Multimodal imaging and electrophysiological study in the differential diagnosis of rest tremor.

Introduction: Distinguishing tremor-dominant Parkinson's disease (tPD) from essential tremor with rest tremor (rET) can be challenging and often requires dopamine imaging. This study aimed to differentiate between these two diseases through a machine learning (ML) approach based on rest tremor (RT) electrophysiological features and structural MRI data. Methods: We enrolled 72 patients including 40 tPD patients and 32 rET patients, and 45 control subjects (HC). RT electrophysiological features (frequency, amplitude, and phase) were calculated using surface electromyography (sEMG). Several MRI morphometric variables (cortical thickness, surface area, cortical/subcortical volumes, roughness, and mean curvature) were extracted using Freesurfer. ML models based on a tree-based classification algorithm termed XGBoost using MRI and/or electrophysiological data were tested in distinguishing tPD from rET patients. Results: Both structural MRI and sEMG data showed acceptable performance in distinguishing the two patient groups. Models based on electrophysiological data performed slightly better than those based on MRI data only (mean AUC: 0.92 and 0.87, respectively; p = 0.0071). The top-performing model used a combination of sEMG features (amplitude and phase) and MRI data (cortical volumes, surface area, and mean curvature), reaching AUC: 0.97 ± 0.03 and outperforming models using separately either MRI (p = 0.0001) or EMG data (p = 0.0231). In the best model, the most important feature was the RT phase. Conclusion: Machine learning models combining electrophysiological and MRI data showed great potential in distinguishing between tPD and rET patients and may serve as biomarkers to support clinicians in the differential diagnosis of rest tremor syndromes in the absence of expensive and invasive diagnostic procedures such as dopamine imaging.

Machine learning pipeline to analyze clinical and proteomics data: experiences on a prostate cancer case.

Proteomic-based analysis is used to identify biomarkers in blood samples and tissues. Data produced by devices such as mass spectrometry requires platforms to identify and quantify proteins (or peptides). Clinical information can be related to mass spectrometry data to identify diseases at an early stage. Machine learning techniques can be used to support physicians and biologists in studying and classifying pathologies. We present the application of machine learning techniques to define a pipeline aimed at studying and classifying proteomics data enriched using clinical information. The pipeline allows users to relate established blood biomarkers with clinical parameters and proteomics data. The proposed pipeline entails three main phases: (i) feature selection, (ii) models training, and (iii) models ensembling. We report the experience of applying such a pipeline to prostate-related diseases. Models have been trained on several biological datasets. We report experimental results about two datasets that result from the integration of clinical and mass spectrometry-based data in the contexts of serum and urine analysis. The pipeline receives input data from blood analytes, tissue samples, proteomic analysis, and urine biomarkers. It then trains different models for feature selection, classification and voting. The presented pipeline has been applied on two datasets obtained in a 2 years research project which aimed to extract hidden information from mass spectrometry, serum, and urine samples from hundreds of patients. We report results on analyzing prostate datasets serum with 143 samples, including 79 PCa and 84 BPH patients, and an urine dataset with 121 samples, including 67 PCa and 54 BPH patients. As results pipeline allowed to identify interesting peptides in the two datasets, 6 for the first one and 2 for the second one. The best model for both serum (AUC=0.87, Accuracy=0.83, F1=0.81, Sensitivity=0.84, Specificity=0.81) and urine (AUC=0.88, Accuracy=0.83, F1=0.83, Sensitivity=0.85, Specificity=0.80) datasets showed good predictive performances. We made the pipeline code available on GitHub and we are confident that it will be successfully adopted in similar clinical setups.

Sex Differences in Conversion Risk from Mild Cognitive Impairment to Alzheimer's Disease: An Explainable Machine Learning Study with Random Survival Forests and SHAP.

Alzheimer's disease (AD) exhibits sex-linked variations, with women having a higher prevalence, and little is known about the sexual dimorphism in progressing from Mild Cognitive Impairment (MCI) to AD. The main aim of our study was to shed light on the sex-specific conversion-to-AD risk factors using Random Survival Forests (RSF), a Machine Learning survival approach, and Shapley Additive Explanations (SHAP) on dementia biomarkers in stable (sMCI) and progressive (pMCI) patients. With this purpose, we built two separate models for male (M-RSF) and female (F-RSF) cohorts to assess whether global explanations differ between the sexes. Similarly, SHAP local explanations were obtained to investigate changes across sexes in feature contributions to individual risk predictions. The M-RSF achieved higher performance on the test set (0.87) than the F-RSF (0.79), and global explanations of male and female models had limited similarity (<71.1%). Common influential variables across the sexes included brain glucose metabolism and CSF biomarkers. Conversely, the M-RSF had a notable contribution from hippocampus, which had a lower impact on the F-RSF, while verbal memory and executive function were key contributors only in F-RSF. Our findings confirmed that females had a higher risk of progressing to dementia; moreover, we highlighted distinct sex-driven patterns of variable importance, uncovering different feature contribution risks across sexes that decrease/increase the conversion-to-AD risk.

Development of a predictive model to distinguish prostate cancer from benign prostatic hyperplasia by integrating serum glycoproteomics and clinical variables.

BACKGROUND: Prostate Cancer (PCa) represents the second leading cause of cancer-related death in men. Prostate-specific antigen (PSA) serum testing, currently used for PCa screening, lacks the necessary sensitivity and specificity. New non-invasive diagnostic tools able to discriminate tumoral from benign conditions and aggressive (AG-PCa) from indolent forms of PCa (NAG-PCa) are required to avoid unnecessary biopsies. METHODS: In this work, 32 formerly N-glycosylated peptides were quantified by PRM (parallel reaction monitoring) in 163 serum samples (79 from PCa patients and 84 from individuals affected by benign prostatic hyperplasia (BPH)) in two technical replicates. These potential biomarker candidates were prioritized through a multi-stage biomarker discovery pipeline articulated in: discovery, LC-PRM assay development and verification phases. Because of the well-established involvement of glycoproteins in cancer development and progression, the proteomic analysis was focused on glycoproteins enriched by TiO2 (titanium dioxide) strategy. RESULTS: Machine learning algorithms have been applied to the combined matrix comprising proteomic and clinical variables, resulting in a predictive model based on six proteomic variables (RNASE1, LAMP2, LUM, MASP1, NCAM1, GPLD1) and five clinical variables (prostate dimension, proPSA, free-PSA, total-PSA, free/total-PSA) able to distinguish PCa from BPH with an area under the Receiver Operating Characteristic (ROC) curve of 0.93. This model outperformed PSA alone which, on the same sample set, was able to discriminate PCa from BPH with an AUC of 0.79. To improve the clinical managing of PCa patients, an explorative small-scale analysis (79 samples) aimed at distinguishing AG-PCa from NAG-PCa was conducted. A predictor of PCa aggressiveness based on the combination of 7 proteomic variables (FCN3, LGALS3BP, AZU1, C6, LAMB1, CHL1, POSTN) and proPSA was developed (AUC of 0.69). CONCLUSIONS: To address the impelling need of more sensitive and specific serum diagnostic tests, a predictive model combining proteomic and clinical variables was developed. A preliminary evaluation to build a new tool able to discriminate aggressive presentations of PCa from tumors with benign behavior was exploited. This predictor displayed moderate performances, but no conclusions can be drawn due to the limited number of the sample cohort. Data are available via ProteomeXchange with identifier PXD035935.

Explainability of random survival forests in predicting conversion risk from mild cognitive impairment to Alzheimer's disease.

Random Survival Forests (RSF) has recently showed better performance than statistical survival methods as Cox proportional hazard (CPH) in predicting conversion risk from mild cognitive impairment (MCI) to Alzheimer's disease (AD). However, RSF application in real-world clinical setting is still limited due to its black-box nature.For this reason, we aimed at providing a comprehensive study of RSF explainability with SHapley Additive exPlanations (SHAP) on biomarkers of stable and progressive patients (sMCI and pMCI) from Alzheimer's Disease Neuroimaging Initiative. We evaluated three global explanations-RSF feature importance, permutation importance and SHAP importance-and we quantitatively compared them with Rank-Biased Overlap (RBO). Moreover, we assessed whether multicollinearity among variables may perturb SHAP outcome. Lastly, we stratified pMCI test patients in high, medium and low risk grade, to investigate individual SHAP explanation of one pMCI patient per risk group.We confirmed that RSF had higher accuracy (0.890) than CPH (0.819), and its stability and robustness was demonstrated by high overlap (RBO > 90%) between feature rankings within first eight features. SHAP local explanations with and without correlated variables had no substantial difference, showing that multicollinearity did not alter the model. FDG, ABETA42 and HCI were the first important features in global explanations, with the highest contribution also in local explanation. FAQ, mPACCdigit, mPACCtrailsB and RAVLT immediate had the highest influence among all clinical and neuropsychological assessments in increasing progression risk, as particularly evident in pMCI patients' individual explanation. In conclusion, our findings suggest that RSF represents a useful tool to support clinicians in estimating conversion-to-AD risk and that SHAP explainer boosts its clinical utility with intelligible and interpretable individual outcomes that highlights key features associated with AD prognosis.

Development of a New Wearable Device for the Characterization of Hand Tremor.

Rest tremor (RT) is observed in subjects with Parkinson's disease (PD) and Essential Tremor (ET). Electromyography (EMG) studies have shown that PD subjects exhibit alternating contractions of antagonistic muscles involved in tremors, while the contraction pattern of antagonistic muscles is synchronous in ET subjects. Therefore, the RT pattern can be used as a potential biomarker for differentiating PD from ET subjects. In this study, we developed a new wearable device and method for differentiating alternating from a synchronous RT pattern using inertial data. The novelty of our approach relies on the fact that the evaluation of synchronous or alternating tremor patterns using inertial sensors has never been described so far, and current approaches to evaluate the tremor patterns are based on surface EMG, which may be difficult to carry out for non-specialized operators. This new device, named "RT-Ring", is based on a six-axis inertial measurement unit and a Bluetooth Low-Energy microprocessor, and can be worn on a finger of the tremulous hand. A mobile app guides the operator through the whole acquisition process of inertial data from the hand with RT, and the prediction of tremor patterns is performed on a remote server through machine learning (ML) models. We used two decision tree-based algorithms, XGBoost and Random Forest, which were trained on features extracted from inertial data and achieved a classification accuracy of 92% and 89%, respectively, in differentiating alternating from synchronous tremor segments in the validation set. Finally, the classification response (alternating or synchronous RT pattern) is shown to the operator on the mobile app within a few seconds. This study is the first to demonstrate that different electromyographic tremor patterns have their counterparts in terms of rhythmic movement features, thus making inertial data suitable for predicting the muscular contraction pattern of tremors.

Neuroimaging correlates of postural instability in Progressive Supranuclear Palsy.

OBJECTIVE: We aimed to identify the brain structures associated with postural instability (PI) in Progressive Supranuclear Palsy (PSP). METHODS: Forty-seven PSP patients and 45 control subjects were enrolled in this study. PI was assessed using the items 27 and 28 of the PSP rating scale (postural instability score, PIS). PSP patients were compared with controls using voxel-based morphometry (VBM). In PSP patients, LASSO regression model was used to investigate associations between VBM-based Region-Of-Interest grey matter (GM) volumes and different categories of the PSP rating scale. A whole-brain multi-regression analysis was also used to identify brain areas where GM volumes correlated with the PIS in PSP patients. RESULTS: VBM analysis showed widespread GM atrophy (fronto-temporal-parietal-occipital regions, limbic lobes, insula, cerebellum, and basal ganglia) in PSP patients compared with control subjects. In PSP patients, LASSO regression analysis showed associations of the right cerebellar lobules IV-V with ocular motor category score, and the left Rolandic area with bulbar category score, while the right inferior frontal gyrus (IFG) was negatively correlated with the PIS. The whole-brain multi-regression analysis identified the right IFG as the only area significantly associated with the PIS. CONCLUSIONS: In our study, two different approaches demonstrated that the IFG volume was associated with PIS in PSP patients, suggesting that this area may play a role in the pathophysiological mechanisms underlying PI. Our findings may have important implications for developing optimal Transcranial Magnetic Stimulation protocols targeting IFG in parkinsonism with postural disorders.

Differentiating between common PSP phenotypes using structural MRI: a machine learning study.

BACKGROUND: Differentiating Progressive supranuclear palsy-Richardson's syndrome (PSP-RS) from PSP-Parkinsonism (PSP-P) may be extremely challenging. In this study, we aimed to distinguish these two PSP phenotypes using MRI structural data. METHODS: Sixty-two PSP-RS, 40 PSP-P patients and 33 control subjects were enrolled. All patients underwent brain 3 T-MRI; cortical thickness and cortical/subcortical volumes were extracted using Freesurfer on T1-weighted images. We calculated the automated MR Parkinsonism Index (MRPI) and its second version including also the third ventricle width (MRPI 2.0) and tested their classification performance. We also employed a Machine learning (ML) classification approach using two decision tree-based algorithms (eXtreme Gradient Boosting [XGBoost] and Random Forest) with different combinations of structural MRI data in differentiating between PSP phenotypes. RESULTS: MRPI and MRPI 2.0 had AUC of 0.88 and 0.81, respectively, in differentiating PSP-RS from PSP-P. ML models demonstrated that the combination of MRPI and volumetric/thickness data was more powerful than each feature alone. The two ML algorithms showed comparable results, and the best ML model in differentiating between PSP phenotypes used XGBoost with a combination of MRPI, cortical thickness and subcortical volumes (AUC 0.93 ± 0.04). Similar performance (AUC 0.93 ± 0.06) was also obtained in a sub-cohort of 59 early PSP patients. CONCLUSION: The combined use of MRPI and volumetric/thickness data was more accurate than each MRI feature alone in differentiating between PSP-RS and PSP-P. Our study supports the use of structural MRI to improve the early differential diagnosis between common PSP phenotypes, which may be relevant for prognostic implications and patient inclusion in clinical trials.

Cortical involvement in essential tremor with and without rest tremor: a machine learning study.

INTRODUCTION: There is some debate on the relationship between essential tremor with rest tremor (rET) and the classic ET syndrome, and only few MRI studies compared ET and rET patients. This study aimed to explore structural cortical differences between ET and rET, to improve the knowledge of these tremor syndromes. METHODS: Thirty-three ET patients, 30 rET patients and 45 control subjects (HC) were enrolled. Several MR morphometric variables (thickness, surface area, volume, roughness, mean curvature) of brain cortical regions were extracted using Freesurfer on T1-weighted images and compared among groups. The performance of a machine learning approach (XGBoost) using the extracted morphometric features was tested in discriminating between ET and rET patients. RESULTS: rET patients showed increased roughness and mean curvature in some fronto-temporal areas compared with HC and ET, and these metrics significantly correlated with cognitive scores. Cortical volume in the left pars opercularis was also lower in rET than in ET patients. No differences were found between ET and HC. XGBoost discriminated between rET and ET with mean AUC of 0.86 ± 0.11 in cross-validation analysis, using a model based on cortical volume. Cortical volume in the left pars opercularis was the most informative feature for classification between the two ET groups. CONCLUSION: Our study demonstrated higher cortical involvement in fronto-temporal areas in rET than in ET patients, which may be linked to the cognitive status. A machine learning approach based on MR volumetric data demonstrated that these two ET subtypes can be distinguished using structural cortical features.

Data-Independent Acquisition Mass Spectrometry of EPS-Urine Coupled to Machine Learning: A Predictive Model for Prostate Cancer.

Prostate cancer (PCa) is annually the most frequently diagnosed cancer in the male population. To date, the diagnostic path for PCa detection includes the dosage of serum prostate-specific antigen (PSA) and the digital rectal exam (DRE). However, PSA-based screening has insufficient specificity and sensitivity; besides, it cannot discriminate between the aggressive and indolent types of PCa. For this reason, the improvement of new clinical approaches and the discovery of new biomarkers are necessary. In this work, expressed prostatic secretion (EPS)-urine samples from PCa patients and benign prostatic hyperplasia (BPH) patients were analyzed with the aim of detecting differentially expressed proteins between the two analyzed groups. To map the urinary proteome, EPS-urine samples were analyzed by data-independent acquisition (DIA), a high-sensitivity method particularly suitable for detecting proteins at low abundance. Overall, in our analysis, 2615 proteins were identified in 133 EPS-urine specimens obtaining the highest proteomic coverage for this type of sample; of these 2615 proteins, 1670 were consistently identified across the entire data set. The matrix containing the quantified proteins in each patient was integrated with clinical parameters such as the PSA level and gland size, and the complete matrix was analyzed by machine learning algorithms (by exploiting 90% of samples for training/testing using a 10-fold cross-validation approach, and 10% of samples for validation). The best predictive model was based on the following components: semaphorin-7A (sema7A), secreted protein acidic and rich in cysteine (SPARC), FT ratio, and prostate gland size. The classifier could predict disease conditions (BPH, PCa) correctly in 83% of samples in the validation set. Data are available via ProteomeXchange with the identifier PXD035942.

Cortical atrophy distinguishes idiopathic normal-pressure hydrocephalus from progressive supranuclear palsy: A machine learning approach.

INTRODUCTION: Progressive supranuclear palsy (PSP) and idiopathic normal pressure hydrocephalus (iNPH) share several clinical and radiological features, making the differential diagnosis challenging. In this study, we aimed to differentiate between these two diseases using a machine learning approach based on cortical thickness and volumetric data. METHODS: Twenty-three iNPH patients, 50 PSP patients and 55 control subjects were enrolled. All participants underwent a brain 3T-MRI, and cortical thickness and volumes were extracted using Freesurfer 6 on T1-weighted images and compared among groups. Finally, the performance of a machine learning approach with random forest using the extracted cortical features was investigated to differentiate between iNPH and PSP patients. RESULTS: iNPH patients showed cortical thinning and volume loss in the frontal lobe, temporal lobe and cingulate cortex, and thickening in the superior parietal gyrus in comparison with controls and PSP patients. PSP patients only showed mild thickness and volume reduction in the frontal lobe, compared to control subjects. Random Forest algorithm distinguished iNPH patients from controls with AUC of 0.96 and from PSP patients with AUC of 0.95, while a lower performance (AUC 0.76) was reached in distinguishing PSP from controls. CONCLUSION: This study demonstrated a more severe and widespread cortical involvement in iNPH than in PSP, possibly due to the marked lateral ventricular enlargement which characterizes iNPH. A machine learning model using thickness and volumetric data led to accurate differentiation between iNPH and PSP patients, which may help clinicians in the differential diagnosis and in the selection of patients for shunt procedures.

Magnetic Resonance Planimetry in the Differential Diagnosis between Parkinson's Disease and Progressive Supranuclear Palsy.

The clinical differential diagnosis between Parkinson's disease (PD) and progressive supranuclear palsy (PSP) is often challenging. The description of milder PSP phenotypes strongly resembling PD, such as PSP-Parkinsonism, further increased the diagnostic challenge and the need for reliable neuroimaging biomarkers to enhance the diagnostic certainty. This review aims to summarize the contribution of a relatively simple and widely available imaging technique such as MR planimetry in the differential diagnosis between PD and PSP, focusing on the recent advancements in this field. The development of accurate MR planimetric biomarkers, together with the implementation of automated algorithms, led to robust and objective measures for the differential diagnosis of PSP and PD at the individual level. Evidence from longitudinal studies also suggests a role of MR planimetry in predicting the development of the PSP clinical signs, allowing to identify PSP patients before they meet diagnostic criteria when their clinical phenotype can be indistinguishable from PD. Finally, promising evidence exists on the possible association between MR planimetric measures and the underlying pathology, with important implications for trials with new disease-modifying target therapies.