I am very proud to announce our first paper that was entire­ly planned, con­duct­ed, analysed and writ­ten up since our group has been in exis­tence. Julia joined me as the first PhD stu­dent in Decem­ber 2010, and has since been busy doing awe­some work. Check out her first paper!

Audi­to­ry skills and brain mor­phol­o­gy pre­dict indi­vid­ual dif­fer­ences in adap­ta­tion to degrad­ed speech

Noise-vocod­ed speech is a spec­tral­ly high­ly degrad­ed sig­nal, but it pre­serves the tem­po­ral enve­lope of speech. Lis­ten­ers vary con­sid­er­ably in their abil­i­ty to adapt to this degrad­ed speech sig­nal. Here, we hypoth­e­sized that indi­vid­ual dif­fer­ences in adap­ta­tion to vocod­ed speech should be pre­dictable by non-speech audi­to­ry, cog­ni­tive, and neu­roanatom­i­cal fac­tors. We test­ed eigh­teen nor­mal-hear­ing par­tic­i­pants in a short-term vocod­ed speech-learn­ing par­a­digm (lis­ten­ing to 100 4- band-vocod­ed sen­tences). Non-speech audi­to­ry skills were assessed using ampli­tude mod­u­la­tion (AM) rate dis­crim­i­na­tion, where mod­u­la­tion rates were cen­tered on the speech-rel­e­vant rate of 4 Hz. Work­ing mem­o­ry capac­i­ties were eval­u­at­ed, and struc­tur­al MRI scans were exam­ined for anatom­i­cal pre­dic­tors of vocod­ed speech learn­ing using vox­el-based mor­phom­e­try. Lis­ten­ers who learned faster to under­stand degrad­ed speech showed small­er thresh­olds in the AM dis­crim­i­na­tion task. Anatom­i­cal brain scans revealed that faster learn­ers had increased vol­ume in the left thal­a­mus (pul­v­inar). These results sug­gest that adap­ta­tion to vocod­ed speech ben­e­fits from indi­vid­ual AM dis­crim­i­na­tion skills. This abil­i­ty to adjust to degrad­ed speech is fur­ther­more reflect­ed anatom­i­cal­ly in an increased vol­ume in an area of the thal­a­mus, which is strong­ly con­nect­ed to the audi­to­ry and pre­frontal cor­tex. Thus, indi­vid­ual audi­to­ry skills that are not speech-spe­cif­ic and left thal­a­mus gray mat­ter vol­ume can pre­dict how quick­ly a lis­ten­er adapts to degrad­ed speech. Please be in touch with Julia Erb if you are inter­est­ed in a preprint as soon as we get hold of the final, type­set manuscript.

[Update#1]: Julia has also pub­lished a blog post on her work.

[Update#2] Paper is avail­able here.