Imaging and Report Review
From the report to the next clinical step
Imaging and Report Review starts from a report, or from the image itself, and ends with a clinical decision. It separates what the radiologist or pathologist saw from what they concluded, applies the classification that governs management of each finding, deals with every incidental finding explicitly, and says whether the study has answered the question it was requested for and, if not, what would.
Who it is for
- Physicians and surgeons who receive a report with an unexpected finding and need to know what follow-up it calls for
- Residents and registrars preparing to discuss a scan with a consultant or at a multidisciplinary meeting
- Emergency and acute medicine doctors looking at a radiograph or CT before the formal report is available
What it does
Findings kept apart from the impression
Reads a report in layers: the technique and its limits, such as contrast phase and image quality, what was described, and what was concluded, so that hedged wording is not taken as certainty and an insensitive normal study is not taken as exclusion.
Classification systems applied by name
Applies the system that governs management of each finding: Fleischner Society guidance and Lung-RADS for lung nodules, BI-RADS, TI-RADS, PI-RADS, LI-RADS, O-RADS and the Bosniak classification, and the equivalent tumour grading and staging schemes for pathology reports.
Incidental findings handled one by one
Lists each incidental finding with whether it needs follow-up, by which test and at what interval, or whether it can be left alone, so that an adrenal nodule or a renal cyst does not fall between teams.
A systematic read of an attached image
Names the modality and projection and judges whether the image is adequate, then works through the whole film methodically without stopping at the first finding, and rechecks places where abnormalities are often missed, such as the lung apices, hila, retrocardiac region and costophrenic angles.
The test that answers the question
Recommends the investigation most likely to answer the clinical question, drawing on appropriateness guidance such as the American College of Radiology criteria or iRefer, and covers contrast and kidney function, metformin, pregnancy, radiation dose and MRI safety.
Questions to take back to the reporter
Suggests what to ask the radiologist or pathologist when a report leaves the clinical question open, such as a comparison with earlier imaging, a volumetric measurement or an additional stain the next decision depends on.
How it works
- Paste the report text, or attach the report or the image, and state the clinical question the test was meant to answer.
- Add the history that changes interpretation, such as smoking, previous cancer and kidney function, and any earlier imaging for comparison.
- The engine works through technique, findings and impression, then applies the governing classification to each significant finding.
- Incidental findings are listed separately with the follow-up each needs, and a differential is given for the main abnormality.
- The report closes with the recommended next step and its urgency, the information that would sharpen the answer, and verified references.
What to include
- The report text or an uploaded report or image
- The clinical question that prompted the test
- Relevant history, including smoking and previous cancer
- Previous imaging for comparison
Example cases
- CT chest: 7 mm solid nodule in the right upper lobe, 58-year-old ex-smoker of 30 pack-years. Follow-up?
- MRI lumbar spine: L4/5 disc protrusion contacting the left L5 root, 6 weeks of sciatica without weakness. What does this change?
- Incidental 2.1 cm adrenal nodule of 8 Hounsfield units on unenhanced CT. Work-up?
What the report contains
- Conclusion and Recommendation
- What the Study Shows
- Significant Findings
- Incidental Findings
- Differential Diagnosis
- Recommended Next Step
- Information Needed
- Key Evidence & References
Limits
- An image read here supports, but does not replace, the formal radiology or pathology report. Findings that need urgent formal review are named as such, and the reporting radiologist or pathologist has the final word.
- The reading is only as good as what is supplied: a photograph of a screen, a single slice, or a report without the clinical question limits what can responsibly be said.
- Follow-up intervals and classification thresholds differ between societies, versions and countries. Local radiology protocols and multidisciplinary team decisions take precedence.
Frequently asked questions
When should Fleischner guidance be used instead of Lung-RADS?
The Fleischner Society guidance is for nodules found incidentally on CT performed for another reason in adults aged 35 or over. It does not apply to lung cancer screening, to patients with known cancer or to immunocompromised patients. Lung-RADS is the reporting and management system for CT scans performed within a screening programme. The engine names which system it is applying and why.
How should an incidental adrenal nodule be worked up?
Two questions apply: is the lesion benign on imaging, and is it secreting hormone? A homogeneous nodule of 10 Hounsfield units or less on unenhanced CT is consistent with a lipid-rich adenoma. Hormone assessment, including an overnight dexamethasone suppression test and, depending on imaging and blood pressure, tests for phaeochromocytoma and primary aldosteronism, follows the European Society of Endocrinology guideline and local endocrine advice.
What does the Bosniak classification of a renal cyst mean?
It grades cystic renal masses on CT or MRI features, such as wall thickness, septa, calcification and enhancing solid components, into categories that rise with the likelihood of malignancy. The lower categories are benign and need no follow-up, the intermediate category needs imaging surveillance, and the highest categories are usually referred to urology. The engine names the version used, because the 2019 update changed some definitions.
Can I upload an X-ray or a CT image rather than the report?
Yes. Images can be attached to a question. The engine begins by identifying the modality and projection and whether the image is of diagnostic quality, then reads it systematically, describing appearances before interpreting them, offering a differential and stating its confidence. It also makes clear that this does not stand in for the formal report.
Related articles
Clinical article
The incidental pulmonary nodule: applying the Fleischner Society guidance
When Fleischner applies and when Lung-RADS does, how to measure and risk-stratify a nodule, follow-up intervals for solid nodules, and when to use PET-CT.
7 min read
Related engines
Breathlessness, airways disease and the pleura
Emergencies, staging and treatment toxicity
Patterns across a panel, and what to test next