analiza aiqbench
SleepVLM addresses the critical barrier to clinical adoption in automated sleep staging by introducing a rule-grounded vision-language model that generates auditable rationales aligned with AASM criteria. The model matches state-of-the-art accuracy on multiple datasets while achieving high expert scores for factual accuracy and logical coherence, demonstrating that interpretability can be integrated without sacrificing performance. By releasing the MASS-EX dataset alongside the model, the authors provide essential resources for advancing research in explainable medical AI.
SleepVLM: Explainable and Rule-Grounded Sleep Staging via a Vision-Language Model
arXiv:2603.26738v3 Announce Type: replace-cross Abstract: While automated sleep staging has achieved expert-level accuracy, its clinical adoption is hindered by a lack of auditable reasoning. We introduce SleepVLM, a rule-grounded vision-language model (VLM) that stages sleep from multi-channel polysomnography (PSG) waveform images and generates clinician-readable rationales based on American Academy of Sleep Medicine (AASM) scoring criteria. Utilizing waveform-perceptual pre-training and rule-grounded supervised fine-tuning, SleepVLM achieved Cohen's kappa of 0.767 on a held-out test set (MASS-SS1) and 0.743 on an external cohort (ZUAMHCS), matching state-of-the-art performance. Independent expert evaluation by two trained sleep technologists further validated the model's reasoning quality, with mean scores of 3.75-3.96 out of 5 across factual accuracy, evidence comprehensiveness, and logical coherence on both datasets. By coupling competitive performance with transparent, rule-based explanations, SleepVLM may improve the trustworthiness and auditability of automated sleep staging in clinical workflows. To facilitate further research in interpretable sleep medicine, we release MASS-EX, a novel expert-annotated dataset.
Abstract:While automated sleep staging has achieved expert-level accuracy, its clinical adoption is hindered by a lack of auditable reasoning. We introduce SleepVLM, a rule-grounded vision-language model (VLM) that stages sleep from multi-channel polysomnography (PSG) waveform images and generates clinician-readable rationales based on American Academy of Sleep Medicine (AASM) scoring criteria. Utilizing waveform-perceptual pre-training and rule-grounded supervised fine-tuning, SleepVLM achieved Cohen's kappa of 0.767 on a held-out test set (MASS-SS1) and 0.743 on an external cohort (ZUAMHCS), matching state-of-the-art performance. Independent expert evaluation by two trained sleep technologists further validated the model's reasoning quality, with mean scores of 3.75-3.96 out of 5 across factual accuracy, evidence comprehensiveness, and logical coherence on both datasets. By coupling competitive performance with transparent, rule-based explanations, SleepVLM may improve the trustworthiness and auditability of automated sleep staging in clinical workflows. To facilitate further research in interpretable sleep medicine, we release MASS-EX, a novel expert-annotated dataset.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2603.26738 [cs.CV]
(or arXiv:2603.26738v3 [cs.CV] for this version)
Text integral sub licența deschisă a publicației · 170 cuvinte
Subiecte conexe
SleepVLM: A Rule-Grounded Vision-Language Model for Auditable Sleep Staging
arXiv · 6 aug. 2026
Models Can't Remember Their Training. Neither Can You.
LessWrong · 19 iul. 2026
AI in healthcare: applications and best practices
Databricks · 24 iul. 2026
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
METR · 26 aug. 2026