From the information available, Chanieldx appears to offer a very practical solution for chairside diagnostics that respects the clinician’s final say. The fact that every output is framed as a suggestion rather than an autonomous diagnosis is a significant strength, as it integrates smoothly into existing workflows without demanding a paradigm shift. The emphasis on offline capability is particularly appealing; storing images before inference ensures that patients are never left waiting or losing data due to connectivity hiccups. The colour-coded overlays are designed to differentiate findings without implying a pass or fail verdict, which feels like a nuanced approach to visual data. One minor hesitation is the reliance on a bootstrap dataset for some metrics, which suggests that while the infrastructure is solid, the clinical evidence base is still growing. However, the commitment to full provenance and the clear distinction between built features and future roadmap items builds a good deal of trust in the product’s honesty.
Questions about Chanieldx
What is Chanieldx?
Chanieldx provides chairside diagnostic AI software for dental radiography that assigns FDI tooth numbers to detected teeth on radiographs. It runs quality pre-checks on exposure and contrast, and surfaces colour-coded lesion overlays across bitewing, periapical, and panoramic modalities. The platform is designed as a decision-support tool rather than an autonomous diagnostic system, allowing licensed clinicians to accept, correct, or dismiss findings with a full audit trail.
How does the tooth numbering feature work?
The system performs quality pre-checks on exposure, contrast, and sharpness before sending anything for analysis. It uses a stage 0 model committed to the process, with a browser heuristic standing in today. Every detected tooth gets an FDI number, and a wrong number can be corrected with one tap.
What are the supported imaging modalities and how are findings displayed?
Chanieldx supports three modalities within the Cariology scope: bitewing, periapical, and panoramic. The platform deliberately avoids red or green overlays to prevent the findings from reading as a pass/fail clinical verdict. Each finding type gets a distinct hue chosen for differentiation, with related findings sharing a family.
How does the clinician intervention process function?
Clinicians can accept, correct, or dismiss findings, and a dismissal stays distinguishable from the model finding nothing. Every decision is attributable with a full audit trail. The review and record process ensures that each action is logged and traceable.
Who built Chanieldx and what hardware requirements are there?
The platform was built by Sephar-Innovations for Chaniel Digital Health. It is designed for modest hardware and for connectivity that comes and goes. The system aims to provide a second opinion anchored to the tooth for the dental chair.
What are the current accuracy metrics for the caries detection models?
The caries detection model has a mAP50 of 0.35, which is described as real but weak. This metric comes from training on bootstrap data only and is expected to improve once proprietary Nigerian radiographs are annotated. The deep caries model has a mAP50 of 0.50 as a coarse binary depth signal.
How accurate is the impacted tooth and periapical lesion detection?
The impacted tooth detection model achieves a mAP50 of 0.91, but this comes along for free from the shared dataset. It is presented muted as it is outside the project's committed clinical scope. The periapical lesion model currently shows a mAP50 of 0.31 based on nine validation boxes.
What are the current accuracy claims for the system?
No accuracy claims are made for any finding type until the Nigerian validation set is scored. The platform states that model figures shown are training-set metrics from public bootstrap datasets and have not been measured against Nigerian radiographs. This approach ensures transparency in the current performance levels.
How does the system handle audit trails and record keeping?
The platform provides a full audit trail for every decision made by the clinician. This ensures that all actions are recorded and can be reviewed later. The system is built to provide a second opinion for a licensed clinician.
What technology is used for the initial model processing?
The platform uses a browser heuristic to stand in for the stage 0 model today. It is designed to work with connectivity that comes and goes. The system is built for the chair and focuses on modest hardware requirements.
How can I contact Chanieldx?
You can reach Chanieldx by e-mail at [email protected].
What is the official website of Chanieldx?
The official website of Chanieldx is chanieldx.com.
How is Chanieldx rated on Trustburn?
On Trustburn, Chanieldx has a rating of 3.8 out of 5 based on 5 reviews.
Post your question publicly — other visitors, and whoever manages this page, can answer. It will be visible on this page.
Want to hear about this company again? Follow it and we’ll email you when a new review appears.
Based on reviews, listing data and the company's website. Verify anything important with the company directly.
While the platform claims high accuracy for tooth numbering, the performance metrics for actual disease detection seem less robust. The cariometry mAP50 is listed at 0.35, which indicates weak performance, and the periapical lesion detection is at 0.31. These numbers are derived from bootstrap data or a limited validation set, meaning the clinical utility for diagnosing actual decay is still unproven. It is crucial for clinicians to understand that these early metrics do not guarantee reliable diagnostic support yet.
The FDI tooth numbering accuracy is fantastic. It really anchors every finding perfectly and makes the whole diagnostic process feel much more reliable and precise for daily use in the clinic.
The way Chanieldx presents its capabilities is refreshingly honest. I appreciate that they explicitly state when a feature is built versus when it is still in development. It makes it easy to understand what is actually available right now versus what is coming later, without any hidden marketing fluff. The distinction between what is committed and what is merely on the roadmap is very clear.
Reading through the technical documentation for Chanieldx, what stands out is the thoughtful approach to connectivity issues in a busy clinic environment. It is reassuring to see that images are stored before inference is attempted, ensuring that a service outage never results in the loss of a patient capture. This design decision addresses a very real pain point for dental professionals who cannot afford to lose diagnostic data just because the internet decided to disconnect. The platform also seems to handle varying hardware conditions well, with a quality pre-check that measures exposure, contrast, and sharpness before sending data for analysis. Rather than simply blocking the workflow with a bare retake request, the system provides a specific reason for the issue. This level of detail suggests a product built by people who understand the practical realities of chairside diagnostics. The commitment to providing a second opinion anchored to the tooth, rather than an autonomous diagnosis, feels like a responsible way to integrate AI into clinical workflows. It respects the clinician’s authority while offering a powerful analytical tool that can help catch details that might otherwise be missed during a busy shift.
Questions & Answers
No questions yet — be the first to ask.
Sign in to post
Your text is saved — you'll come right back to post.
Don't have an account? Create one
News & Press
No news yet. This is where the business shares its own updates and press.
Photos
No photos yet. Photos shared by reviewers will show up here.
About Chanieldx
Chanieldx provides chairside diagnostic AI software for dental radiography. The platform assigns an FDI tooth number to each detected tooth on a radiograph, runs quality pre-checks on exposure and contrast, and surfaces colour-coded lesion overlays across bitewing, periapical, and panoramic modalities. It is designed as a decision-support tool rather than an autonomous diagnostic system, allowing licensed clinicians to accept, correct, or dismiss findings with a full audit trail. The site notes that a Nigerian validation set is intended to establish accuracy claims, which are not currently asserted.
- Website
- chanieldx.com
- [email protected]
- Categories
- Computer software, Biotechnology, Medical practice
