Before We Can Live Longer, We Need to Diagnose Better
This piece was first published in Meditech Today.
Much of the longevity conversation focuses on what we might take, change or invest in to live longer. Before any of that, we need to get better at finding problems while there is still time to act. My first 3D breast ultrasound, at BeSound, made me think about how we should judge progress in diagnostics: by the evidence, not only by how the appointment feels.
AI is already making a measurable difference. In Sweden's MASAI trial, AI-supported mammography screening found 6.4 cancers per 1,000 women versus 5.0 with standard double reading, and cut screen-reading workload by 44%. More detection is not the same as fewer deaths, so false positives and overdiagnosis still need tracking.
Early detection changes outcomes. Five-year relative survival is close to 100% when breast cancer is found while still localized, and about 34% once it has spread to distant sites.
Dense breasts need more than a mammogram. In the SomoInsight study of 15,318 women with dense breasts, adding automated 3D ultrasound to mammography found 30 more cancers, with more recalls. Ultrasound supplements mammography, it does not replace it.
Women's health deserves better evidence. Women were historically underrepresented in clinical research, so new diagnostic tools, including AI, should be tested across ages and patient groups and be clear about their limits.
As an investor and as a patient, I see meaningful progress as reliable detection, clear explanations, appropriate follow-up and appointments that are easier to approach.
Read the full article, including comments from Ariel Ganz, PhD, co-founder of Arben Ventures, on Meditech Today.