Camyla: Scaling Autonomous Research in Medical Image Segmentation

2026-04-12

AI Summary

This work addresses key challenges in autonomous medical image segmentation research—namely exploration drift, knowledge degradation, and inefficient failure recovery—by proposing an end-to-end autonomous research system. The system establishes a fully automated research pipeline from data to publication through three core mechanisms: quality-weighted branch exploration, hierarchical reflective memory, and divergent diagnostic feedback. Integrated with autonomous experiment scheduling, multi-granularity knowledge compression, and failure-driven re-exploration strategies, the framework operates without human intervention on an 8-GPU cluster. It generates over 2,700 novel models and 40 manuscripts across 31 datasets, achieving state-of-the-art performance on 24 datasets—surpassing strong baselines such as nnU-Net—and producing results commensurate with T1/T2 journal standards.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Learning to Search

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

We present Camyla, a system for fully autonomous research within the scientific domain of medical image segmentation. Camyla transforms raw datasets into literature-grounded research proposals, executable experiments, and complete manuscripts without human intervention. Autonomous experimentation over long horizons poses three interrelated challenges: search effort drifts toward unpromising directions, knowledge from earlier trials degrades as context accumulates, and recovery from failures collapses into repetitive incremental fixes. To address these challenges, the system combines three coupled mechanisms: Quality-Weighted Branch Exploration for allocating effort across competing proposals, Layered Reflective Memory for retaining and compressing cross-trial knowledge at multiple granularities, and Divergent Diagnostic Feedback for diversifying recovery after underperforming trials. The system is evaluated on CamylaBench, a contamination-free benchmark of 31 datasets constructed exclusively from 2025 publications, under a strict zero-intervention protocol across two independent runs within a total of 28 days on an 8-GPU cluster. Across the two runs, Camyla generates more than 2,700 novel model implementations and 40 complete manuscripts, and surpasses the strongest per-dataset baseline selected from 14 established architectures, including nnU-Net, on 22 and 18 of 31 datasets under identical training budgets, respectively (union: 24/31). Senior human reviewers score the generated manuscripts at the T1/T2 boundary of contemporary medical imaging journals. Relative to automated baselines, Camyla outperforms AutoML and NAS systems on aggregate segmentation performance and exceeds six open-ended research agents on both task completion and baseline-surpassing frequency. These results suggest that domain-scale autonomous research is achievable in medical image segmentation.

Problem

• autonomous research
• medical image segmentation
• long-horizon experimentation
• knowledge degradation
• search drift

Innovation

• Autonomous Research
• Medical Image Segmentation
• Quality-Weighted Branch Exploration
• Layered Reflective Memory
• Divergent Diagnostic Feedback