New AI- and liquid-biopsy-adjacent tools are being positioned to improve cancer risk stratification and diagnostic throughput. City of Hope and UC Berkeley researchers reported a microfluidic platform, mechano-NPS, that squeezes individual mammary epithelial cells to generate “mechanical age” signals and estimate breast cancer susceptibility using machine learning classifiers (MechanoAge and Mechano-RISQ). Elsewhere in the pipeline, Myriad Genetics presented updates around broader precise MRD efforts in breast cancer using ultrasensitive ctDNA, with researchers discussing how ctDNA can inform relapse risk and treatment escalation or de-escalation decisions. Together, the updates show a continuing shift toward functional readouts and ultra-sensitive molecular measurements that can be scaled beyond traditional imaging and single-marker assays—if clinical validation supports utility across patient populations.