9/ Let us know your thoughts about the preprint. We’re excited to discuss and collaborate further! 🧑‍🔬 #DisordersOfConsciousness

8/ Many thanks to my incredible collaborators across a few different countries. This work wouldn’t be possible without such an inspiring and multidisciplinary team. 🙏

7/ The ultimate goal? A step towards personalized medicine for DoC patients—providing clinicians with better tools to detect consciousness and predict recovery, enabling informed care decisions. #ERANET_PerMed

6/ Our approach goes beyond accuracy:
- We explore feature importance and univariate analyses to understand why certain signals are critical.
- Observed disagreements between modalities might reveal patient-specific paths to recovery. 🔍

5/ With data from two other international centers, we also validated how models trained in one hospital can generalize to others! 🌍 This highlights the potential for scaling these methods across healthcare systems. 🏥

4/ Key insights:
- EEG (task-based) & PET are strong for diagnosing a patient's current state.
- Structural scans like dMRI excel in predicting patient recovery.
- Some features predict better diagnosis and others prognosis.
- Combining modalities improves accuracy 📊

3/ Our work uses multimodal neuroimaging (EEG, fMRI, PET, dMRI, aMRI) + interpretable machine learning to evaluate diagnostic and prognostic predictions for DoC patients. Each modality provides a unique window into brain function and structure. 🧠

2/ Let's break down the key findings and why they matter.
Disorders of consciousness (DoC) like unresponsive wakefulness syndrome and minimally conscious state pose a major diagnostic challenge. Behavioral assessments can be ambiguous, with many factors influencing responses.

Hi Nikolas, could you add me as well? Thanks!