Radiology & Medical Imaging
CT, MRI, ultrasound, and mammography: instant recognition of sequences (T1, T2, FLAIR, DWI), classifications (BI-RADS, Lung-RADS, TI-RADS), and 4-section structuring.
Dictate right at the cursor.
WannaMed ASR: Medical AI Dictation Bar
A 46-pixel floating bar that types directly into your medical software on Windows and Mac, calibrates to your voice, leverages a predictive dictionary of 14,000+ terms, and guards clinical consistency in every report.
Decision-support software for physicians, who remain solely responsible for validating every clinical report.
Zero keyboard focus loss
On both Windows and Mac, clicking the floating bar or pressing global hotkeys never steals focus from your PACS, RIS, or Word document.
Active idle bar (820 × 46 px)
One-click access to dictation, microphone selector, medical specialty, Normal Exam macro, and the AI structuring studio.
Live cursor injection
As you speak, the DIRECT drawer displays recognized speech in real time and injects it simultaneously into your target window, with automatic punctuation and continuous tone calibration.
Compact capsule mode
In compact mode (240 × 32 px), the bar tucks at the top of the screen and confirms each normalized medical term, acronym, or abbreviation with a micro-notification.
Calibrated to your voice
Every physician has a distinct speaking pace, accent, and reporting style. Trained on over 10,000 hours of medical dictation and up to 50,000 anonymized patient cases, our AI adjusts its acoustic processing live without ever using your patients’ data for model training.
10,000+ hours of medical dictation
Acoustic and clinical learning across more than 10,000 hours of medical dictation and up to 50,000 anonymized cases to capture natural physician speech.
Predictive AI lexicon (14,000+ terms)
Over 14,000 medical terms, abbreviations, and clinical acronyms dynamically suggested and normalized according to the active specialty.
Accent, tone & auto punctuation
Fine adaptation to the physician’s accent, automatic audio processing calibrated to vocal tone, and intelligent punctuation insertion.
Sovereign locally or in cloud
Deployed on a local server inside your clinic or on a dedicated sovereign cloud, with full encryption and zero reuse of patient data.
Supported specialties
Select your specialty directly on the bar: WannaMed ASR immediately activates the corresponding predictive lexicon, clinical acronyms, and structured report templates.
Cardiology & Vascular
Doppler echocardiography, coronary CT, Holter, and stress testing: precise transcription of ejection fractions (LVEF), valvular gradients, coronary segments, and hemodynamic metrics.
Pulmonology & Thoracic
Chest CT, pulmonary function tests (PFT), and bronchoscopy: mastery of ventilatory acronyms (FEV1, FVC), nodule measurements, and lobar anatomy.
Gastro, Neuro & Surgery
Digestive endoscopy, abdominopelvic ultrasound, brain and spine MRI, operative reports, and emergency medicine: immediate adaptation of specialty abbreviations and templates.
From raw speech to report
Scroll to see how continuous radiological dictation is cleaned of hesitations and organized into standardized clinical headings.
Raw voice dictation (continuous stream)
Lumbar spine MRI performed for right L5 radiculopathy evolving for six weeks in a 48-year-old patient. Sagittal T1 T2 and axial T2 sequences without contrast. Normal spinal canal caliber. At L4-L5 degenerative disc disease with right posterolateral disc protrusion contacting the right L5 nerve root. L5-S1 level unremarkable. Normal conus medullaris signal. In summary right posterolateral L4-L5 disc protrusion impinging on the right L5 nerve root.
Structured report by WannaMed ASR
INDICATION
Right L5 radiculopathy evolving over 6 weeks in a 48-year-old patient.
TECHNIQUE
Lumbar spine MRI acquired in sagittal T1, T2 and axial T2 weightings without gadolinium injection.
FINDINGS
Preserved lumbar canal dimensions. At L4-L5: degenerative disc disease with a right posterolateral disc protrusion contacting the right L5 nerve root. Unremarkable L5-S1 level. Normal morphology and signal of the conus medullaris.
IMPRESSION
Right posterolateral L4-L5 disc protrusion in radicular conflict with the right L5 nerve root.
Clinical contradiction guard
Before inserting text into the patient record, WannaMed ASR automatically cross-checks the findings described in the body against the conclusion. Whether it detects a clinical contradiction between a described lesion and a "normal" conclusion, an anatomical mismatch, or a Right / Left inversion, an amber banner blocks the oversight. On the right, the medical AI chat lets you refine tone, condense the impression, or apply a custom template in one click.
SpeechMike & mobile mic
Keep your familiar dictation ergonomics: USB Philips SpeechMike microphones (LFH3000, LFH3200, LFH3500, SMP3700, and SMP3800 series) are driven directly via USB HID with live visual testing for all 13 physical buttons. Away from your desk, open the built-in QR code to turn your smartphone into a low-latency wireless microphone.
System specifications
- Client OS
- Windows 10, Windows 11, and Mac (macOS Apple Silicon & Intel)
- Cursor injection
- Native keyboard injection (Windows SendInput & macOS Accessibility) + secure clipboard with Citrix, RDP, and VMware Horizon detection
- Training & Lexicon
- 10,000+ hours of medical dictation, up to 50,000 anonymized cases, and predictive AI dictionary of 14,000+ terms, abbreviations, and acronyms
- Adaptive voice profile
- Physician accent adaptation, automatic audio calibration by vocal tone, and automatic punctuation
- Built-in specialties
- Radiology, Cardiology, Pulmonology, Gastroenterology, Neurology, Surgery, Emergency, and General Medicine
- Sovereign deployment
- Local on-premise server or dedicated sovereign cloud, AES-256 encryption, and zero use of patient data for training
Request a live demo
We reply within 24 hours to arrange an on-site trial on your workstations or a live video walkthrough.
Decision-support software for physicians, who remain solely responsible for validating every clinical report.