A hundred lung nodules a day. Which three need a phone call?
Lung nodule queues are ordered by scan date, nodule size and guideline interval. Those are the signals available, and they do not separate the patients who need attention now from the ones who can safely wait. RevealAI-Lung returns an mSI™ with each finding, at the point of nodule, so the queue can be prioritized with more information on malignancy risk.
Point of nodule.
The high end of the mSI™ scale
Above 0.9, the lung nodule carries features closely resembling confirmed malignant lung nodules. This is the finding that moves a patient up the queue. The device returns a continuous 0–1 value; the high band is how RevealDx validation studies have grouped it.
Risk assessed on the day of the scan.
Between a recommendation and the follow-up scan, there is no new information to work with. In 696 incidental lung nodules, 34% of patients had no documented follow-up and 5 early-stage cancers progressed to stage III or IV. The mSI™ puts a risk assessment in the chart at the time of the read. In our retrospective clinical studies, mSI™ was predicted to accelerate cancer follow-up in 45% of cases where guidelines recommend interval surveillance.
Follow-up: Pannu et al., CHEST Pulmonary, 2025; 696 incidental pulmonary nodules, Ohio State University. Putting the lung nodule in the report impression reduced non-concordant follow-up by 45% in the same series. Reclassification: Adams et al., Journal of the American College of Radiology, 2023; Duerden et al., Clinical Radiology, 2026.
The highest-scoring lung nodules rise to the top.
A navigator working a hundred incidental findings cannot call everyone with equal urgency, and today has nothing to prioritize on except millimeters. When the radiology report carries a Malignancy Similarity Index (mSI™) value, the worklist can be ordered on something that actually distinguishes one nodule from another.
The effect is not that the program does more. It is that the same team spends its first hours of the day on the patients most likely to need them — and can document why.
A nodule scoring below 0.1 carries features closely resembling confirmed benign nodules. That supports continuing surveillance at the guideline-recommended interval.
What mSI™ changes for each role.
“Who in this queue needs a call today?”
You are working a list, and there are only so many calls in a day. The mSI™ is in the radiology report, so it travels with that report into the system you track nodules with. The worklist can then be ordered by risk instead of by scan date or nodule size. The patients with a high mSI™ come to the top, and those are the calls you make first.
Making the evidence visible.
Shared decision making is hardest when the evidence is qualitative. You are weighing size, growth, history, and how much waiting the patient can tolerate. What they hear is “it’s probably nothing, but let’s keep an eye on it.” The mSI™ provides a more objective, quantitative value to weigh the risk. A high score sharpens the case for moving sooner; a low one supports staying the course with confidence rather than anxiety.
“What is the pre-test picture before this goes to the multidisciplinary team?”
You are deciding whether this patient needs tissue. The mSI™ is already in front of you, measured on the scan that found the lung nodule rather than reconstructed months later. A high score sharpens the case for intervention; a low one supports continued surveillance. It gives the team and the patient the same thing to look at.
“How do we know the program is working?”
A recorded, comparable value on every characterized nodule lets the program be reviewed on consistent data — follow-up concordance, time to workup, and where patients fall out of the pathway.
Add revenue, prioritize your worklist, and lose fewer patients to follow-up.
Lung nodule programs are usually funded as overhead — navigator salaries, tracking systems, coordinator time — against a benefit that is real but hard to attribute. mSI™ gives the program what it has never had: a billable service attached to every characterized nodule, and a quantitative value to manage risk and prioritize patients.
| What mSI™ adds | How it works |
|---|---|
| A billable service | 0721T covers characterization performed on a previously acquired CT and is reported on its own. 0722T is the add-on used when characterization is performed alongside a concurrent CT. Either is billed once per patient per encounter, regardless of how many nodules are characterized. Both are carrier-priced Category III codes, so the rate is set by your MAC and varies by locality. |
| An order to the queue | Every patient on the list needs a call. What is usually missing is a reason to make one call before another, so the order ends up being whoever answers the phone, or alphabetical, or however the list is exported. A reported value on each nodule gives the queue a clinical order. The same list gets worked, in a sequence the navigator can explain and defend. |
| Fewer patients worked up who did not need it | A low mSI™ adds evidence in support of the guideline-recommended interval, so a patient can stay in routine surveillance rather than being escalated to further imaging or biopsy. |
| Quality assurance | Every characterized nodule carries the same objective measure, so the program can be reviewed on consistent data — follow-up concordance, time to workup, and where patients drop out. |
Program design, not just software.
The software installs in five to ten business days. The program around it takes longer to build: who orders the characterization, where the result lands, who watches the queue, and what happens when a patient does not come back. We work through each of those with your team.
| What we help with | Detail |
|---|---|
| Pathway design | How detection, characterization, navigation and intervention fit together as one pathway |
| Order construction | How 0721T and 0722T orders are built, and who the appropriate ordering provider is. We work this through with your billing and compliance teams, because it depends on your setting. |
| Report integration | Structured output and addendum templates so the finding lands in the impression, where it drives follow-up |
| Tracking | Surfacing the reported value in existing incidental-findings dashboards and navigator worklists |
| AI governance | Material for committees distinguishing detection from characterization, and where each sits in the pathway |
We aren’t just software, we are a part of your program.
We work with radiology, pulmonology and navigation together, because a nodule program only functions when all three are in the room.