# CarDSScan — AI enabled ECG analysis > On-device AI ECG analysis for research and study workflows. (iPhone/iPad app by Rohan Khera.) - Source: https://appshunter.io/ios/app/cardsscan/id6751209518 (this page in markdown: same URL + `.md`) - Developer: [Rohan Khera](https://appshunter.io/developer/1709545475) - Category: Health & Fitness - Price: Free - Rating: 5.00/5 from 1 App Store ratings - Age rating: 17+ - Requires: iOS 17.6 · 231 MB - Languages: American English - Released: 2025-08-21 - Data updated: 2026-08-31 - Monetization: free - User reviews in markdown: https://appshunter.io/ios/app/cardsscan/id6751209518/reviews.md ## What is CarDSScan? CarDSScan delivers research-grade, on-device ECG analysis from the Cardiovascular Data Science (CarDS) Lab. Built for reliability and privacy, CarDSScan runs its machine-learning models entirely on your iPhone - no internet required and no cloud uploads. Two ECG analysis engines ECG-Vision ATTR: Capture or import clear photos of printed or screen-displayed ECGs for fast, fully on-device analysis. ECG-GPT: A transformer-based ECG interpretation model designed for research-grade signal understanding. ECG-GPT runs locally on your device and may require approval for access. Private by design All inference happens locally on your device. ECG images and outputs stay with you unless you explicitly export them. Certain advanced models may be downloaded after approval. Developed by CarDS Lab Models are developed by the Cardiovascular Data Science (CarDS) Lab and have been evaluated across multi-center research datasets. CarDSScan is intended for research use only and is not a medical device. Key Features On-device ML inference (offline; no cloud processing) 12-lead ECG image capture and import ECG-Vision ATTR CNN model and ECG-GPT transformer-based model for research workflows Secure model download for approved users Clean, professional UI designed for study workflows Fast processing with reproducible outputs Optional export for research documentation Who It’s For Research teams, engineers, and clinicians conducting feasibility studies, validation work, and algorithm benchmarking in controlled settings. Privacy at a Glance No automatic data collection No ads, no trackers, no data sale Data stays on device unless you export it Requirements iPhone with camera (iOS 17.6 or later) Important CarDSScan is for research use only. It does not provide medical advice and is not intended for diagnosis or treatment. Do not make clinical decisions based on the app’s outputs. ## Key features - On-device AI ECG analysis - Offline processing - ECG image capture and import - Two AI analysis engines (ECG-Vision ATTR, ECG-GPT) - Privacy-focused design - Secure model downloads - Research documentation export ## Frequently asked questions about CarDSScan ### What is CarDSScan used for? CarDSScan is designed for research use only, enabling researchers, engineers, and clinicians to perform on-device ECG analysis for feasibility studies, validation, and algorithm benchmarking. It is not intended for medical diagnosis or treatment. ### Does CarDSScan require an internet connection? No, CarDSScan performs all its machine learning inference entirely on your iPhone. This means no internet connection is required, and no cloud uploads are necessary, ensuring your data remains private. ### How does CarDSScan ensure user privacy? CarDSScan is built with privacy at its core. All ECG analysis happens locally on your device, and your ECG images and outputs stay with you unless you explicitly choose to export them. There is no automatic data collection, ads, trackers, or data sale. ### What are the two ECG analysis engines in CarDSScan? CarDSScan features ECG-Vision ATTR, a CNN model for analyzing photos of ECGs, and ECG-GPT, a transformer-based model for advanced signal interpretation. Both run locally on your device. ### What devices is CarDSScan compatible with? CarDSScan is compatible with iPhones running iOS 17.6 or later. It requires a camera for capturing ECG images. ### Is CarDSScan a medical device? No, CarDSScan is intended for research use only and is not a medical device. It does not provide medical advice, and users should not make clinical decisions based on its outputs. ### How often is CarDSScan updated? The latest version, 4.0.3, was last updated on February 21, 2026. Updates typically focus on model improvements, UI enhancements, and compatibility with the latest iOS versions. ### Does CarDSScan have ads? No, CarDSScan is ad-free. The app is developed with a focus on user privacy and research integrity, and it does not contain advertisements or trackers. ## Version history (last 5 releases) ### 4.1.0 — 2026-08-27 Minor bug fixes. Improvement to ECG detecting algorithm. ### 4.0.3 — 2026-02-21 New ECG-GPT model added to the list of models available for inference! ### 3.0.1 — 2025-11-20 AI Model optimization to run in an optimal manner on the neural engine. ### 3 — 2025-11-05 Account creation and report sharing features for compliance! ### 2 — 2025-10-17 We’ve made this update even smarter! Added an advanced YOLO-powered auto-cropper that automatically detects and crops ECG regions for faster, cleaner analysis. Improved image-processing accuracy and model stability for better on-device performance. General performance and UI refinements for a smoother experience. ## Apps similar to CarDSScan | App | Rating | Price | Category | | --- | --- | --- | --- | | [Qaly \| ECG Reader](https://appshunter.io/ios/app/qaly-ecg-reader/id1542440769) | 4.7 (853) | Free | Health & Fitness | ## More apps by Rohan Khera - [CarDSPlus](https://appshunter.io/ios/app/cardsplus/id6470764242) All apps by Rohan Khera: https://appshunter.io/developer/1709545475 ## Related topics [ecg buddy research](https://appshunter.io/ios/topics/ecg-buddy-research) · [ekg reader](https://appshunter.io/ios/topics/ekg-reader) · [ekg interpretation](https://appshunter.io/ios/topics/ekg-interpretation) --- *Data collected daily from the US App Store and indexed by [AppsHunter](https://appshunter.io/). User reviews are verbatim App Store reviews. Ratings, prices and chart positions refresh continuously; this snapshot is from 2026-08-31.*