Sharper reading of scanned and photographed reports in PDF

Not every report is a clean digital PDF. Many are scans or phone photos of a printed sheet, and those are harder to read accurately. We reworked how ReportSense handles these so more values come through correctly on the first try.

What changed

• Better handling of faint or overprinted text on scanned pages.
• Fewer misreads on reports photographed at an angle or in poor light.
• More reliable capture of parameter names that labs write in shorthand.

Getting a value right matters, so every reading is checked against a plausible range before it is saved.