DigestAID
DigestAID AI Software is a clinical workspace bringing eighteen AI applications for digestive endoscopy, endoscopic ultrasound, motility testing, bronchoscopy and anoscopy into one platform that detects, differentiates and prioritises lesions before the clinician validates and reports.
What is DigestAID?
DigestAID AI Software is the clinical workspace built by DigestAID, a Portuguese company based in Porto and founded by gastroenterologists and engineers, led by Miguel Mascarenhas, MD, PhD. The vendor presents it as one integrated AI platform extending across multiple clinical specialties and procedures, and it is best understood as an application ecosystem rather than a single algorithm.
Eighteen named clinical applications sit inside that ecosystem, arranged in six clusters. Capsule endoscopy is covered by Deep Capsule, GI Spy, IBD Smart and Deep Bowel. The pancreatobiliary tract is covered by Smart Cholangio, Smart Cyst, Smart Pancreas, Sonomind and Smart Node. Digestive motility is covered by DigestiFlow AI, Smart Mano and Smart Esophagus, pulmonology by Smart Bronchus and Smart Node EBUS, coloproctology by HPV Smart and Anoscopy Genius, and surgical work by LeakSight and PrecisionRect AI. Between them they address small-bowel and colon capsule video, device-assisted enteroscopy, digital single-operator cholangioscopy, endoscopic and endobronchial ultrasound, anorectal and oesophageal manometry, FLIP panometry, high-resolution anoscopy, colposcopy, hysteroscopy and laparoscopy.
The software layer is what ties them together. DigestAID describes a single journey: an examination is brought into its patient context, AI-assisted findings are reviewed while remaining anchored to the source image, and validated information becomes a structured report. Patient context, study progression and reporting activity stay in one view, and every specialised application keeps a shared visual language.
The underlying technology is described as tailor-made convolutional neural networks. The homepage claims twenty-one patent-protected AI technologies, a dataset of ten million examinations, global diagnostic accuracy above 98 percent and more than five hundred scientific publications and awards. Individual products carry their own figures: Deep Capsule reports a 96.1 percent validated lesion-detection rate and a mean AI reading time of 203 seconds across seven centres in four countries, with peer-reviewed validation in The American Journal of Gastroenterology.
The vendor insists the clinician stays in charge, keeping final interpretation in expert hands. Industrial partnerships with Olympus and Jinshan are announced. There is no public pricing, no self-service trial and no API: access starts with a contact form.
What it does
- Detect lesions automatically across capsule, enteroscopy, cholangioscopy, ultrasound and anoscopy images
- Differentiate findings, including benign from malignant strictures, cysts, lymph nodes and mucosal lesions
- Stratify the haemorrhagic potential of small-bowel lesions on a validated scale
- Prioritise clinically relevant frames so a long examination can be reviewed quickly
- Identify and classify digestive motility patterns, including Chicago 4.0 oesophageal classification
- Guide tissue sampling with bounding boxes drawn around the relevant area during cholangioscopy
- Produce a standardised diagnostic report from the findings the clinician has validated
When to use DigestAID / When not to
A quick filter to help you decide if DigestAID is the right fit.
When to use DigestAID
- Gastroenterology and endoscopy units that read long capsule endoscopy videos and want relevant frames surfaced first
- Pancreatobiliary teams performing digital single-operator cholangioscopy or endoscopic ultrasound on strictures, cysts and lymph nodes
- Digestive motility laboratories running anorectal manometry, high-resolution oesophageal manometry or FLIP panometry
- Interventional pulmonology services working with endobronchial ultrasound images
- Coloproctology and HPV screening clinics performing high-resolution anoscopy, and research centres running multicentre clinical validation studies
When not to use DigestAID
- Anyone looking for a self-service tool: there is no sign-up, no free trial and no published price anywhere on the site
- Buyers who need a firm regulatory answer today, since the vendor itself says several technologies are still in its research and development pipeline
- Developers wanting to embed the models, because no public API, developer documentation or integration is offered
- Clinicians hoping for a mobile companion app, as none exists on either the App Store or Google Play
- Procurement teams that require a published contractual framework, since the site publishes no terms and conditions and no legal notice
How to use DigestAID
A typical end-to-end flow, from setup to results.
- Identify which of the eighteen clinical applications matches your procedure, using the Products page and its six clusters
- Send an enquiry through the contact form, choosing an area of interest such as product information, clinical collaboration or commercial partnership
- Discuss scope, market availability and regulatory status with the vendor's team, since none of this is published
- Receive an account for the web application; registration asks for hospital or institution, country, licence number and medical specialty
- Enter the workspace, where each clinical application opens into its own focused environment
- Bring an examination into context, either a previously acquired video or a live procedure depending on the product
- Let the AI analyse the study and separate normal appearances from clinically relevant findings
- Review the prioritised frames, with each AI-assisted finding kept anchored to its source image
- Validate or reject the findings in their clinical context, since final interpretation remains yours
- Turn the validated information into a standardised diagnostic report from the same workspace
Pros & Cons
Pros
- Unusually wide coverage for a single vendor: eighteen applications across six clusters and several specialties beyond gastroenterology
- Named peer-reviewed validation, including Deep Capsule in The American Journal of Gastroenterology and Smart Cholangio in Clinical and Translational Gastroenterology
- Performance figures given product by product rather than only in headline claims
- Multicentre clinical validation announced for the flagship application, across seven centres in four countries
- Industrial partnerships with Olympus and Jinshan, with a verifiable press release for the first
- Explicit assistive positioning, with final interpretation left to the clinician
- One platform rather than isolated tools, keeping patient context, study progress and reporting connected
Cons
- No public pricing, no plans and no trial: every path leads to a contact form
- No terms and conditions and no legal notice, so no public contractual framework exists
- The CE marking and FDA 510(k) claim carries no certificate number and no notified body, and covers only four of the eighteen products
- The vendor states elsewhere that its showcased technologies belong to a research and development pipeline and may be pending regulatory approval
- No hosting country, region or named subprocessor is disclosed, and no data processing agreement is offered
- The privacy policy dates from August 2021, still names an outdated legal form and links to a page that no longer exists
- Nothing is published about the examination data and patient images the products process, nor about model training
Pricing & Plans
No price is published. The site displays no pricing page, no plan and no amount on any page, and there is no free plan or free trial. Every call to action leads to the contact form, and the privacy policy confirms that the service is sold on subscription by referring to the end of a subscription period. Pricing must therefore be obtained from the vendor's team.
Data, GDPR & hosting
A consolidated view of how DigestAID handles your data.
GDPR overview
The privacy policy is written explicitly under Articles 13 and 14 of the GDPR, and the About page lists governance as GDPR compliant, with international ethics committee approval. The six data subject rights are named individually, with a thirty-day response commitment and a single contact address, info@digestaid.health. Beyond that the implementation is thin. No data protection officer is designated, no supervisory authority is named and no right to lodge a complaint is mentioned. No hosting jurisdiction is disclosed, no subprocessor list is published and no data processing agreement is offered. The publisher is established in Portugal, so no Article 27 representative is required or named. The policy was last updated on 25 August 2021, still names the company as a Lda. and links to a contact page that now returns an error.
Who owns the data?
DIGESTAID is named as the data controller for its website, and the privacy policy states that all copyright, trademarks and other intellectual property rights belong to DIGESTAID - Artificial Intelligence Development, Lda. Users keep the six GDPR rights the policy lists: access, rectification, erasure, restriction, objection and portability, with a stated thirty-day response time and a possible small fee for supplying copies. Requests go to a single mailbox, info@digestaid.health. The policy addresses website visitors and web-application account holders only. It says nothing about who owns the examination data, the endoscopic video or the patient images that the AI products actually process.
Reuse rights
No terms and conditions are published, so no document states what an end user may do with the material the platform produces. The privacy policy covers personal data alone. It lists what is collected: name, email address and telephone number from the website, and name, email address, telephone number, hospital or healthcare institution, country, licence number, medical specialty and profile picture from web-application accounts. The declared purposes are handling demo requests, answering product questions, sending news and offers, improving the user experience and, in the vendor's own words, to make scientific and historical research about our services. Google Analytics and Hotjar are named as indirect sources. Marketing consent can be withdrawn at any time. Nothing in the policy addresses reuse rights over clinical outputs, and nothing addresses the training of AI models on customer data.
Data retention & training
Hosting summary
Hosting is not disclosed. The privacy policy says only that DIGESTAID stores data at its data centres and secures it with SSL certificates issued by the cPanel, Inc. Certification Authority. No country, region or datacentre operator is named, and no jurisdiction is stated. Third parties are described by category rather than by name: website hosting, web analytics, cloud compute, file storage and email marketing providers are said to act as data processors, and some data may be transferred to and stored by them. The only named third parties anywhere are Google Analytics for the website and Hotjar for the web application, both cited as indirect sources of data rather than as hosting providers. No subprocessor list is published, no transfer mechanism outside the European Economic Area is discussed, and no data processing agreement is offered. Since the publisher is established in Portugal, an EU jurisdiction is plausible but is never confirmed by the site itself.
Things to keep in mind
Risks and trade-offs to weigh before adopting DigestAID.
- Automation bias: prioritised frames make review faster, but a reader who trusts the ranking may stop looking at what the model did not flag
- Deskilling: trainees who learn to read capsule or cholangioscopy studies through an AI overlay may build weaker unaided pattern recognition
- Regulatory uncertainty: the CE and FDA claim covers one cluster only, carries no certificate number, and the vendor says other technologies may still be pending approval in your market
- No contractual protection: with no terms and conditions and no legal notice published, liability, service levels and data responsibilities are undefined until a contract is negotiated
- Opaque data handling: no hosting country, no subprocessor list and no data processing agreement, on a product that by nature touches patient images
- Silence on model training: nothing states whether customer or examination data may be used to improve the models, and no opt-out is offered
- Stale legal information: a privacy policy from 2021 that names an outdated legal form and links to a dead page is a poor basis for a compliance review
Setup & Integrations
Technical difficulty
There is nothing to install by yourself, and nothing you can start alone either. Access is a web application reached after an enquiry, and account creation asks for licence number and medical specialty, which implies vendor-side validation. The real effort is organisational: agreeing scope, checking regulatory status for your market, and connecting the products that run during a procedure to your endoscopy acquisition chain, which the site never describes. No documentation, no API and no alternative deployment mode are published, so expect the vendor's team to lead the setup.
Deployment
Behind DigestAID
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What exactly is DigestAID AI Software?
Which specialties and procedures does it cover?
Does the platform make a diagnosis?
How much does it cost?
Is there a free trial or a free plan?
Is there an API?
Is there a mobile application?
Are the products CE marked or FDA cleared?
What data is collected when an account is created?
How long is data kept?
Should you pick DigestAID?
DigestAID is one of the more substantial clinical AI catalogues a single vendor publishes: eighteen applications, six procedure clusters, several specialties, and named peer-reviewed validation for its flagship products rather than headline percentages alone. The academic backing is visible, the partnerships with Olympus and Jinshan are real and documented, and the assistive framing is consistent: the platform surfaces and characterises findings, the clinician decides.
The weakness is transparency rather than substance. Nothing about the commercial offer is public. There is no price, no plan, no trial and no self-service path, and there is neither a terms page nor a legal notice, which leaves a buyer with no published contractual framework. The regulatory picture needs care: CE marking and FDA 510(k) clearance are claimed for the capsule endoscopy cluster only, without a certificate number or a notified body, while other pages describe the showcased technologies as part of a research and development pipeline that may still be pending approval in a given market. Those two statements are not contradictory, but they mean the regulatory status has to be established product by product and country by country before any purchase.
Data governance is similarly incomplete. The privacy policy is written under the GDPR and lists the data subject rights properly, but it dates from 2021, names an outdated legal form, links to a page that no longer exists, and covers only the website and the web application. The examination data and patient images the products process are never addressed, and no hosting jurisdiction, subprocessor list or data processing agreement is published.
For an endoscopy department already convinced by the published evidence, this is a serious candidate worth a conversation. For anyone who needs to compare offers on documented terms, the evaluation cannot start on the website alone.
- Choosing a selection results in a full page refresh.
- Opens in a new window.