Introduction to CTC - Interactive Guide
Step-by-step guide on the use of Harrison.ai CT Chest (CTC)
0Frequently Asked Questions
- For what findings is volumetry available?
CTC will provide volumetry for all nodular and lesion findings. At present only lung nodule volumetry is able to be copied to clipboard directly from the application.
- Do I use CTC in place of my PACs viewer?
No, the CTC viewer displays a compressed JPEG rather than the full DICOM image. We recommend using the Harrison Viewer along-side your PACs to interrogate the findings listed.
- Can I expand the application automatically to reduce clicks?
Yes, the application behaviour can be adjusted using the settings (cog) icon. See the Settings Guide (interactive or video) for more details.
- How should I interpret the confidence bar and the detection threshold?
The confidence bar provides a human element to the AI output. This can be used to inform your clinical judgement as to the presence/absence of each finding. Each finding must meet a threshold to be listed in the viewer; this threshold is shown as the vertical white bar on the confidence scale.
- Is the confidence of a detection the same for all instances of a finding, i.e. for each lung nodule detected?
No, confidence for each nodule, or each presentation of a finding, is calculated individually.
- Can I change the group which a finding belongs to?
Yes, this can be updated to match your practices clinical needs. Contact your customer success manager for more information.
- Is it possible to adjust the thresholds of findings?
In most cases yes, contact your customer success manager for more information.
- What findings does CTC detect?
There is a full ontology tree (findings list) available here. Keep in mind that your specific practice may have a customised findings list, contact your system administrator or customer success manager for the exact configuration.
- What findings aren't detected by CT-Chest?
The CT-Chest model will cover from the cervical vertebrae down to the top of the adrenal glands. Findings outside that anatomical boundary won't be detected by the model. For a full ontology see here. With the field the following findings are not detected by the model:
- Emphysema
- Inflammatory lung disease
- Bronchiectasis
- ILA scores
- Right heart strain
- Osteoporosis