Browsing by Subject "Signal Processing, Computer-Assisted"
Now showing 1 - 2 of 2
- Results Per Page
- Sort Options
Item Advanced MR Imaging of Bone Marrow: Quantification of Signal Alterations on T1-Weighted Dixon and T2-Weighted Dixon Sequences in Red Marrow, Yellow Marrow and Pathologic Marrow Lesions(2019-03-27) Sasiponganan, Chayanit; Chhabra, Avneesh; Pezeshk, Parham; Ashikyan, OganesBACKGROUND: Characterization of bone marrow through qualitative changes in signal intensity (SI) have not been evaluated on T2-weighted (W) Dixon imaging. OBJECTIVE: We hypothesized that T2W Dixon imaging SI changes are similar to those on T1W Dixon imaging and can be used to effectively characterize bone marrow. METHODS: 144 patients (77 controls with yellow and red marrow regions, 64 lesions--33 benign and 31 malignant) between January 2016 - December 2017 were retrospectively identified. For the control group, fixed 2 cm2 region of interests (ROI) were drawn at L5, bilateral ilium and femurs on in-phase and opposed-phase T1W and T2W Dixon images. For the lesion group, ROI of best fit were drawn around each lesion on in-phase and opposed-phase T2W Dixon images. SI changes for each group were compared. Inter-reader analysis was performed. RESULTS: Yellow marrow exhibited smaller SI changes as compared to red marrow on both T1W and T2W Dixon at all locations (p < 0.0001) except at L5 on T2W Dixon (p = 0.206). Both benign and malignant lesions showed significantly smaller SI changes as compared to both yellow (p = 0.0087, p <0.0001) and red marrow (p = 0.0004, p < 0.0001) on T2W Dixon. Malignant lesions exhibited smaller SI change as compared to benign lesions on T2W Dixon (p = 0.0005). Signal intensity loss on both red and yellow marrow were smaller on T1W Dixon as compared to T2W Dixon (0.49-0.64, 0.27-0.31 vs 0.70-0.74, 0.48-0.71). Inter-reader agreements were excellent (0.91-0.97). CONCLUSION: SI change calculated from T2W Dixon imaging can adequately differentiate between yellow marrow, red marrow and osseous lesions--both benign and malignant.Item Quantifying the Perceptual Quality of Strain: An Electroglottographic Analysis of Continuous Dysphonic Speech(2015-04-03) Somanath, Keerthan; Mau, TedINTRODUCTION: Excessive vocal effort perceived as strain may involve increased vocal fold compression. Electroglottography (EGG) in principle could provide a measure of strain directly connected to oscillatory kinematics at the sound source. The utility of EGG has been limited by considerable inter- and intra-subject variability. OBJECTIVE: This work aims to (1) develop an algorithm to analyze EGG signal in continuous, dysphonic speech and (2) identify parameters that correlate with strain. METHODS: EGG signal from 8 normal speakers and 8 subjects with adductor spasmodic dysphonia (ADSD) reading two-sentence excerpts from the Rainbow Passage was processed by the new software developed in MATLAB. The contact quotient (CQ), pulse width at the 50% amplitude level (EGGW50), and various closing slope and opening slope measures were extracted from selected speech segments. Intra-subject and inter-subject comparisons were then made. RESULTS: None of the EGG parameters differed between normal and ADSD speakers. Within- subject comparison among ADSD speakers showed that the opening slope measure SO7525 distinguished between the strained and unstrained syllables. CONCLUSION: These results provide further insight into the utility and limitations of EGG. While EGG may have limited utility in inter-subject comparison, it may provide a useful objective measure of vocal strain in the same subject with variable degrees of strain or over time.