An echo, a DEXA, a coronary calcium score, trended over time

Three kinds of record, kept separate

Each scan you log is keyed to a procedure from a curated catalog and holds three kinds of information. Readings are what the scan measured (LVEF, carotid IMT, liver fat fraction): numbers, classified against a sex-specific reference band by the same engine bloodwork uses, and trended across your studies. Findings are what the radiologist concluded, coded where a code exists and in free text where it doesn't, since "partial supraspinatus tear, articular surface" has no number. Files are the report PDFs, exported stills and de-identified DICOM series, which you can page through in the app.

A number can be graphed and a radiologist's conclusion cannot, so keeping them apart keeps clinical judgments off the trend chart.

AI-extracted readings wait for you

A reading MuscleBuddy extracts from an uploaded report with AI arrives unconfirmed. Until you accept it, it does not trend, flag, feed a rank or count as verification evidence. AI does not write findings: a finding is a clinical conclusion, and generating one would amount to a diagnosis. Each finding on your record was entered by a person reading the report.

DICOM is de-identified before it is stored

Upload the DICOM files a facility gives you on a CD or through a portal, and they are de-identified to the PS3.15 Basic Confidentiality Profile before anything is written to storage. The pass keeps an allow-list of tags, so an unexpected tag cannot slip through, and it regenerates every UID. A file whose header declares burned-in annotation is refused: patient details printed into the pixels can't be removed without reading the image, so the file is not stored at all.

What a scan writes elsewhere

A whole-body DEXA writes its body composition onto your weigh-in for that day, so your bodyweight has one source instead of two. Each study is also stamped with the phase you were in when it was taken (a blast, PCT or recovery block), so you can read a change against what was running at the time.

Imaging does not drive training. When you or a coach start a rehab episode by hand, a finding can pre-select the matching rehab pack, and that is as far as it goes. No scan creates a rehab episode, advances a stage or writes a training constraint, because a scan steering a program would be medical advice.

What a coach and an export can see

Imaging follows the same privacy rules as the rest of Health: it is not posted to a feed, not shown on a leaderboard, and not included in any export except your own. A coach sees your studies only through a scope you grant, and loses access when the relationship ends. If you attach a study as evidence in a natural-verification check, only the measured part is published: procedure, modality, region, date and confirmed readings. Findings, notes, the facility, the ordering clinician, the report and the DICOM series stay in your account; a reviewer who needs more has to ask you.

There are two exports. Your own data export includes unconfirmed readings, labelled with their source and whether you confirmed them. A clinical FHIR export to a provider leaves out anything unconfirmed, because a FHIR Observation is a clinical assertion and another clinician might act on it.