ImageRive (Switzerland)

Abstract Feature selection plays a vital role in promoting classifier performance. However, current methods cannot effectively distinguish the complex interaction in the selected features. To further remove these hidden negative interactions, we propose a Genetic Algorithm (GA)-like Dynamic Probability (GADP) method with mutual information method (MI), which has a two-layer structure. The first layer applies MRMR to obtain a primary feature subset. Based on the former candidate features, the GADP, as the second layer, can mine more supportive features. Instead of the commonly used approaches that they frequently focus on improving the operators of GA to enhance the search ability and lowering the convergence time, we improve GA by boldly abandoning the operators and employing the dynamic probability that relies on the performance of each chromosome to determine feature selection in the new generation. The dynamic probability mechanism significantly reduces the parameter number in GA and makes it easy to use. As each gene's probability is independent, the chromosome variety in GADP is more notable than in traditional GA, which ensures GADP has a wider search space and can effectively and accurately select relevant features. To verify the superiority of our method, we evaluate it under multiple conditions on 15 datasets. The results demonstrate the outperformance of the proposed method. Generally, it has the best accuracy. Further, we also compare the proposed model to the state-of-the-art heuristic methods like Harris Hawk Optimization (HHO), Salp Swarm Algorithm (SSA), and White Shark Optimizer (WSO). Our model still owns advantages over them.

BackgroundRepetitive transcranial magnetic stimulation (rTMS) has proven to be an efficient treatment option for patients with treatment-resistant depression (TRD). However, the success rate of this method is still low, and the treatment outcome is unpredictable. The objective of this study was to explore clinical and structural neuroimaging factors as potential biomarkers of the efficacy of high-frequency (HF) rTMS (20 Hz) over the left dorso-lateral pre-frontal cortex (DLPFC).MethodsWe analyzed the records of 131 patients with mood disorders who were treated with rTMS and were assessed at baseline at the end of the stimulation and at 1 month after the end of the treatment. The response is defined as a 50% decrease in the MADRS score between the first and the last assessment. Each of these patients underwent a T1 MRI scan of the brain, which was subsequently segmented with FreeSurfer. Whole-brain analyses [Query, Design, Estimate, Contrast (QDEC)] were conducted and corrected for multiple comparisons. Additionally, the responder status was also analyzed using binomial multivariate regression models. The explored variables were clinical and anatomical features of the rTMS target obtained from T1 MRI: target-scalp distance, DLPFC gray matter thickness, and various cortical measures of interest previously studied.ResultsThe results of a binomial multivariate regression model indicated that depression type (p = 0.025), gender (p = 0.010), and the severity of depression (p = 0.027) were found to be associated with response to rTMS. Additionally, the resistance stage showed a significant trend (p = 0.055). Whole-brain analyses on volume revealed that the average volume of the left part of the superior frontal and the caudal middle frontal regions is associated with the response status. Other MRI-based measures are not significantly associated with response to rTMS in our population.ConclusionIn this study, we investigated the clinical and neuroimaging biomarkers associated with responsiveness to high-frequency rTMS over the left DLPFC in a large sample of patients with TRD. Women, patients with bipolar depressive disorder (BDD), and patients who are less resistant to HF rTMS respond better. Responders present a lower volume of the left part of the superior frontal gyrus and the caudal middle frontal gyrus. These findings support further investigation into the use of clinical variables and structural MRI as possible biomarkers of rTMS treatment response.

Background and objectivesFocal high intensity focused ultrasound (HIFU) is an emerging treatment for selected men with localized prostate cancer. A limitation of HIFU is the absence of a reliable tool to measure treatment effect intraoperatively. Contrast-enhanced ultrasound (CEUS) has been shown to be a promising modality for assessing the extent and boundaries of tissue ablation. The aim of this study was to assess the value of CEUS immediately after focal HIFU. Materials and methodsRetrospective analysis of a prospectively maintained registry including consecutive men undergoing focal HIFU (Focal One). Candidates for focal HIFU were treatment naive men with ≥10 years life expectancy, prostate-specific antigen (PSA) ≤ 20 ng/ml, TNM primary tumor, regional lymph nodes, distant metastasis stage ≤ T2c N0 M0 with a multiparametric MRI (mpMRI) visible lesion concordant with histologically proven prostate cancer. CEUS evaluation was performed immediately at the end of the procedure. Based on the surgeon's estimation of CEUS imaging, re-HIFU was performed, followed by another CEUS evaluation. To test our hypothesis, the results of the CEUS were compared to the results of early mpMRI to rule out clinically significant cancer. The concordance between the 2 tests was measured using the Cohen's kappa. The best model including relevant predictors was calculated with CEUS or with mpMRI to determine their respective added value. ResultsOf 66 men who underwent HIFU, 32 met eligibility criteria. Bifocal treatment was performed in 1 man, increasing the number of treated lesions to 33. Further ablation based on CEUS was delivered intraoperatively to 13 lesions (39%). The positive biopsy rate for clinically significant cancer in the treated zones was 30% (10/33). The negative predictive value of CEUS and early mpMRI was 71% (95% confidence interval: 59%–82%). Concordance between CEUS and mpMRI was significant with a 72.7% agreement (P = 0.001). The model with CEUS showed the best accuracy with an area under the curve of 0.881. ConclusionCEUS has a higher added value compared to early mpMRI in ruling out clinically significant cancer after focal HIFU. It should be evaluated whether the use of CEUS intraoperatively enhances the efficacy of focal HIFU.