{"id":1215,"date":"2023-10-19T19:06:06","date_gmt":"2023-10-19T19:06:06","guid":{"rendered":"https:\/\/bcchrmicro.wpengine.com\/?page_id=1215"},"modified":"2026-04-09T20:33:51","modified_gmt":"2026-04-09T20:33:51","slug":"projects","status":"publish","type":"page","link":"https:\/\/www.bcchr.ca\/adsu\/projects\/","title":{"rendered":"ADSU Projects"},"content":{"rendered":"
\n\t
\n\t
\n\t
\n\t
\n\t<\/span>\n\tAI and Data Science Unit<\/span><\/a><\/span>News & Events<\/span><\/a><\/span><\/div>\n

ADSU Projects<\/h1><\/div>\n
\n\t
\n\t\t\t<\/figure>\n<\/div>\n<\/div><\/div>\n<\/div>\n\n\n
\n\t
\n\t
\n\t\n
\n

ADSU collaborates with internal and external partners from various sectors. We support clinical research collaborations at BC Children\u2019s Hospital Research Institute, while also advancing AI and data science with ADSU\u2019s very own research program. Our past and ongoing projects spans data pipeline building, to integrating AI models in clinical research settings. ADSU has worked on with BCCHR researchers\/clinicians to:
<\/p>\n<\/div>

\"\"<\/figure><\/div>\n\n
\n
    \n
  • Leverage optical character recognition and LLMs to extract relevant data from clinical notes<\/li>\n\n\n\n
  • Integrate data from clinical notes, medical imaging, and clinical registries to predict pediatric epilepsy surgery success<\/li>\n\n\n\n
  • Automate data cleaning pipelines to develop predictive models to forecast treatment trajectories of pediatric orthopedic patients<\/li>\n\n\n\n
  • Consolidate of data from multiple sources (REDCap, network drives, and PDF reports) to streamline data analysis and reporting<\/li>\n\n\n\n
  • Develop a regular program evaluation tool to improve workflow and reporting timelines<\/li>\n<\/ul>\n<\/div>
    \"\"<\/figure><\/div>\n\n
    \n
      \n
    • Integrate semantic and vector-based AI search, conversational interfaces, and user-centred design to enhance a document retrieval system on PHSA platforms<\/li>\n\n\n\n
    • Automate the analysis workflow of BCCHR\u2019s DNA Sequencing Core Facility\u2019s new whole plasmid sequencing service<\/em><\/li>\n<\/ul>\n<\/div>
      \"\"<\/figure><\/div>\n<\/div>\n<\/div>
      \n\t
      \n\t
      \n\t

      Current Projects<\/h2>

      Pediatric Epilepsy: AI-Assisted Surgical Decision Support<\/h3>

      Applying natural language processing and machine learning to pediatric epilepsy surgery, this program automates the extraction of clinically relevant features from MRI reports and develops predictive models to forecast surgical outcomes \u2014 helping clinicians identify the right candidates for surgery and anticipate results.<\/p>

      Research to date:<\/h4>

      Predictive Models for Postoperative Seizure Freedom in Pediatric Epilepsy: Merging MRI Saliency Features to Enhance Model Parsimony.<\/em><\/h4>

      Bahel VR, Phung TAT, Sujan J, G\u00f6rges M, van Rooij T, Kim D, Bjornson B, Go C, Connoly M, Tamber M. In: Proceedings of the 2025 AANS Annual Scientific Meeting, Boston, MA: American Association of Neurological Surgeons AANS; 2025. P104.<\/em><\/p>

      \n\t\t\t<\/figure>\n

      Key Points<\/h4>
        \n\tWhen using detailed MRI measurements in a small group of kids, the model \u201cmemorizes\u201d those patients and then poorly predicts results for new kids<\/li>\n\n\tIf we combine many smaller brain areas into a few larger brain areas, we end up with fewer inputs, and the model becomes simpler and tends to work better overall<\/li>\n\n\tAfter combining regions, the model is less \u201cover\u2011tuned\u201d to the original data (its performance on training and test sets is more similar), which means it\u2019s more likely to work well on future patients.<\/li>\n<\/ul>

        Feature extraction from MRI clinical reports in Pediatric Epilepsy: Rule-based, Open-weight LLMs and Hybrid Methods.<\/em><\/h4>

        King B, Bahel VR, Paul S, Abraham AP, Kim D, G\u00f6rges M, Bjornson BH, Sujan J, van Rooij T, Tamber MS. In: Proceedings of CLAE 2025 Annual Scientific Meeting. Calgary, AB: Canadian League Against Epilepsy; 2025.<\/em><\/p>

        \n\t\t\t<\/figure>\n

        Key Points<\/h4>