What Bassir is, what it is not, and how it handles your data.
Every safety claim on our website links to this page. When something changes, we change the version and the date.
- 01Intended use
- 02How your image travels
- 03Your data
- 04How well Bassir performs
- 05Certifications and regulation
- 06Documents
- 07Report a problem
01Intended use
What Bassir is
- AI software that analyses adult chest X-rays (PA and AP views) and suggests findings with their location and confidence.
- For doctors and medical trainees who read chest X-rays.
- Decision support: the doctor confirms, edits or rejects every finding.
What Bassir is not
- Not a certified medical device. Do not use it for patient-care decisions during the beta.
- Not a replacement for a reading by a qualified doctor, and not for emergencies on its own.
- Not validated for children, lateral views, CT or other imaging.
02How your image travels
- You uploadDICOM, PNG or JPEG, over an encrypted connection.
- Personal data is removedNames, IDs and dates are removed from the file header before anything else.
- Quality checkImages that are too blurred, dark or cropped are not analysed. Bassir tells you why.
- AI analysisFindings, heatmap and explanation. Each one is marked as an AI suggestion until you review it.
- Kept, then deletedKept for [30] days for your case history, then deleted. You can delete it earlier.
03Your data
- Where it is stored
- [Hosting provider and region to decide]
- Encryption
- In transit and at rest [method to state]
- AI training
- We do not train on your uploads without a written data agreement.
- Who can see your cases
- You. Our clinical reviewers see a case only when you send feedback on it.
Companies that process data for us
| Company | Purpose | Region | Data |
|---|---|---|---|
| [Cloud provider] | Hosting and AI analysis | [Region] | De-identified images, results |
| [Email provider] | Account and notification emails | [Region] | Name, email |
| [Analytics provider] | Product usage counts | [Region] | Events, no images |
04How well Bassir performs
Results on [test set: number of images, sources, period], read by [number] radiologists as the reference. We publish every finding, including the weak ones. Until the validation is complete, the figures stay empty.
| Finding | Images with it | Sensitivity | Specificity | Known weak points |
|---|---|---|---|---|
| Consolidation | [n] | [pending validation] | [pending validation] | [to describe] |
| Pleural effusion | [n] | [pending validation] | [pending validation] | [to describe] |
| [Finding 3] | [n] | [pending validation] | [pending validation] | [to describe] |
| [Finding 4] | [n] | [pending validation] | [pending validation] | [to describe] |
| [Finding 5] | [n] | [pending validation] | [pending validation] | [to describe] |
| [Finding 6] | [n] | [pending validation] | [pending validation] | [to describe] |
| Image quality check | [n] | Rejection rate: [x %] of uploads | Photographs of film are rejected more often | |
Not tested on children. Lower performance is expected on AP portable films, on photographs of film and screens, and on images with lines, tubes or devices. Rare findings outside the six listed are not detected.
05Certifications and regulation
We show the real status of each item. "Planned" means that we have not started the certification yet.
06Documents
07Report a problem
Use "Report a problem" on any result, or write to [safety email]. A clinician from our team answers within 24 hours on working days.
Report a problem