Research Technologies

Magnetic Resonance Imaging (MRI)

MRI image

Vocal tract images are made in multiple planes during sustained speech sounds. Tagged Cine-MRI (tMRI) allows tracking of tissue points in oral structures.

Ultrasound


Head and Transducer Support (HATS) system

Ultra high-frequency sound waves are used to create movies of planar tongue motion. 

Electropalatography (EPG)


Screenshot of computer animation of electropalatography scan

Acrylic palates, custom made to fit each subject's palate, record tongue-palate contact in real time.

Research Areas

Modeling the Tongue

3D Finite Element Model of the tongue designed for Restore Medical Inc, Minneapolis, in conjuction with Reiner Wilhelms-Tricarico, Paul Buscemi, Mark Carlson, for the purpose of modeling sleep apnea. The model contains a jaw, hyoid and tongue with 3 extrinsic and 4 intrinsic tongue muscles (not palatoglossus); and three jaw/hyoid muscles (mylohyoid, geniohyoid, digastric).

 Modeling the interdigitated tongue muscles tests the many muscle combinations used to produce speech. 

Tongue Segmentation

Screenshot from a movie simulating uncorrolated segmental movement of the tongue

The tongue can be segmented into regions that produce a speech gesture using coordinated muscle activation strategies. 

Swallowing

3d diagram of tongue surface

Tongue surface deformation creates a transverse waveform that propels the bolus from the oral cavity (front) into the pharynx (back).

Glossectomy Study

Patient with mouth open, displaying a tumor on the tongue

We study speech adaptation to glossectomy surgery, which removes a malignant tumor from the tongue. A ‘flap’ and may be used to replace the tissue.

Geometry to Acoustics

Profile view of MRI demonstrating speech

Vocal tract airway sections are calculated from reconstructed 3D super-resolution volumes, and used to generate speech spectra.