RWTH Aachen University

E. Madan led experiments, data analysis, and manuscript writing. A.M.P., V.V., A.K., and P.B. contributed to experiments, data analysis, and manuscript preparation. J.W. handled biostatistical analysis and data interpretation. T.B. and D.A. focused on microscopy, data interpretation, and figure preparation. P.C.M., G.B., M.D.B., I.D., H.K., and D.O. assisted with biostatistical analysis and data interpretation. A. Gogna and A.B. helped with figure preparation and live-cell videos. M.L.P., M.Y., D.Z., S.P., S.K., K.B., P.M., I.C., E.A.F., A.K.P., S.M., C.J.P., A.L., M.N., S.C., K.G., K.C.G., and K.S.S. contributed to pathology data collection and analysis. T.F.A., E.G., A. Gustafsson, G.L., L.N., E.Y., and R.S.-F. worked on 3D scaffold and organoid experiments. K.D.P.D. assisted with PDX experiments. N.M., M.J.P., K.J.-J., and S.H. focused on exosome biology experiments. Z.S. and W.L.R. worked on multiplexing experiments. M.B. and P.J. handled FISH experiments. H.O.R., F.A.T., and R.B. contributed to MR imaging data analysis. K.G.A., J.A., G.X., G.S., and S. Singh designed and produced mAbs. S.D., C.L.d.S., A.M.F., D.T., S.R.G., K.P., G.G., G.L.W.G.R., and H.W. edited the manuscript. A.M.H., A.R., and C.L. worked on mass spec analysis. A.S. and P.B.F. provided laboratory resources and collaboration. G.M.C., Y.L., A.L.O., L.L., and A.A.T. handled CRISPR-off experiments. A.S.P., U.M., R.D., A.K.G., K.S., and D. Chelmow contributed to ovarian cancer expertise. R.P., J.K., S.R., H.R., S. Sayeed, K.P.K., D. Connolly, K.P.N., M.O.I., M.P.K., C.M., M.A., A.L.-B., D.J., K.D.M., A. Tzankov, S.B., E.A.R., and R.C.M. assisted with pathology data collection. A. Tsung and D.S. fostered clinical collaborations. J.T. provided mouse tumor models and funding. A.P.S., R.W., K.J.W., and E. Moreno helped with manuscript writing and funding. R.G. designed the study, organized collaborations, and provided overall supervision and funding.

New guidelines for AI use in brain cancer diagnosis and treatment ensure reliable clinical trial results and patient protection, addressing variability in radiologist interpretations and promoting objective AI analysis of tumor images.