A note for people who read for a living

You've heard me complain about PACS for years.
I finally did something about it.

The short version — updated July 2026, after a week where most of it stopped being slideware. Forward it to anyone who reads for a living.

The problem, in one sentence

We spend our days inside software built by people who have never had a list of 40 unread studies at 6pm — and it shows: something like 45 minutes a day per radiologist lost to clicking, waiting, re-hanging, and fighting the PACS-to-dictation glue.

Sectra is genuinely good — if you're a 400-bed system with a seven-figure budget. If you're a teleradiology group or an imaging center, you're choosing between platforms that were modern in 2012 and cloud products built by people who have never watched someone read. That's the gap.

What Slate.MD is

A reading-first PACS + reporting system for teleradiology groups and imaging centers, built by a radiologist who still reads daily — tested nightly on my own multi-monitor rig, bugs fixed while I'm still sitting there. The whole product is organized around one question: what makes the next study open faster and read cleaner?

What it does

The read

  • No diagnostic monitor is ever dark; hanging protocols fill the glass.
  • Scroll never waits. The worklist pre-downloads into workstation memory — a real workstation's worth, tens of studies — and when you do outrun the cache, you get an instant preview frame with an honest "loading full fidelity" tag. The system refuses to let you sign until every frame you actually looked at has been shown to you at full quality. Speed never quietly costs diagnostic confidence.
  • STAT mode: sign, and the next study is already hung — the one after that is already downloading. The switch overhead between studies is measured in a corner of the screen, because that overhead is the product.
  • Tools feel right: scroll on wheel and drag, zoom anchored under your cursor, measurements that don't fight you — every binding user-remappable, multiple keys per action, because no two rads agree.

The dictation

  • On-site speech engine — your audio never leaves the building. Fast draft pass + accuracy pass on every utterance.
  • It handles the PowerScribe classics: "sigmoid colon" never becomes "sigmoid:", "radiograph" never becomes a paragraph break, and hypo/hyper confusions can never be silently auto-"corrected" — words that invert clinical meaning are structurally barred from automatic substitution and get flagged for your eyes instead.
  • A per-user voice profile — the thing everyone misses from Dragon, modernized: your vocabulary, your corrections, and (with your consent) acoustic adaptation that only ships an update if it measurably beats your current accuracy on your own dictations. You watch your error rate drop month over month.

The report

  • Dictate findings in whatever order you read; the system routes them to the right organ sections deterministically — and when it isn't sure, it asks instead of guessing.
  • The bright line, enforced in the architecture itself: the system can never assert a clinical finding you didn't make. Templates are built from your own prior reports in your own words; suggestions can prompt and style, never author. A complete verbatim record of exactly what you said exists at every instant.
  • Chart-aware guardrails: dictate "gallbladder is unremarkable" on a patient with a 2019 cholecystectomy in the chart and you get an amber "verify" before it ever reaches a referrer. Same for left/right vs the order, contrast vs the tech's notes, and "the clinical question was never addressed."
  • A sign-to-send hold your group configures — instant for STAT (the ED is standing there), a couple of minutes for routine — so the oops you catch as you hit Sign never reaches the chart.

What it deliberately does NOT do

  • No black-box AI reads. Anything that detects disease is a regulated device; we don't ship uncleared versions dressed up as "assistance." Everything that makes you faster — context, reformats, drafting from your own findings with fabricated numbers structurally blocked — is built in.
  • No cloud PHI. The appliance lives on-site; your voice, your reports, your patients' charts stay in the building.
  • No pretending. The diagnostic seat rides on an FDA-cleared viewer partner while our own clearance path proceeds. Unglamorous, correct order.

Where it stands, honestly

Real pixels on real monitors from a real archive; the reading loop — worklist → hang → scroll → measure → dictate-adjacent → sign — runs end-to-end today on my rig; an eight-hour "zero degradation across a shift" endurance graph runs nightly.

Design-partner pilots this fall — teleradiology groups and imaging centers, 5–25 rads, paid but design-partner priced, weekly feedback call, my cell number. Then standard contracts into 2027.

What I'd love from you

  • If you're in a terad group or imaging center on an aging platform: give me 30 minutes to read a real case in front of you — an actual read, open to sign, not a feature tour.
  • If you're not: tell me the three things about your current PACS that make you want to throw a monitor. That list is literally the roadmap — half of what's above started as exactly that kind of complaint.
  • Know a group owner who should see this? Send it along.

— Eli · elisha.lim@gmail.com