When you upload a lab result, medical report or radiology description to Looms, the platform's AI reads the document and generates a plain-language summary highlighting key findings. Here's what the summary covers — and where you should use your own judgement.
What the AI summary includes
- Key findings — values or observations flagged as abnormal by the document itself
- Plain-language explanation of each finding
- Whether values are above or below the reference range
- Suggested context — what a finding might indicate, framed as possibilities, not diagnoses
- Any urgent-looking flags — values the AI identifies as requiring prompt attention
What the AI summary does not do
- It does not diagnose you — findings are contextual observations, not clinical diagnoses
- It does not replace your doctor's interpretation — clinical context, history and examination are essential for diagnosis
- It cannot read images — the AI reads text (including text in PDF documents) but not X-ray or MRI images
- It may miss findings that require specialist knowledge to interpret
- It cannot account for trends in your personal history unless you also ask the AI assistant
Improving AI accuracy
- Upload clear, high-quality scans — blurry or skewed documents reduce text recognition accuracy
- If the document is handwritten, the AI may not parse it well — type out the key values manually in the notes field
- Check that the record type is correctly set (e.g. Lab Result, not Medical Report) — the AI adjusts its analysis by type
If a summary seems wrong
If the AI has misread a value or labelled something incorrectly, tap the feedback option below the summary. You can also correct the finding manually in the record notes. The AI summary doesn't alter the original uploaded document — it's an overlay that you can disregard. For a broader look at what AI analysis surfaces across your health record, see our article on what an AI health summary shows, and for a deeper dive into how the analysis works, read how AI analysis works.