Article

Imaging and diagnostics centers: get paid for your records

Updated

This page is for the owner or administrator of an independent imaging or diagnostics center. You run expensive machines on a tight schedule, and every scan sits inside rules about patient privacy. So the natural assumption is that nothing you hold could ever be licensed. For the patient side of the business, start from that assumption. But a center also runs a business around the scans, and that business has kept its own record for years. Buyers license operating records like that to train AI models.

What a center usually overlooks is the equipment and workflow trail. The service ticket on the scanner that went down at 7:40 on a Monday. The work order the field engineer closed. The daily quality check, the protocol that was changed after a run of repeat scans, the scheduling template that was rebuilt twice. Our page on what data buyers want from a trade business lays out the buyer’s test with trades as the example. In an imaging center the test runs on the operations, never on the patient chart.

Statements about Polyshares here are quoted from its published pages. Statements about a center’s files are our own description. What health privacy law allows is your counsel’s call, not ours.

The records an imaging and diagnostics center keeps

Our examples, not a buyer’s list. They are the records a center produces as it works, day by day, outside the patient file.

The equipment service tickets. The scanner throws an error. A technologist logs it: what happened, which sequence, what the console said. The service desk opens a ticket. A remote engineer tries a reset. A field engineer comes out, replaces a part, runs a test, closes the ticket. Each ticket is a diagnosis from first symptom to fix, with times on every step. Over ten years and several machines, the tickets show which faults repeat, which fixes held and how long the room was dark.

The work orders and the preventive maintenance log. Scheduled service visits, the checklist for each, the parts replaced, the readings taken, the item deferred to the next visit and the reason. Our page on job files and work orders explains why a record of what was done, against what was planned, is worth more than a list of completed jobs. In a center, the planned visit next to the unplanned call-out tells that story.

The quality control logs. The daily and weekly checks a technologist runs before the first patient. The reading, the tolerance, the pass or the fail, and what happened after a fail. A failed check that led to a service call is a problem caught before it reached a scan. That chain of check, finding and action is reasoning written down as it happened.

The protocol library and its revisions. How a given exam is run on a given machine. The version that changed after the radiologists asked for a different sequence, the version that cut ten minutes from a slot, the version that was rolled back. The protocol text itself is the center’s method. The revision note says why it changed.

The scheduling template. How many slots of each exam type per room per day, the buffer after a contrast study, the block held for urgent referrals. The template that was rebuilt when a second machine arrived. Read by year, it is a record of how a center fitted demand to scarce equipment time.

The referral-office CRM. Which offices send which exam types, how a new office was brought on, the outreach visits, the complaints about turnaround and how each was handled, the office that stopped referring and why. Our page on CRM records explains which parts of a CRM carry reasoning and which parts are only contact details. In a center, the turnaround complaints and their resolution are the reasoning.

The incident and improvement file. A near miss, the review that followed, the change made to the checklist, the follow-up check a month later. The accreditation preparation folder, with the gaps found and closed. Each is a small case of a problem raised and resolved.

How a fault was diagnosed, how a protocol was revised, how a schedule was rebuilt: that is the how. Which patient was on the table never is.

See your estimate

What counts, in the buyer’s own words

Polyshares has no page about imaging centers. Its pages say this, and here is how it applies.

On sectors, Polyshares’ home page names “healthcare administration” among the businesses with “real operating history” that qualify, “in any sector.” The operating side of an imaging center, its equipment, its schedules and its referral relationships, is administration in that sense. Whether a particular record fits is Polyshares’ decision.

On what it pays for, its data licensing page: “A support ticket that shows a problem being misdiagnosed, argued over, and finally resolved is worth more than a clean table of finished numbers.” An equipment service ticket is exactly that kind of record, with a machine in place of a customer. The same page: “The messy working record is the point, not a defect in it.” The ticket with three wrong guesses before the right part is the valuable one.

On the systems it names: “Slack and Teams, Drive and SharePoint, email, CRM history, support tickets, design files, call recordings, accounting and ERP exports.” Service tickets and CRM history are both on that list.

Its Zendesk guide describes what a ticket holds: “A Zendesk ticket holds a customer’s problem in their own words, every agent reply, private notes between agents and the final status.” And “That full path from question to resolution shows how real support problems get solved.” A center’s internal service desk, in whatever tool it runs, follows the same shape.

Its Salesforce guide says “Cases carry the full path from customer report to resolution,” and “Custom objects often hold the most specific process data a company has.” A referral CRM with custom objects for turnaround issues is that kind of process data.

What is never included

Patient information is out. That means the images, the reports, the orders, the prior authorisations, the patient schedule, the billing and anything else tied to a person. Our page on what is never included covers the general lines, and our page on health, payment and other regulated data covers records that carry their own law.

The federal rule on de-identifying health information is at 45 CFR 164.514, published on the eCFR. It says health information “that does not identify an individual and with respect to which there is no reasonable basis to believe that the information can be used to identify an individual is not individually identifiable health information.” It then sets out two routes. One is a determination by “A person with appropriate knowledge of and experience with generally accepted statistical and scientific principles and methods.” The other is removal of a list of identifiers that includes “Names,” “Medical record numbers,” “Health plan beneficiary numbers,” “Device identifiers and serial numbers,” and “Full face photographic images and any comparable images.” We quote it so you can see how specific the rule is. We don’t interpret it. Ask counsel, with health privacy experience, before you count anything near a patient.

Why care, if the plan is to keep patients out? Because patient details leak into operational records. A service ticket that says which exam was running when the fault hit may name the patient. A scheduling note may carry a medical record number. Polyshares’ answer page on value says one thing that caps a record is “Personal information woven through the record in a way that cannot be cleanly stripped.” Its legal page says “No buyer with regulatory exposure will accept a record carrying personal information.”

Polyshares’ data handling page lists what its PII partner removes: “Names, emails, phone numbers, account numbers, and addresses.” Referring physicians and office staff are people too, so their names come out with the rest.

Then the tools. Polyshares’ Box guide says “You choose which users and folders are exported, so HR, legal and executive folders can stay out of scope.” Its Zendesk guide says sensitive queues “can stay out entirely.” Its Salesforce guide: “Data Export lets you choose which data types to include, so objects that hold sensitive fields can be left out before the files are built.” A center can scope the equipment desk and the referral CRM and leave every patient-facing system out.

Anything a vendor or a hospital partner owns is a third line. A service contract may say who owns the machine’s logs. A reading agreement may say who owns the reports. Our page on client-owned work and NDAs explains how contracts decide it.

How de-identification works

Here is the order Polyshares publishes.

First, nothing moves. Polyshares’ data handling page: “Intake asks for record counts, not files or exports.” A center counts service tickets, work orders and CRM records from the reports its systems already run.

Second, the PII partner. Polyshares says the record is anonymised by “A specialist PII partner” before Polyshares or any buyer sees it, and that “Roles, timestamps, and the work itself are kept.” A service ticket keeps the technologist’s role, the fault, the times and the fix.

Third, a sample is read. “Someone on the team reads the anonymized sample before Polyshares prices the record.” If anything in it reads as identifying, it goes back for another pass.

Fourth, the listing. Polyshares says listings carry no company name, website, people or exact locations, and that buyers “see de-identified listings only.”

A caution of our own. Polyshares’ PII process removes identifying fields. Health privacy law sets its own standard for health information, quoted above. The two are not the same question, and counsel answers the second. Our page on how de-identification works explains the general process.

What decides the price

Polyshares prices after it has read the record.

The range: on its answer page about value, Polyshares says deals “typically land between $100K and $2M paid to the company” (indicative, not an offer). Its home page adds that deals “run higher for large or long records.”

The published deals: four, each a perpetual exclusive licence anonymised before delivery, in home services, construction, energy and professional services. No imaging or health company is among them, and we would not read any of them as a guide to a center.

What raises it, in Polyshares’ words: “Length of continuous operating history,” “Visible disagreement and correction,” and “A specialized domain, because material from a trade with few written records is scarce in a way general business correspondence is not.” Equipment service in a diagnostic center is a specialised domain. Ten years of tickets on the same machines is length.

What lowers it: “A short operating history,” records “the company holds but does not own,” and personal information that cannot be cleanly stripped. For a center, the third is the one to watch. A ticket system where patient names sit in every second note is a harder record than one where they never appear.

Polyshares says it does “not pay for something we have not seen.” Our page on what decides the price goes through the general factors. How many years of service tickets does your equipment desk still hold?

Which programme fits

On this site, Polyshares’ programme is for US companies with fewer than 30 people. A single-site center with a small team is often inside that line; count before you assume. Our page on the programme for 29 or fewer people explains it.

The estimate asks how many people work at the center, counting everyone: technologists, front desk, billing, management, and any radiologists on staff. It names the programme that fits. A larger group, or a center outside the US, is pointed to the programme that fits it.

Polyshares says “The length and texture of the record matter more than headcount or revenue.” Headcount decides where you apply. The record decides the price.

Design and web agencies keep the same kind of ticket-and-revision trail, on client sites instead of scanners; see design and web agencies. Construction companies hold work orders and service records with the same gap between plan and outcome; see construction companies. Law firms face the same task of separating a regulated client file from the firm’s own operations; see law firms.

TrustPast

Questions and answers

Do the scans or the reports count?

Treat them as out and ask counsel. Images, reports and orders are patient health information, and federal rules at 45 CFR 164.514 set what de-identified health information means. This page is about the center’s own operating record.

What is the center’s own record, then?

This is our own description: equipment service tickets and work orders, QC logs, protocol revisions, scheduling templates and the referral-office CRM. Polyshares says the value is in reasoning that was captured while the work happened.

Does Polyshares see patient or referrer names?

Polyshares says names, emails, phone numbers, account numbers and addresses are removed or replaced by a specialist PII partner before Polyshares or any buyer sees the record.

Can we keep whole folders out of scope?

Yes. Polyshares’ Box guide says you choose which users and folders are exported, and its Salesforce guide says objects that hold sensitive fields can be left out before the files are built.

Do we send anything to apply?

No. Polyshares says intake asks for record counts, not files or exports, and nothing leaves your systems until it has made an offer and you have accepted it.

What does Polyshares pay?

Polyshares says deals typically land between $100K and $2M (indicative, not an offer). It reads the record before it prices it, and no imaging center is among its published deals.

See your estimate

*Purchases are made by Polyshares under its data licensing programme. TrustPast is an independent referral partner of Polyshares.

Sources

  • Polyshares, home page: deal range, healthcare administration among the sectors, the licence terms (polyshares.com)
  • Polyshares, data licensing: what it pays for, the operational data it names, the messy working record (polyshares.com)
  • Polyshares, how we handle your data: counts at intake, the PII partner, the sample review, who sees an unmasked record (polyshares.com)
  • Polyshares, case studies: the four published deals, anonymised before delivery (polyshares.com)
  • Polyshares, how much is my company’s data worth?: what moves the number, what caps it, personal information that cannot be cleanly stripped (polyshares.com)
  • Polyshares, is selling company data legal?: privacy regimes and personal information, anonymisation before anything moves (polyshares.com)
  • Polyshares, Zendesk data guide: the path from question to resolution, sensitive queues kept out (polyshares.com)
  • Polyshares, Box data guide: controlled documents and regulated records, folders kept out of scope (polyshares.com)
  • Polyshares, Salesforce data guide: cases from report to resolution, sensitive objects left out of Data Export (polyshares.com)
  • eCFR, 45 CFR 164.514: the de-identification standard for protected health information and the listed identifiers