Objective: To evaluate the capacity of semiquantitative striatal [¹²³I]FP-CIT SPECT-derived metrics to improve the clinical differentiation of degenerative parkinsonism through an integrated machine learning model.

Background: Differentiating Parkinson’s disease (PD) from atypical parkinsonian syndromes (APS) remains challenging because of overlapping motor features. Although [¹²³I]FP-CIT SPECT is widely used to assess presynaptic dopaminergic dysfunction, its role in distinguishing degenerative parkinsonian subtypes remains unclear.

Method: This cross-sectional study included 487 patients with PD and 219 with APS, including 127 progressive supranuclear palsy (PSP), 37 multiple system atrophy parkinsonian type (MSA-P), 12 multiple system atrophy cerebellar type (MSA-C), and 43 corticobasal degeneration (CBD). All participants underwent a [¹²³I]FP-CIT SPECT scan. Striatal [¹²³I]FP-CIT uptake was quantified using anatomical (caudate, putamen, ventral striatum) and functional (limbic, executive, sensorimotor) parcellations to calculate specific binding ratios, asymmetry indices, and inter-regional ratios. Discriminative performance of each metric was evaluated using receiver operating characteristic (ROC) curves analyses. A random forest classifier integrating all semiquantitative metrics was trained and validated, enabling data-driven identification of clinically meaningful diagnostic pathways.

Results: The caudate-to-posterior putamen and sensorimotor-to-limbic inter-regional ratios showed the strongest discriminative performance (AUC up to 0.89) for differentiating PD from APS. The random forest model achieved a classification error rate of 8.6% and revealed two diagnostic pathways. The first, driven by higher sensorimotor-to-limbic ratios and preserved posterior putaminal uptake, primarily classified MSA-C and CBD, with asymmetry in cognitive striatal subregions further distinguishing CBD from MSA-C. The second pathway, characterized by lower sensorimotor-to-limbic ratios and reduced caudate-to-posterior putamen ratios, mainly differentiate PD from PSP.

Conclusion: Integrating anatomical and functional [¹²³I]FP-CIT SPECT metrics within a machine learning framework enhances the clinical differentiation of degenerative parkinsonisms and supports [¹²³I]FP-CIT SPECT as a robust in vivo disease biomarker.

References: This abstract has also been submitted to the 42nd Congress of the Spanish Society of Nuclear Medicine and Molecular Imaging (SEMNIM), to be held in Córdoba, Spain, on 10-12 June 2026, and is currently pending acceptance.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/semiquantitative-%c2%b9%c2%b2%c2%b3ifp-cit-spect-metrics-combined-with-machine-learning-improve-clinical-differentiation-of-parkinsons-disease-and-atypical-parkinsonian-syndrome/