Trunk Control Test as a Main Predictor of the Modified Barthel Index Score at Discharge From Intensive Post-acute Stroke Rehabilitation: Results From a Multicenter Italian Study

OBJECTIVES

To verify whether trunk control test (TCT) upon admission to intensive inpatient post-stroke rehabilitation, combined with other confounding variables, is independently associated with discharge mBI.

DESIGN

Multicentric retrospective observational cohort study.

SETTING

Two Italian inpatient rehabilitation units.

PARTICIPANTS

A total of 220 post-stroke adult patients, within 30 days from the acute event, were consecutively enrolled.

INTERVENTIONS

Not applicable.

MAIN OUTCOME MEASURE

The outcome measure considered was the modified Barthel Index (mBI), one of the most widely recommended tools for assessing stroke rehabilitation functional outcomes.

RESULTS

All variables collected at admission and significantly associated with mBI at discharge in the univariate analysis (TCT, mBI at admission, pre-stroke modified Rankin Scale [mRS], sex, age, communication ability, time from the event, Cumulative Illness Rating Scale, bladder catheter, and pressure ulcers) entered the multivariate analysis. TCT, mBI at admission, premorbid disability (mRS), communication ability and pressure ulcers (P<.001) independently predicted discharge mBI (adjusted R2=68.5%). Concerning the role of TCT, the model with all covariates and without TCT presented an R2 of 65.1%. On the other side, the model with the TCT only presented an R2 of 53.1%. Finally, with the inclusion of both TCT and all covariates, the model showed an R2 increase up to 68.5%.

CONCLUSIONS

TCT, with other features suggesting functional/clinical complexity, collected upon admission to post-acute intensive inpatient stroke rehabilitation, independently predicted discharge mBI.

Artificial intelligence for the recognition of benign lesions of vocal folds from audio recordings

Objective

The diagnosis of benign lesions of the vocal fold (BLVF) is still challenging. The analysis of the acoustic signals through the implementation of machine learning models can be a viable solution aimed at offering support for clinical diagnosis.

Materials and methods

In this study, a support vector machine was trained and cross-validated (10-fold cross-validation) using 138 features extracted from the acoustic signals of 418 patients with polyps, nodules, oedema, and cysts. The model's performance was presented as accuracy and average F1-score. The results were also analysed in male (M) and female (F) subgroups.

Results

The validation accuracy was 55%, 80%, and 54% on the overall cohort, and in M and F, respectively. Better performances were observed in the detection of cysts and nodules (58% and 62%, respectively) vs polyps and oedema (47% and 53%, respectively). The results on each lesion and the different patterns of the model on M and F are in line with clinical observations, obtaining better results on F and detection of sensitive polyps in M.

Conclusions

This study showed moderately accurate detection of four types of BLVF using acoustic signals. The analysis of the diagnostic results on gender subgroups highlights different behaviours of the diagnostic model.