−2.76, p=2.4×10 −38 ), fire_fat (d=+2.18, p=1.2×10 −28 ). VSD severity showed no correlation with days-from-pathological-diagnosis (r=−0.008, p=0.944) across a range of day −1,394 to day +249, indicating a temporally stable tissue state. In an exploratory observation, patient C3N-01375, scanned 1,394 days (3.8 years) before pathological confirmation, showed VSD severity 2.8 standard deviations above the healthy mean. Conclusions VSD with Dendritic Binary Gating detects a stable pancreatic tissue composition signature on standard CT that is present on pre-pathological imaging, validated across three independent datasets without parameter adjustment. The six sigmoid channels map to biologically meaningful tissue components through a fully transparent interpretability chain. The temporal stability of the signal across available time points suggests that VSD detects an early, persistent tissue state rather than a progressively worsening process. If confirmed in prospective cohorts with truly incidental pre-diagnostic CTs, VSD could function as a single-scan screening tool applicable to abdominal CT performed during the pre-clinical window. Prospective longitudinal validation is the critical next step.