Max Research/Masters Updates

1 Agenda

• Masters Updates
• Research Updates
  • Imaging
    • MMII
    • MRF-CLIN
    • MRF-SEEG
  • Non-Imaging
    • HEMI-SLIM
    • FB-GEN
• Research Ideas
  • EPBiome v2?
  • MRF-Histo

2 Masters

• Nearing the end of coursework, now shifting focus to thesis (“capstone”)
• Required courses:
  • Intro to Data Structures and Algorithms in Python
  • Computing in Biomedical and Health Informatics
  • Statistics I
• Domain-specific Courses
  • Epidemiology
  • Machine Learning & Data Mining (in progress)
• Electives
  • Statistics II
  • Categorical Data Analysis (summer)
  • May need 1-2 more, have to check with program director
• Thesis (3-6 credits)
  • Committee formed: Satya Sahoo, Andreas Alexopoulos, Irene Wang
  • Focus will be MRF-SEEG project

3 Research Updates

3.1 Imaging Projects

3.1.1 MMII

• Educational report from ILAE neuroimaging task force on the importance of multimodal imaging integration
• Co-written with Italian group (University of Modena)
• Manuscript submitted, first round of revisions just completed. Should be out soon I hope!

3.1.2 MRF-CLIN

• Project to summarize our experience with MRF, clinical yield, etc.
• Goal is N ~= 100, we’re now at N=90
• Here are the main takeaways:

3.1.2.1 Additional findings in over 1/3 of patients

3.1.2.2 Additional Findings event if negative on intial clinical MRI

3.1.2.3 Increased lesion conspicuity and extent most common additional finding

3.1.2.4 Next steps

• 1-2 further review sessions with Dr Jones to increase our total N to ~100
• Manuscript in prep

3.1.3 MRF-SEEG

3.1.3.1 Who are these patients

3.1.3.2 N=2639 SEEG contacts, 1858 without WM/CSF/OUT

3.1.3.3 At group level, EZ contacts have higher T1 in Frontal

3.1.3.4 Including pathology

It appears that for patients with PVNH in particular EZ has higher average T1 GM value, this is not normalized to side/lobe/sublobar region, so take with a grain of salt… Irritative zone may show higher T1 value as well across FCD, PVNH, and nonlesional.

3.1.3.5 May be important to look at individual-level…

3.2 Non-imaging Projects

3.2.1 HEMI-SLIM

• Project started with Dr Moosa during fellowship.
• Basic question - large lesion, PMC opinions include hemispherectomy vs smaller resection. Which patients can “get away with” a smaller resection and which actually need the larger surgery? How can we predict this using presurgical data?

3.2.1.1 Results so far

3.2.1.1.1 PET vs Visible Lesion

- If PET extends beyond visible lesion, this is a poor prognostic outcome for smaller surgery

3.2.1.2 Current issues…

• Short follow up time for many patients (<1 year) - working on leveraging epilepsy surgery outcome register

• Went over these results with Poturalski who suggested that current approach of comparing PET data using text of report is not robust/reproducible/objective. Would be best if one person reviewed everything

• Even better would be a quantitative approach for defining volume of MRI lesion vs PET hypometabolism vs resection

  • Exploring SWANe software from Italian group currently
  • Victoria Morgan lab (Vanderbilt) has a different approach, may be able to use this but difficulty is anatomy

3.2.2 FB-GEN

• Project started with Dr Pestana towards the end of fellowship, have since recruited an MD/PhD student who I am supervising
• Online survey on caregiver social media use in families affected by genetic epilepsies.
• So far, N=202

• Main questions - how often they use social media? what kind of information is being exchanged? Healthcare utilization - ED visits etc.
• Hypothesis - those that are “more connected” have better support and may show reduced use of the emergency visits

4 Research Ideas

4.1 EPBiome

• During residency, I obtained seed funding from Weston Foundation $150,000 for microbiome study, called “EPBiome”
• Epilepsy is associated with microbiome changes. Many ancedotal and small studies support this, as well as animal studies. E.g. protective effect of KD on epilepsy is abolished in germ-free mice or if microbiome is cleared out, this is rescued when specific strains are re-introduced.
• Considering this, can we supplement with dietary fibre to “push” the profile to a healthier state, and thereby reduce seizures?

4.1.1 Trial design

• This was overly ambitious, especially given that COVID-19 was right around the corner. Difficulties with recruitment.

4.1.2 Preliminary Results

4.1.3 Possible spin-off? LIFE study adjacent?

• Larger pool of patients at CCF, would not limit to pediatrics
• Could look at:

1. Microbiome of epilepsy patients in general
2. Larger trial of inulin as a therapeutic intervention
3. KD - I think this is too difficult at one center, would need to have at least 5-6 centers

• Apply for CTSC grant or perhaps Catalyst grant

4.2 MRF-Histo

• As I’m looking at the MRF data, we are finding good clinical utility and potentially some promising data on correlation with epileptogenicity. However, what strikes me as a “missing piece” is the histology. Could we leverage our surgical volumes to combine MRF scans pre-op and co-register to pathology?
• Coming back again to the idea of a “gradient” - we should be looking at this at the macro and micro scales, to corroborate what we’re seeing under the microscope to what we can measure with MRF. Could there be a better delineation of lesion boundary? This is especially interesting in the pediatric epilepsy population where we have large diffuse lesions with difficult to define boundaries.