Transoesophageal Echocardiography (TEE)

What it does

Transoesophageal Echocardiography (TEE) reads a recorded TEE study: the multiplane 2D clips, and 3D mitral datasets where present, acquired from the mid-oesophageal and transgastric positions. TEE is a procedure of its own, with sedation, probe intubation and a 28-view standard, and it is ordered to answer a short list of questions. The run is organised around those questions rather than around a full chamber quantification.

It labels each clip with its standard TEE view, then returns three focused reads: the left atrial appendage (thrombus, sludge or dense spontaneous echo contrast), the mitral valve (scallop map, prolapse or flail, annulus and regurgitation mechanism) and endocarditis (vegetations with size, perivalvular abscess flags).

Intended use

Decision support for the reader of a recorded TEE: the cardiologist clearing a patient for cardioversion or ablation, the heart team describing a mitral valve before repair, and the physician assessing suspected infective endocarditis. Pectinate muscles and sludge can mimic appendage thrombus, so the appendage result is a probability with the frames it rests on.

Who it is for

Echocardiographers and imaging cardiologists, electrophysiologists, structural-heart teams, cardiac surgeons and cardiac anaesthesiologists who perform intra-operative TEE. Developers call the run to add TEE findings to an echo reporting system or a procedure registry.

Accepted input

• 2D multiplane clips: Ultrasound Multi-frame Image objects from the standard mid-oesophageal and transgastric views, including colour Doppler of the mitral and aortic valves.
• 3D datasets: Enhanced US Volume objects of the mitral valve, used by the mitral section.
• Supporting TTE clips: transthoracic clips added to the upload are read with the TEE for the endocarditis question.

Requirements

Requirement	Why
Cine loops with frame timing	Leaflet motion, prolapse and appendage contents are read over the cardiac cycle
Appendage imaged in several multiplane angles	Thrombus is judged across angles, not on one plane
SequenceOfUltrasoundRegions calibration	Vegetation size and mitral annulus dimensions in millimetres
Standard DICOM objects, not screen recordings	Multiplane angle and 3D data stored only in vendor-private formats cannot be read

Optional context

The indication (pre-cardioversion, suspected endocarditis, mitral repair planning), heart rhythm and whether ultrasound contrast was given go in clinical_context. Contrast-enhanced appendage clips are read as such.

The result

One JSON result in sections with findings, measurements and a per-clip view and quality list. Views that the questions need but the study lacks are listed, so a missing appendage sweep is visible rather than silent.

Section	Content
TEE views	One of the 28 standard views per clip (ME 4C, 2C and LAX, TG SAX, bicaval, LAA, AV SAX and LAX and others) with quality
Left atrial appendage thrombus	Probability of thrombus, sludge or dense spontaneous echo contrast, with the supporting frames
Mitral valve anatomy	Scallop map A1–P3, prolapse or flail segments, annulus dimensions and MR mechanism by Carpentier class
Endocarditis	Vegetation detection and size, perivalvular abscess flags, mapped to the imaging criteria
Draft report	Opt-in (options.report): an English narrative that interprets the result for the reader (findings, impression, limitations); every number is checked against the findings

Standards

• DICOM: SR Echocardiography Procedure Report TID 5200; GSPS calipers on the source frames; key frames and MP4 overlays for review; Encapsulated PDF.
• Protocol: ASE/SCA 2013 comprehensive TEE examination (view names and acquisition).
• Endocarditis: ESC 2023 endocarditis guideline, with vegetation and abscess findings mapped to the Duke-ISCVID imaging criteria.
• Mitral valve: Carpentier functional classification for the regurgitation mechanism, with lesions located on the A1–A3 and P1–P3 scallops.

Notes on reading

The appendage result keeps thrombus, sludge and dense spontaneous echo contrast apart, because they lead to different decisions before cardioversion or ablation. Prolapse and flail are reported per scallop, which is how a surgeon plans a repair.

Related models

• us/echo reads the transthoracic study, including LV function and regurgitation grading.
• us/vascular completes the stroke-source work-up with carotid duplex.