Papers for

medical imaging research teams

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Medical image model evaluation improves with three separate scores

M3-Score: Fidelity, Memorization and Coverage as Separate Axes for Evaluating Generative Radiology Image Models

Abstract: Quantitative evaluation of generative models for radiology remains challenging. Clinically relevant structures are often small and infrequent, feature spaces learned from natural images may represent them poorly, and a single summary score cannot distinguish limited fidelity from limited diversity. This study proposes the Medical Multi-axis Maximum Mean Discrepancy score (M3-Score), an evaluation framework based on RadioDINO-s16, a frozen vision transformer pretrained on radiology images. M3-Score reports three complementary axes computed at pre-specified encoder depths: \emph{fidelity}, measured by an unbiased multi-bandwidth radial basis function (RBF) MMD$^2$ at the final block; \emph{memorization}, measured by nearest-neighbor distances at 75\% depth; and \emph{coverage}, defined as the fraction of real images with a generated neighbor within their $k$-nearest-neighbor radius at 33\% depth. Reference sets are sampled across subjects to limit the influence of correlated slices. On BraTS brain MRI, the fidelity axis ordered five comparison sets of increasing severity (Spearman $ρ= 1.00$), and a subject-disjoint real set yielded $\mathrm{MMD}^2 = 0$ (permutation $p = 1$). An unconditional denoising diffusion probabilistic model achieved $\mathrm{MMD}^2 = 0.073$ (95\% confidence interval $[0.071, 0.080]$) but covered only 38\% of the real distribution. Under progressive mode dropping, $k$-NN manifold recall increased at all twelve encoder blocks, whereas the proposed coverage estimator decreased monotonically ($ρ= -1.00$). RadioDINO-s16 features separated real brain MRI from generated samples with a ROC-AUC of 0.819, compared with 0.555 for InceptionV3 and 0.582 for CLIP. Across a twentyfold range of sample sizes, the mean M3 value varied by a factor of 1.05, compared with 2.52 for the Fréchet Inception Distance.

Sun 27 SeptComputer Vision and Pattern Recognition
The gist
Evaluating computer programs that create medical images is hard because important details are small and rare. The authors introduced a new way to measure these programs by looking separately at how realistic the images are, whether the program just remembers training images, and how well it covers all types of images it should. They tested their method on brain MRI data and showed it can better tell which models work well and which don’t, compared to older methods. This helps ensure medical image generators produce useful and varied images.
Open → 2609.33769v1