Search

(1) Background: This meta-analysis assessed the diagnostic accuracy of deep learning model-based osteoporosis prediction using plain X-ray images. (2) Methods: We searched PubMed, Web of Science, SCOPUS, and Google Scholar from no set beginning date to 28 February 2023, for eligible studies that applied deep learning methods for diagnosing osteoporosis using X-ray images. The quality of studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 criteria. The area under the receiver operating characteristic curve (AUROC) was used to quantify the predictive performance. Subgroup, meta-regression, and sensitivity analyses were performed to identify the potential sources of study heterogeneity. (3) Results: Six studies were included; the pooled AUROC, sensitivity, and specificity were 0.88 (95% confidence interval [CI] 0.85-0.91), 0.81 (95% CI 0.78-0.84), and 0.87 (95% CI 0.81-0.92), respectively, indicating good performance. Moderate heterogeneity was observed. Mega-regression and subgroup analyses were not performed due to the limited number of studies included. (4) Conclusion: Deep learning methods effectively extract bone density information from plain radiographs, highlighting their potential for opportunistic screening. Nevertheless, additional prospective multicenter studies involving diverse patient populations are required to confirm the applicability of this novel technique.

Background: Stroke and spinal cord injury are neurological disorders that cause disability and exert tremendous social and economic effects. Robot-assisted training (RAT), which may reduce spasticity, is widely applied in neurorehabilitation. The combined effects of RAT and antispasticity therapies, such as botulinum toxin A injection therapy, on functional recovery remain unclear. This review evaluated the effects of combined therapy on functional recovery and spasticity reduction., Materials and Methods: Studies evaluating the efficacy of RAT and antispasticity therapy in promoting functional recovery and reducing spasticity were systemically reviewed. Five randomized controlled trials (RCTs) were included. The modified Jadad scale was applied for quality assessment. Functional assessments, such as the Berg Balance Scale, were used to measure the primary outcome. Spasticity assessments, such as the modified Ashworth Scale, were used to measure the secondary outcome., Results: Combined therapy improves functional recovery in the lower limbs but does not reduce spasticity in the upper or lower limbs., Conclusions: The evidence supports that combined therapy improves lower limb function but does not reduce spasticity. The considerable risk of bias among the included studies and the enrolled patients who did not receive interventions within the golden period of intervention are two major factors that should be considered when interpreting these results. Additional high-quality RCTs are required.

Unilateral vocal fold paralysis (UVFP) causes glottal incompetence and poor vocal efficiency. The influence of laryngeal neuromuscular control on aerodynamics in UVFP remains unclear. This study investigated the relationship between laryngeal muscle activities using quantitative laryngeal electromyography (LEMG) and aerodynamics in UVFP. This prospective study recruited patients with UVFP, and the diagnosis was confirmed with videolaryngostroboscopy and LEMG. The patient received aerodynamic assessment and LEMG of the thyroarytenoid-lateral cricoarytenoid (TA-LCA) muscle complex and the cricothyroid (CT) muscle. The relationship between quantitative LEMG and aerodynamic parameters was analyzed. A total of 134 UVFP patients without concurrent CT muscle involvement were enrolled. Compared with the normal side, the peak turn frequency of the lesioned side was lower in the TA-LCA (p < 0.001) and CT (p = 0.048) muscles. Stepwise linear regression revealed that the turn ratio of TA-LCA muscles was a robust factor in the decrease in peak expiratory airflow (β = −0.34, p = 0.036), mean airflow during voicing (β = −0.28, p = 0.014), and aerodynamic power (β = −0.42, p = 0.019), and an increase in aerodynamic efficiency (β = 27.91, p = 0.012). In addition, the turn ratio of CT muscles was a potent factor in inducing an increase in aerodynamic resistance (β = 14.93, p = 0.029). UVFP without CT involvement still showed suppression of CT muscles on the lesioned side, suggesting that neurological impairment of the TA-LCA complex could cause asymmetrical compensation of CT muscles, further impeding aerodynamics. The residual function of TA-LCA muscle complexes facilitates less air leakage and power dissipation, enhancing aerodynamic efficiency. On the other hand, the symmetrical compensation of the CT muscles improves aerodynamic resistance.

Bi-manual therapy (BT), mirror therapy (MT), and robot-assisted rehabilitation have been conducted in hand training in a wide range of stages in stroke patients; however, the mechanisms of action during training remain unclear. In the present study, participants performed hand tasks under different intervention conditions to study bilateral sensorimotor cortical communication, and EEG was recorded. A multifactorial design of the experiment was used with the factors of manipulating objects (O), robot-assisted bimanual training (RT), and MT. The sum of spectral coherence was applied to analyze the C3 and C4 signals to measure the level of bilateral corticocortical communication. We included stroke patients with onset <6 months (n = 6), between 6 months and 1 year (n = 14), and onset >1 year (n = 20), and their Brunnstrom recovery stage ranged from 2 to 4. The results showed that stroke duration might influence the effects of hand rehabilitation in bilateral cortical corticocortical communication with significant main effects under different conditions in the alpha and beta bands. Therefore, stroke duration may influence the effects of hand rehabilitation on interhemispheric coherence.

In this cross-sectional study, we compared voice tone and activities relating to the laryngeal muscle between unilateral vocal fold paralysis (UVFP) patients with and without cricothyroid (CT) muscle dysfunction to define how CT dysfunction affects language tone. Eighty-eight female surgery-related UVFP patients were recruited and received acoustic voice analysis and laryngeal electromyography (LEMG) when the patient was producing the four Mandarin tones. The statistical analysis was compared between UVFP patients with (CT+ group, 17 patients) and without CT muscle (CT- group, 71 patients) involvement. When producing Mandarin Tone 2, the voice tone in the CT+ group had smaller rise range ( p = 0.007), lower rise rate ( p = 0.002), and lower fundamental frequency (F0) at the offset point of the voice ( p = 0.023). When producing Mandarin Tone 4, the voice tone in the CT+ group had smaller drop range ( p = 0.019), lower drop rate ( p = 0.005), and lower F0 at voice onset ( p = 0.025). The CT+ group had significantly lower CT muscle activity when producing the four Mandarin tones. In conclusion, CT dysfunction causes a limitation of high-rising tone in Tone 2 and high-falling tone in Tone 4, a property that dramatically limits the tonal characteristics in Mandarin, a tonal language. This limitation could further impair the patient's communication ability.

Pain assessment is essential for preclinical and clinical studies on pain. The mouse grimace scale (MGS), consisting of five grimace action units, is a reliable measurement of spontaneous pain in mice. However, MGS scoring is labor-intensive and time-consuming. Deep learning can be applied for the automatic assessment of spontaneous pain. We developed a deep learning model, the DeepMGS, that automatically crops mouse face images, predicts action unit scores and total scores on the MGS, and finally infers whether pain exists. We then compared the performance of DeepMGS with that of experienced and apprentice human scorers. The DeepMGS achieved an accuracy of 70-90% in identifying the five action units of the MGS, and its performance (correlation coefficient = 0.83) highly correlated with that of an experienced human scorer in total MGS scores. In classifying pain and no pain conditions, the DeepMGS is comparable to the experienced human scorer and superior to the apprentice human scorers. Heatmaps generated by gradient-weighted class activation mapping indicate that the DeepMGS accurately focuses on MGS-relevant areas in mouse face images. These findings support that the DeepMGS can be applied for quantifying spontaneous pain in mice, implying its potential application for predicting other painful conditions from facial images.

Prediction of post-stroke functional outcomes is crucial for allocating medical resources. In this study, a total of 577 patients were enrolled in the Post-Acute Care-Cerebrovascular Disease (PAC-CVD) program, and 77 predictors were collected at admission. The outcome was whether a patient could achieve a Barthel Index (BI) score of >60 upon discharge. Eight machine-learning (ML) methods were applied, and their results were integrated by stacking method. The area under the curve (AUC) of the eight ML models ranged from 0.83 to 0.887, with random forest, stacking, logistic regression, and support vector machine demonstrating superior performance. The feature importance analysis indicated that the initial Berg Balance Test (BBS-I), initial BI (BI-I), and initial Concise Chinese Aphasia Test (CCAT-I) were the top three predictors of BI scores at discharge. The partial dependence plot (PDP) and individual conditional expectation (ICE) plot indicated that the predictors' ability to predict outcomes was the most pronounced within a specific value range (e.g., BBS-I < 40 and BI-I < 60). BI at discharge could be predicted by information collected at admission with the aid of various ML models, and the PDP and ICE plots indicated that the predictors could predict outcomes at a certain value range.

(1) Background: severe weight loss was reported to be related to unilateral vocal fold paralysis (UVFP) after esophagectomy and could thus impair survival. Concomitant radical lymph node dissection along the recurrent laryngeal nerve during esophageal cancer surgery is controversial, as it might induce UVFP. Early intervention for esophagectomy-related UVFP by administering intracordal injections of temporal agents has recently become popular. This study investigated the survival outcomes of esophagectomy for esophageal squamous cell carcinoma (ESCC) after the introduction of early injection laryngoplasty (EIL). (2) Methods: a retrospective review of patients with ESCC after curative-intent esophagectomy was conducted in a tertiary referral medical center. The necessity of EIL with hyaluronic acid was comprehensively discussed for all symptomatic UVFP patients. The survival outcomes and related risk factors of ESCC were evaluated. (3) Results: among the cohort of 358 consecutive patients who underwent esophagectomy for ESCC, 42 (11.7%) showed postsurgical UVFP. Twenty-nine of them received office-based EIL. After EIL, the glottal gap area, maximum phonation time and voice outcome survey showed significant improvement at one, three and six months measurements. The number of lymph nodes in the resected specimen was higher in those with UVFP than in those without UVFP (30.1 ± 15.7 vs. 24.6 ± 12.7, p = 0.011). The Kaplan-Meier overall survival was significantly better in patients who had UVFP ( p = 0.014), received neck anastomosis ( p = 0.004), underwent endoscopic resection ( p < 0.001) and had early-stage cancer ( p < 0.001). Multivariate Cox logistic regression analysis showed two independent predictors of OS, showing that the primary stage and anastomosis type are the two independent predictors of OS. (4) Conclusion: EIL is effective in improving UVFP-related symptoms, thus providing compensatory and palliative measures to ensure the patient's postsurgical quality of life. The emerging use of EIL might encourage cancer surgeons to radically dissect lymph nodes along the recurrent laryngeal nerve, thus changing the survival trend.