Objective: To support cancer screening and identify precancerous conditions, such as atypical hyperplasia, to improve cure rates and reduce mortality, by analyzing the performance of a previously confirmed procedure. Patients and Methods: We established 204 short-term blood-derived cell lines from cancer patients between 2013 and 2022.A dataset of phenotypic patterns, cytopathological variables, and proliferation profiles was used to train a neural network model.Comparative analysis of standard optical, functional, and Machine Learning-supported diagnosis was performed to verify reproducibility and clinical translatability.Results: Tumor heterogeneity was classified into seven phenotypic patterns (Pn1-Pn7); the Pn6-7 group showed an overall survival of 8 months (95% CI, 6-9).Among variables Vc1-Vc8, Vc3 (AUC=1, p35% cancer.AI-supported cytology showed positive and negative predictive values of 0.990.015and 10, compared to histopathological specimens (0.940.1 and 0.850.04).The algorithm achieved 10 sensitivity and 0.980.04specificity, with respect to traditional diagnosis (0.880.1 of specificity and 0.930.04 of sensitivity). Conclusion:The model demonstrated fast adaptive performance in predicting cancer risk and primary source assessment.The results suggest this screening model is sufficient to detect atypical hyperplasia compared to models based on oligo-analysis for single or double mutations.
Malara et al. (Wed,) studied this question.