222 lines
6.9 KiB
Markdown
222 lines
6.9 KiB
Markdown
# inou MCP Connector - Usage Examples
|
|
|
|
These examples demonstrate how Claude can help users explore their health data through the inou connector.
|
|
|
|
---
|
|
|
|
## Example 1: Brain MRI Analysis
|
|
|
|
**User prompt:**
|
|
> "I had a brain MRI last year. Can you look at it and tell me if anything looks concerning? I've been having headaches."
|
|
|
|
**Claude's approach:**
|
|
|
|
1. Lists the user's dossiers to find their health record
|
|
2. Lists imaging studies, identifies the brain MRI
|
|
3. Fetches the contact sheet to see all slices at a glance
|
|
4. Identifies relevant axial slices through the ventricles and key structures
|
|
5. Fetches full-resolution images of specific slices for detailed analysis
|
|
6. Provides observations about brain structures, symmetry, and any visible findings
|
|
|
|
**Sample interaction:**
|
|
|
|
```
|
|
Claude: I found your brain MRI from May 2024. Let me take a look at the axial
|
|
T2-weighted images, which are best for seeing brain tissue detail.
|
|
|
|
[Fetches contact sheet, then specific slices]
|
|
|
|
Looking at these images, I can see:
|
|
- The lateral ventricles appear symmetric and normal in size
|
|
- No midline shift is present
|
|
- The brain parenchyma shows normal signal intensity
|
|
- I don't see any obvious mass lesions or areas of abnormal enhancement
|
|
|
|
However, I want to be clear: I'm an AI assistant, not a radiologist. For
|
|
headaches, the MRI findings are just one piece of the puzzle. Your doctor
|
|
can correlate these images with your symptoms, medical history, and physical
|
|
exam for a complete picture.
|
|
|
|
Would you like me to look at any specific region more closely?
|
|
```
|
|
|
|
---
|
|
|
|
## Example 2: Lab Result Trends
|
|
|
|
**User prompt:**
|
|
> "Can you check my vitamin D and iron levels? I've been feeling tired lately."
|
|
|
|
**Claude's approach:**
|
|
|
|
1. Lists available lab tests for the user
|
|
2. Fetches vitamin D and iron panel results (ferritin, serum iron, TIBC)
|
|
3. Identifies trends over time if multiple results exist
|
|
4. Explains the values in context with reference ranges
|
|
|
|
**Sample interaction:**
|
|
|
|
```
|
|
Claude: I found your lab results. Here's what I see:
|
|
|
|
**Vitamin D (25-OH):**
|
|
- Jan 2024: 18 ng/mL (low - reference: 30-100)
|
|
- Jul 2024: 32 ng/mL (normal)
|
|
- Dec 2024: 24 ng/mL (insufficient)
|
|
|
|
Your vitamin D dropped again after summer, which is common in northern
|
|
latitudes. The insufficient level could contribute to fatigue.
|
|
|
|
**Ferritin:**
|
|
- Dec 2024: 45 ng/mL (normal range: 12-150)
|
|
|
|
**Serum Iron:**
|
|
- Dec 2024: 85 mcg/dL (normal range: 60-170)
|
|
|
|
Your iron stores look healthy. The fatigue is more likely related to your
|
|
vitamin D insufficiency. Many people benefit from supplementing with D3
|
|
during winter months - typically 1000-2000 IU daily, though your doctor
|
|
can recommend the right dose for you.
|
|
|
|
Want me to look at any other lab values?
|
|
```
|
|
|
|
---
|
|
|
|
## Example 3: Genetic Insights for Medication Response
|
|
|
|
**User prompt:**
|
|
> "I'm about to start an antidepressant. Can you check if I have any genetic factors that might affect how I respond to medications?"
|
|
|
|
**Claude's approach:**
|
|
|
|
1. Queries the genome data for pharmacogenomics-related variants
|
|
2. Looks specifically at CYP2D6, CYP2C19 (drug metabolism enzymes)
|
|
3. Checks for MTHFR variants that affect folate metabolism
|
|
4. Explains findings in practical terms
|
|
|
|
**Sample interaction:**
|
|
|
|
```
|
|
Claude: I found some relevant genetic variants in your data:
|
|
|
|
**CYP2D6 - Drug Metabolism:**
|
|
You have rs3892097 (CYP2D6*4) - one copy. This makes you an "intermediate
|
|
metabolizer" for many medications processed by CYP2D6, including some
|
|
antidepressants like:
|
|
- Fluoxetine (Prozac)
|
|
- Paroxetine (Paxil)
|
|
- Venlafaxine (Effexor)
|
|
|
|
This means you may process these medications more slowly than average. Your
|
|
doctor might start with a lower dose.
|
|
|
|
**CYP2C19:**
|
|
You have rs4244285 (CYP2C19*2) - one copy. This affects metabolism of:
|
|
- Citalopram (Celexa)
|
|
- Escitalopram (Lexapro)
|
|
- Sertraline (Zoloft)
|
|
|
|
Again, intermediate metabolism - not a problem, but good to know.
|
|
|
|
**MTHFR:**
|
|
You have rs1801133 (C677T) - heterozygous. This is very common and mildly
|
|
reduces folate processing. Some research suggests methylfolate supplementation
|
|
may support antidepressant response.
|
|
|
|
I'd recommend sharing these results with your prescribing doctor. This
|
|
information can help them choose the right medication and starting dose
|
|
for you.
|
|
```
|
|
|
|
---
|
|
|
|
## Example 4: Cross-Domain Health Analysis
|
|
|
|
**User prompt:**
|
|
> "I have MTHFR mutations. Can you look at my homocysteine levels and see if they correlate?"
|
|
|
|
**Claude's approach:**
|
|
|
|
1. Queries genome for MTHFR variants
|
|
2. Fetches homocysteine lab results
|
|
3. Correlates the genetic and lab findings
|
|
4. Provides context on the relationship
|
|
|
|
**Sample interaction:**
|
|
|
|
```
|
|
Claude: Let me pull together your genetic and lab data on this.
|
|
|
|
**MTHFR Genetic Status:**
|
|
- rs1801133 (C677T): T/T (homozygous) - reduces enzyme activity by ~70%
|
|
- rs1801131 (A1298C): A/C (heterozygous) - mild additional effect
|
|
|
|
This combination significantly reduces your MTHFR enzyme function.
|
|
|
|
**Homocysteine Lab Results:**
|
|
- Mar 2023: 14.2 umol/L (slightly elevated - reference: <10)
|
|
- Sep 2023: 11.8 umol/L (borderline)
|
|
- Mar 2024: 9.1 umol/L (normal)
|
|
|
|
I can see your levels have improved over time. This pattern is consistent
|
|
with what happens when people with MTHFR variants supplement with
|
|
methylfolate (active B9) and methylcobalamin (active B12), which bypass
|
|
the impaired enzyme.
|
|
|
|
The correlation here is clear: your genetic profile predicts higher
|
|
homocysteine risk, and your labs confirmed it until intervention brought
|
|
levels to normal.
|
|
|
|
Would you like me to check other B-vitamin levels that might be relevant?
|
|
```
|
|
|
|
---
|
|
|
|
## Example 5: Spine MRI Navigation
|
|
|
|
**User prompt:**
|
|
> "Can you look at my lumbar spine MRI? My doctor mentioned something about L4-L5."
|
|
|
|
**Claude's approach:**
|
|
|
|
1. Lists imaging studies, finds the spine MRI
|
|
2. Lists series to find sagittal and axial T2 sequences
|
|
3. Uses contact sheet to identify L4-L5 level
|
|
4. Fetches relevant slices at full resolution
|
|
5. Describes findings at the specific level
|
|
|
|
**Sample interaction:**
|
|
|
|
```
|
|
Claude: I found your lumbar spine MRI from October 2024. Let me look at
|
|
the L4-L5 level specifically.
|
|
|
|
[Fetches sagittal T2 for overview, then axial slices at L4-L5]
|
|
|
|
At the L4-L5 level, I can see:
|
|
- There's a disc bulge that extends posteriorly (toward the back)
|
|
- The bulge appears to contact the thecal sac (the fluid-filled sac
|
|
containing the nerve roots)
|
|
- On the axial view, the bulge is slightly more prominent on the left side
|
|
- The neural foramina (nerve exit tunnels) appear mildly narrowed on the left
|
|
|
|
This pattern is consistent with a disc bulge that could potentially affect
|
|
the L5 nerve root, which can cause symptoms down the leg.
|
|
|
|
For comparison, your L3-L4 level above looks normal with good disc height
|
|
and no bulging.
|
|
|
|
Want me to look at any other levels, or compare the left and right sides
|
|
in more detail?
|
|
```
|
|
|
|
---
|
|
|
|
## Notes for Users
|
|
|
|
- Claude provides informational analysis, not medical diagnoses
|
|
- Always discuss findings with your healthcare provider
|
|
- Image quality and completeness affect analysis accuracy
|
|
- Genetic interpretations are based on current research, which evolves
|