The paper presents Libo Zhang’s algorithmic solution for the autoPETV Grand Challenge, focusing on interactive lesion segmentation in whole-body PET/CT scans. The method encodes user scribbles as two additional input channels and trains a large residual‑encoder U‑Net (≈140 M parameters) through a three‑phase curriculum over 4000 epochs, progressively moving from fully automatic segmentation to handling user‑provided scribbles and finally correcting its own mistakes via simulated error‑driven steps. Using 1811 studies for training and an ensemble of five‑fold checkpoints, the approach achieves a mean AUC‑Dice of 3.836 and a mean AUC‑DMM of 3.869 in interactive five‑fold cross‑validation, with significant gains from the first corrective scribble.
arXiv:2609.01554v1 Announce Type: cross Abstract: Automated lesion segmentation in whole-body PET/CT is complicated by the variety of physiological tracer uptake patterns and by the differing appeara...
arXiv:2608.30844v1 Announce Type: cross Abstract: Interactive lesion segmentation in whole-body PET/CT requires a model to provide a strong initial prediction while also responding efficiently to spa...
The paper introduces an anatomy-aware, promptable segmentation model for whole-body lesion detection in FDG and PSMA PET/CT scans, tailored for the AUTOPET V challenge. The approach builds on nnU-Net, employing a two-stage training process: an initial pre-training phase for strong baseline segmentation and an online interactive phase that refines predictions using scribble prompts. Anatomical context is integrated via organ supervision with a shared head predicting both lesions and organs, reducing false positives, while a tracer classifier directs studies to either a combined FDG+PSMA model or a PSMA-specific model. Cross-validation results show that organ-supervised training yields the most stable performance, the interactive stage consistently improves Dice scores, and PSMA-specific training delivers the best tracer-wise results.
DALE-CT introduces depth‑aware 2D slice encoders that learn an anatomical world model of chest CT scans without 3D or positional supervision. By sampling self‑supervised views across a physical $z$‑axis slab, the encoder captures how anatomy changes between neighboring slices, enabling it to recover slice ordering and distinguish slices by anatomy alone. The model, trained on a large 287k‑scan corpus, achieves state‑of‑the‑art performance on CT‑RATE and is released with full code and evaluation tools.
arXiv:2609.16775v1 Announce Type: new Abstract: Annotating large radiology datasets is bottlenecked by the manual effort of delineating structures slice-by-slice in 3D volumes. Interactive methods re...
arXiv:2609.09477v1 Announce Type: new Abstract: Delineating lung tumours on computed tomography (CT) takes a considerable share of the time spent on radiotherapy planning, and a contour proposed by a...
arXiv:2606. 28392v1 Announce Type: cross Abstract: Accurate lesion segmentation in PET/CT is critical for oncology, yet remains challenging because physiologic tracer uptake and artifacts can mimic malignant signal.
The paper introduces ExiL, a mask‑conditioned progressive learning framework for bone ultrasound segmentation that models annotation as a structured refinement trajectory. ExiL uses a synthetic expert‑like brush simulator and a lightweight U‑Net to learn from imperfect masks, and it can be updated in real time from expert refinements. In experiments on UltraBones100k and a prospective volunteer dataset, ExiL cut average annotation time from 60 to 20 seconds per frame and improved mean Dice by about 0.045, achieving 0.87 Dice and 2.7 px boundary error with 10–50 ms inference.
arXiv:2610.09940v1 Announce Type: cross Abstract: Joint-embedding predictive architectures (JEPAs) predict masked or future observations in representation space, offering a natural source of predicti...
arXiv:2607. 22739v1 Announce Type: cross Abstract: We study how far a deliberately simple behavioral-cloning policy can progress in a visually rich first-person game before adding reinforcement learning or explicit memory.
arXiv:2506. 06006v3 Announce Type: replace-cross Abstract: Can unified vision-language models (VLMs) perform forward dynamics prediction (FDP), i.
GroundingPI is a 4‑billion‑parameter grounding foundation model that generates points and boxes as quantized coordinates in a shared vocabulary. It is trained with multimodal and spatial pretraining, supervised fine‑tuning, and reinforcement learning, achieving a new state‑of‑the‑art average of 73.68% across 34 grounding benchmarks. As a visual backbone, GroundingPI improves performance in robotic manipulation and autonomous driving, outperforming larger models and mainstream backbones in several out‑of‑distribution settings.