Non-invasive Prediction Research Articles

The Performance of Machine Learning for Prediction of H3K27 M Mutation in Midline Gliomas: A Systematic Review and Meta-Analysis

Diffuse midline gliomas (DMGs) encompass a set of tumors, and those tumors with H3K27M mutation carry a poor prognosis. In recent years, machine learning (ML)-based radiomics have shown promising results in predicting gene mutation status non-invasively. Therefore, this study aims to comprehensively evaluate the diagnostic performance of ML-based magnetic resonance imaging radiomics in predicting H3K27M mutation status in DMG patients. A systematic search was conducted using relevant keywords in PubMed/Medline, Scopus, Embase, and Web of Science from inception to May 2023. Original studies evaluating the diagnostic performance of ML models in predicting H3K27M mutation status in DMGs were enrolled. Quality assessment of the enrolled studies was conducted using QUADAS-2. Data were analyzed using STATA version 17.0 to calculate pooled sensitivity, specificity, positive and negative likelihood ratio, diagnostic score, and diagnostic odds ratio. A total of 13 studies, including 12 retrospectives and 1 both retrospective and prospective study, enrolled 1510 (male= 777) DMG patients. Six studies underwent meta-analysis which showed a pooled sensitivity, specificity, positive and negative likelihood ratio, diagnostic score, and diagnostic odds ratio of 0.91 (95% CI 0.77-0.97), 0.81 (95% CI 0.73-0.88), 4.86 (95% CI 3.25-7.24), 0.11 (95% CI 0.04-0.29), 3.75 (95% CI 2.62-4.88), and 42.61 (95% CI 13.77-131.87), respectively. Non-invasive prediction of H3K27M mutation status in patients with DMGs using magnetic resonance imaging radiomics is a promising tool with good diagnostic performance. However, the pooled metrics had a wide confidence interval, which required further studies to enhance ML algorithms' accuracy and facilitate their integration into daily clinical practice.

Dual-Level Augmentation Radiomics Analysis for Multisequence MRI Meningioma Grading.

Preoperative, noninvasive prediction of meningioma grade is important for therapeutic planning and decision making. In this study, we propose a dual-level augmentation strategy incorporating image-level augmentation (IA) and feature-level augmentation (FA) to tackle class imbalance and improve the predictive performance of radiomics for meningioma grading on Magnetic Resonance Imaging (MRI). This study recruited 160 consecutive patients with pathologically proven meningioma (129 low-grade (WHO grade I) tumors; 31 high-grade (WHO grade II and III) tumors) with preoperative multisequence MRI imaging. A dual-level augmentation strategy combining IA and FA was applied and evaluated in 100 repetitions in 3-, 5-, and 10-fold cross-validation. The best area under the receiver operating characteristics curve of our method in 100 repetitions was ≥0.78 in all cross-validations. The corresponding cross-validation sensitivities (cross-validation specificity) were 0.72 (0.69), 0.76 (0.71), and 0.63 (0.82) in 3-, 5-, and 10-fold cross-validation, respectively. The proposed method achieved significantly better performance and distribution of results, outperforming single-level augmentation (IA or FA) or no augmentation in each cross-validation. The dual-level augmentation strategy using IA and FA significantly improves the performance of the radiomics model for meningioma grading on MRI, allowing better radiomics-based preoperative stratification and individualized treatment.

Development and Validation of a Machine Learning–Based Model Using CT Radiomics for Predicting Immune Checkpoint Inhibitor-related Pneumonitis in Patients With NSCLC Receiving Anti-PD1 Immunotherapy: A Multicenter Retrospective Case[sbnd]Control Study

This study aimed to develop and evaluate a radiomics-based model combined with clinical and qualitative radiological (semantic feature [SF]) features to predict immune checkpoint inhibitor-related pneumonitis (CIP) in patients with non-small cell lung cancer (NSCLC) treated with programmed cell death protein 1 inhibitors. This was a multicenter retrospective casecontrol study conducted from January 1, 2018, to December 31, 2022, at three centers. Patients with NSCLC treated with anti-PD1 were enrolled and randomly divided into two groups (7:3): training (n=95) and validation (n=39). Logistic regression (LR) and support vector machine (SVM) algorithms were used to transform features into the models. The study comprised 134 participants from three independent centers (male, 114/134, 85%; mean [±standard deviation] age, 63.92 [±7.9]years). The radiomics score (RS) models built based on the LR and SVM algorithms could accurately predict CIP (area under the receiver operating characteristics curve [AUC], 0.860 [0.780, 0.939] and 0.861 [0.781, 0.941], respectively). The AUCs for the RS-clinic-SF combined model were 0.903 (0.839, 0.967) and 0.826 (0.688, 0.964) in the training and validation cohorts, respectively. Decision curve analysis showed that the combined models achieved high clinical net benefit across the majority of the range of reasonable threshold probabilities. This study demonstrated that the combined model constructed by the identified features of RS, clinical features, and SF has the potential to precisely predict CIP. The RS-clinic-SF combined model has the potential to be used more widely as a practical tool for the noninvasive prediction of CIP to support individualized treatment planning.

Derivation and validation of a noninvasive prediction tool to identify pulmonary hypertension in patients with IPF: Evolution of the model FORD

The administration of inhaled prostanoids to patients with pulmonary hypertension (PH) related to idiopathic pulmonary fibrosis (IPF) and other fibrotic lung diseases improves functional outcomes. Selection of patients with IPF at risk for concomitant PH to undergo right heart catheterization (RHC) remains challenging. We sought to develop a clinical prediction tool based on common noninvasive parameters to identify PH in patients with IPF. A prediction model based on noninvasive parameters was derived from patients enrolled in the ARTEMIS-IPF randomized, placebo-controlled clinical trial. Predictor variables were tested for association with the presence of PH diagnosed based on RHC. The derived multivariable logistic regression model and associated point-score index were then externally validated in a real-world cohort of patients with IPF. Of the 481 patients included in the ARTEMIS-IPF study, 9.8% (N = 47) were diagnosed with PH related to IPF. Four variables were associated with PH and were included in the final model: forced vital capacity/diffusing capacity for carbon monoxide ratio (F), oxygen saturation nadir during 6-minute walk test (6MWT) (O), race (R), and distance ambulated during 6MWT (D). A model containing continuous predictors (FORD calculator) and a simple point-score system (FORD index) performed similarly well in the derivation cohort (area under the curve [AUC]: 0.75 and 0.75, respectively) and validation cohort (AUC: 0.69 and 0.69, respectively). The FORD models are simple, validated tools incorporating noninvasive parameters that can be applied to identify patients at risk of PH related to IPF who may benefit from invasive testing.

Preoperative Prediction of New Vertebral Fractures after Vertebral Augmentation with a Radiomics Nomogram.

The occurrence of new vertebral fractures (NVFs) after vertebral augmentation (VA) procedures is common in patients with osteoporotic vertebral compression fractures (OVCFs), leading to painful experiences and financial burdens. We aim to develop a radiomics nomogram for the preoperative prediction of NVFs after VA. Data from center 1 (training set: n = 153; internal validation set: n = 66) and center 2 (external validation set: n = 44) were retrospectively collected. Radiomics features were extracted from MRI images and radiomics scores (radscores) were constructed for each level-specific vertebra based on least absolute shrinkage and selection operator (LASSO). The radiomics nomogram, integrating radiomics signature with presence of intravertebral cleft and number of previous vertebral fractures, was developed by multivariable logistic regression analysis. The predictive performance of the vertebrae was level-specific based on radscores and was generally superior to clinical variables. RadscoreL2 had the optimal discrimination (AUC ≥ 0.751). The nomogram provided good predictive performance (AUC ≥ 0.834), favorable calibration, and large clinical net benefits in each set. It was used successfully to categorize patients into high- or low-risk subgroups. As a noninvasive preoperative prediction tool, the MRI-based radiomics nomogram holds great promise for individualized prediction of NVFs following VA.

Entirely noninvasive outcome prediction in central nervous system lymphomas using circulating tumor DNA

AbstractState-of-the-art response assessment of central nervous system lymphoma (CNSL) by magnetic resonance imaging is challenging and an insufficient predictor of treatment outcomes. Accordingly, the development of novel risk stratification strategies in CNSL is a high unmet medical need. We applied ultrasensitive circulating tumor DNA (ctDNA) sequencing to 146 plasma and cerebrospinal fluid (CSF) samples from 67 patients, aiming to develop an entirely noninvasive dynamic risk model considering clinical and molecular features of CNSL. Our ultrasensitive method allowed for the detection of CNSL-derived mutations in plasma ctDNA with high concordance to CSF and tumor tissue. Undetectable plasma ctDNA at baseline was associated with favorable outcomes. We tracked tumor-specific mutations in plasma-derived ctDNA over time and developed a novel CNSL biomarker based on this information: peripheral residual disease (PRD). Persistence of PRD after treatment was highly predictive of relapse. Integrating established baseline clinical risk factors with assessment of radiographic response and PRD during treatment resulted in the development and independent validation of a novel tool for risk stratification: molecular prognostic index for CNSL (MOP-C). MOP-C proved to be highly predictive of outcomes in patients with CNSL (failure-free survival hazard ratio per risk group of 6.60; 95% confidence interval, 3.12-13.97; P < .0001) and is publicly available at www.mop-c.com. Our results highlight the role of ctDNA sequencing in CNSL. MOP-C has the potential to improve the current standard of clinical risk stratification and radiographic response assessment in patients with CNSL, ultimately paving the way toward individualized treatment.

Cardiovascular and inflammatory biomarkers for non-invasive detection of coronary artery obstruction and prediction of long-term survival in patients with suspected chronic coronary syndrome

Abstract Introduction and methods: The knowledge about diagnostic and prognostic value of cardiac and inflammatory biomarkers in patients with chronic coronary syndrome (CCS) is limited. To address this, we analyzed serum levels of selected biomarkers in 2536 patients with suspected CCS (35% female) from the Leipzig LIFE Heart Study who were admitted for coronary angiography. The median follow-up was 10 years. The following biomarkers were considered: high sensitive troponin T (hsTNT), N-terminal pro B-type natriuretic peptide (NT-proBNP), copeptin, high sensitive C-reactive protein (hsCRP) and interleukin-6 (IL-6). Patients were stratified according to the angiographic severity of coronary artery disease (CAD): CAD0 (no sclerosis), CAD1 (non-obstructive, i.e., stenosis &amp;lt; 50%), CAD2 (≥ one stenosis ≥ 50%). Group comparison (GC) included GC1: CAD0 + 1 vs. CAD2, GC2: CAD0 vs. CAD1 + 2. Using age-, sex and symptom-based pre-test probability (PTP) table for obstructive CAD (i.e. CAD 2) which was published in the current CCS guidelines, patients were further classified into the following three categories: PTP &amp;lt; 5% (n=559), PTP 5-15% (n=545) and PTP &amp;gt; 15% (n=1432). Results CAD0, CAD1, CAD2 were apparent in 999, 529, and 1008 patients, respectively. Upon adjustment for traditional risk factors (TRF) the levels of hsTNT, NT-proBNP and IL-6 showed significant difference in GC1 (hsTNT: p=2.0x10-10, NT-proBNP: p=0.049, IL6: p=0.030). In GC2 only elevated hsTNT remained significant (p=1.9x10-7). In the ROC analysis only hsTNT slightly improved the AUC for the GC1-comparison in addition to TRF (0.729 with vs 0.712 witt hsTNT, p=0.013). Within the PTP subcategories the following results were obtained: PTP &amp;lt; 5%: AUC 0.747 with vs 0.730 without hsTNT, p=0.017; PTP 5-15 %: AUC 0.636 with vs 0.670 without hsTNT, p&amp;lt;0.001; PTP &amp;gt; 15%: AUC 0.699 with vs 0.677 without hsTNT, p= 0.024). In PTP 5-15% subgroup TropT cut-off value &amp;lt; 3 pg/mL (N= 119 / 23.4%) showed sensitivity of 89.7% and specificity of 73.1% for CAD2 detection. Ten years survival in groups CAD0, CAD1, CAD2 were 88.3%, 77.3%, 72.4%, respectively. In the multivariate analysis elevated hsTNT, NT-proBNP, copeptin and IL-6 remained significant mortality predictors in CAD2 patients with hazard ratios of similar magnitude (i.e. 1.5, 1.3, 1.4 and 1.3 per unit on the log-scale of the parameters, respectively). hsCRP did not reach significance. In the model stratified into tertiles according to the effects of classical risk factors and the joint biomarker levels for except hsCRP, hazard rates were 3.13 (tertile 2 vs 1) and 11.2 (tertile 3 vs 1). Conclusions In the present study hsTNT substantially improved the detection of obstructive CAD particularly in patients with intermediate PTP 5-15%. Furthermore, the studied biomarkers enable fast and precise non-invasive prediction of mortality risk in patients with suspected CCS, allowing tailored primary and secondary CAD prevention in this high-risk group.

Abstract 14855: Noninvasive Prediction of Pulmonary Capillary Wedge Pressure in Patients With Normal LV EF: Comparison of CMR With Echocardiography

Background: Cardiac magnetic resonance (CMR) was recently reported to predict mean pulmonary capillary wedge pressure (PCWP) with the best accuracy in mid-rage or reduced left ventricular (LV) ejection fraction (EF). . There is a paucity of data on the accuracy of CMR estimation of PCWP in patients with normal left ventricular (LV) ejection fraction (EF) or on how CMR compares with echocardiography in these patients. Methods: Stable patients with EF of ≥50%, who underwent right heart catheterization (RHC), CMR, and echocardiographic imaging within one week were included. PCWP was estimated by CMR using a previously validated equation where PCWP is estimated based on the left atrial maximum volume (LAV) and left ventricular mass (LVM) as PCWP=6.1352+ (0.07204*LAV) + (0.02256*LVM). Echocardiographic estimation of PCWP was based on 2016 ASE/EACVI guidelines taking into account the presence of myocardial disease and pulmonary vein atrial velocity. Results: Mean age was (54.6 ± 15.2), 58.2% females. There were 33 patients with PCWP &gt; 15 mmHg by RHC and 46 with a normal PCWP by right heart catheterization. The table shows sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for the two imaging modalities. CMR prediction of PCWP had an AUC=0.72. In comparison, echocardiographic prediction of PCWP had significantly higher accuracy (AUC=0.87 vs. AUC=0.72; P- value=0.008). Conclusion: In patients with normal LV EF, CMR estimation of mean PCWP has an acceptable accuracy for detecting patients with mean PCWP&gt;15 mmHg. Echocardiography predicts elevated PCWP with a significantly higher sensitivity, specificity, and overall accuracy in comparison with CMR.

Development and validation of non-invasive prediction models for migraine in Chinese adults

BackgroundMigraine is a common disabling neurological disorder with severe physical and psychological damage, but there is a lack of convenient and effective non-invasive early prediction methods. This study aimed to develop a new series of non-invasive prediction models for migraine with external validation.MethodsA total of 188 and 94 subjects were included in the training and validation sets, respectively. A standardized professional questionnaire was used to collect the subjects' 9-item traditional Chinese medicine constitution (TCMC) scores, Pittsburgh Sleep Quality Index (PSQI) score, Zung's Self-rating Anxiety Scale and Self-rating Depression Scale scores. Logistic regression was used to analyze the risk predictors of migraine, and a series of prediction models for migraine were developed. Receiver operating characteristic (ROC) curve and calibration curve were used to assess the discrimination and calibration of the models. The predictive performance of the models were further validated using external datasets and subgroup analyses were conducted.ResultsPSQI score and Qi-depression score were significantly and positively associated with the risk of migraine, with the area of the ROC curves (AUCs) predicting migraine of 0.83 (95% CI:0.77–0.89) and 0.76 (95% CI:0.68–0.84), respectively. Eight non-invasive predictive models for migraine containing one to eight variables were developed using logistic regression, with AUCs ranging from 0.83 (95% CI: 0.77–0.89) to 0.92 (95% CI: 0.89–0.96) for the training set and from 0.76 (95% CI: 0.66–0.85) to 0.83 (95% CI: 0.75–0.91) for the validation set. Subgroup analyses showed that the AUCs of the eight prediction models for predicting migraine in the training and validation sets of different gender and age subgroups ranged from 0.80 (95% CI: 0.63–0.97) to 0.95 (95% CI: 0.91–1.00) and 0.73 (95% CI: 0.64–0.84) to 0.93 (95% CI: 0.82–1.00), respectively.ConclusionsThis study developed and validated a series of convenient and novel non-invasive prediction models for migraine, which have good predictive ability for migraine in Chinese adults of different genders and ages. It is of great significance for the early prevention, screening, and diagnosis of migraine.

Noninvasive prediction of perineural invasion in intrahepatic cholangiocarcinoma by clinicoradiological features and computed tomography radiomics based on interpretable machine learning: a multicenter cohort study.

Perineural invasion (PNI) of intrahepatic cholangiocarcinoma (ICC) is a strong independent risk factor for tumour recurrence and long-term patient survival. However, there is a lack of noninvasive tools for accurately predicting the PNI status. The authors develop and validate a combined model incorporating radiomics signature and clinicoradiological features based on machine learning for predicting PNI in ICC, and used the Shapley Additive explanation (SHAP) to visualize the prediction process for clinical application. This retrospective and prospective study included 243 patients with pathologically diagnosed ICC (training, n =136; external validation, n =81; prospective, n =26, respectively) who underwent preoperative contrast-enhanced computed tomography between January 2012 and May 2023 at three institutions (three tertiary referral centres in Guangdong Province, China). The ElasticNet was applied to select radiomics features and construct signature derived from computed tomography images, and univariate and multivariate analyses by logistic regression were used to identify the significant clinical and radiological variables with PNI. A robust combined model incorporating radiomics signature and clinicoradiological features based on machine learning was developed and the SHAP was used to visualize the prediction process. A Kaplan-Meier survival analysis was performed to compare prognostic differences between PNI-positive and PNI-negative groups and was conducted to explore the prognostic information of the combined model. Among 243 patients (mean age, 61.2years ± 11.0 (SD); 152 men and 91 women), 108 (44.4%) were diagnosed as PNI-positive. The radiomics signature was constructed by seven radiomics features, with areas under the curves of 0.792, 0.748, and 0.729 in the training, external validation, and prospective cohorts, respectively. Three significant clinicoradiological features were selected and combined with radiomics signature to construct a combined model using machine learning. The eXtreme Gradient Boosting exhibited improved accuracy and robustness (areas under the curves of 0.884, 0.831, and 0.831, respectively). Survival analysis showed the construction combined model could be used to stratify relapse-free survival (hazard ratio, 1.933; 95% CI: 1.093-3.418; P =0.021). We developed and validated a robust combined model incorporating radiomics signature and clinicoradiological features based on machine learning to accurately identify the PNI statuses of ICC, and visualize the prediction process through SHAP for clinical application.

From Diagnosis to Treatment: A Comprehensive Review of Biomarkers and Therapeutic Advances in Parkinson’s Disease

Background Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by the loss of dopaminergic neurons, resulting in motor symptoms. Ongoing research shows promise for long-term solutions. Summary Studies highlight the dysregulation of Syt11 and α-synuclein (α-syn) in PD. Disrupted α-syn homeostasis due to palmitoylation of Syt11 contributes to its aggregation, potentially playing a role in PD pathology. α-synuclein aggregates in stool samples show promise as an early diagnostic biomarker. Vocal impairments in PD may be linked to α-syn-induced neuropathology. Irisin, produced after exercise, promotes the degradation of pathologic α-syn. Progress has been made in identifying PD biomarkers. Retinal thinning and abnormal protein aggregates in skin biopsies provide noninvasive diagnostic indicators. Blood-based biomarkers like α-syn, DJ-1, and LRRK2 hold promise but face limitations. Artificial intelligence (AI) models enhance mitophagy, detect PD through sleep-breathing signals, and improve survival. AI analysis aids noninvasive assessment and risk prediction. Further understanding of PD involves studying pathological seeds and genetic mutations. Adenosine receptor regulation relates to early-onset PD, and specific gene mutations impact patient survival. Differentiated-induced pluripotent stem cells offer the potential for cell replacement therapy. Autoimmune features and T-cell involvement suggest intervention targets. Stem cell-based therapies and neurostimulation strategies show promise for improving motor function. Imaging reveals increased central inflammation in PD, suggesting an inflammatory role. Machine learning algorithms and home gait speed monitoring aid in diagnosis and disease progression tracking. Abnormal putamen gradients reflect dopaminergic loss and motor dysfunction. Antiepileptic drug prescriptions are associated with an increased PD risk. Personalized medicine, gut–brain axis involvement, and vestibular stimulation therapy offer potential PD treatment avenues. Genetic engineering techniques and deep brain stimulation show promise for alleviating PD symptoms. Key Message Ongoing research and technological advancements promise to improve PD screening, diagnosis, and treatment, bringing hope to affected individuals.

Non-invasive prediction of mango quality using near-infrared spectroscopy: Assessment on spectral interferences of different packaging materials

Non-invasive prediction of mango quality using near-infrared spectroscopy: Assessment on spectral interferences of different packaging materials

Longitudinal dynamic contrast-enhanced MRI radiomic models for early prediction of response to neoadjuvant systemic therapy in triple-negative breast cancer

Early prediction of neoadjuvant systemic therapy (NAST) response for triple-negative breast cancer (TNBC) patients could help oncologists select individualized treatment and avoid toxic effects associated with ineffective therapy in patients unlikely to achieve pathologic complete response (pCR). The objective of this study is to evaluate the performance of radiomic features of the peritumoral and tumoral regions from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) acquired at different time points of NAST for early treatment response prediction in TNBC. This study included 163 Stage I-III patients with TNBC undergoing NAST as part of a prospective clinical trial (NCT02276443). Peritumoral and tumoral regions of interest were segmented on DCE images at baseline (BL) and after two (C2) and four (C4) cycles of NAST. Ten first-order (FO) radiomic features and 300 gray-level-co-occurrence matrix (GLCM) features were calculated. Area under the receiver operating characteristic curve (AUC) and Wilcoxon rank sum test were used to determine the most predictive features. Multivariate logistic regression models were used for performance assessment. Pearson correlation was used to assess intrareader and interreader variability. Seventy-eight patients (48%) had pCR (52 training, 26 testing), and 85 (52%) had non-pCR (57 training, 28 testing). Forty-six radiomic features had AUC at least 0.70, and 13 multivariate models had AUC at least 0.75 for training and testing sets. The Pearson correlation showed significant correlation between readers. In conclusion, Radiomic features from DCE-MRI are useful for differentiating pCR and non-pCR. Similarly, predictive radiomic models based on these features can improve early noninvasive treatment response prediction in TNBC patients undergoing NAST.

Preoperative contrast-enhanced CT imaging and clinicopathological characteristics analysis of mismatch repair-deficient colorectal cancer

BackgroundColorectal cancer (CRC) can develop through various pathogenetic pathways, and one of the primary pathways is high microsatellite instability (MSI-H)/deficient mismatch repair (dMMR). This study investigated the correlation between preoperative contrast-enhanced CT (CECT) and clinicopathologic characteristics of colorectal cancer (CRC) according to different mismatch repair (MMR) statuses.MethodsFrom April 2021 to July 2022, a total of 281 CRC patients with preoperative CECT and available MMR status were enrolled from a single centre for this retrospective study. Preoperative CECT features and clinicopathologic characteristics were analysed. Univariate and multivariate logistic regression analyses were used for statistical analysis. A nomogram was established based on the multivariate logistic regression results. Preoperative and postoperative dynamic nomogram prediction models were established. The C-index, a calibration plot, and clinical applicability of the two models were evaluated, and internal validation was performed using three methods.ResultsIn total, 249 patients were enrolled in the proficient mismatch repair (pMMR) group and 32 patients in the deficient mismatch repair (dMMR) group. In multivariate analysis, tumour location (right-hemi colon vs. left-hemi colon, odds ratio (OR) = 2.90, p = .036), the hypoattenuation-within-tumour ratio (HR) (HR > 2/3 vs. HR < 1/3, OR = 36.7, p < .001; HR 1/3–2/3 vs. HR < 1/3, OR = 6.05, p = .031), the number of lymph nodes with long diameter ≥ 8 mm on CECT (OR = 1.32, p = .01), CEA status (CEA positive vs. CEA negative, OR = 0.07, p = .002) and lymph node metastasis (OR = 0.45, p = .008) were independent risk factors for dMMR. Pre- and postoperative C-statistic were 0.861 and 0.908, respectively.ConclusionThe combination of pre-operative CECT and clinicopathological characteristics of CRC correlates with MMR status, providing possible non-invasive MMR prediction. Particularly for dMMR CRC, tumour-draining lymph node status should be prudently evaluated by CECT.

A comparative proteomic analysis for non-invasive early prediction of hypoxic-ischemic injury in asphyxiated neonates.

Hypoxic Ischemic Encephalopathy (HIE) is one of the principal causes of neonatal mortality and long-term morbidity worldwide. The neonatal signs of mild cerebral injury are subtle, making an early precise diagnosis difficult. Delayed detection, poor prognosis, and lack of specific biomarkers for the disease are increasing mortality rates. In this study, we intended to identify specific biomarkers using comparative proteomic analysis to predict the severity of perinatal asphyxia so that its outcome can also be prevented. A case-control study was conducted on 38 neonates, and urine samples were collected within 24 and 72h of life. A tandem mass spectrometry-based quantitative proteomics approach, followed by validation via sandwich ELISA, was performed. The LC-MS/MS-based proteomics analysis resulted in the identification of 1201 proteins in urine, with 229, 244, and 426 being differentially expressed in HIE-1, HIE-2, and HIE-3, respectively. Axon guidance, Diseases of programmed cell death, and Detoxification of reactive oxygen species pathways were significantly enriched in mild HIE versus severe HIE. Among the differentially expressed proteins in various stages of HIE, we chose to validate four proteins - APP, AGT, FABP1, and FN1 - via sandwich ELISA. Individual and cumulative ROC curves were plotted. AGT and FABP1 together showed high sensitivity, specificity, and accuracy as potential biomarkers for early diagnosis of HIE. Establishing putative urinary biomarkers will facilitate clinicians to more accurately screen neonates for brain injury and monitor the disease progression. Prompt treatment of neonates may reduce mortality and neurodevelopmental impairment.

Optimizing Random Forest Algorithm to Classify Player's Memorisation via In-game Data

Assessment of a player's knowledge in game education has been around for some time. Traditional evaluation in and around a gaming session may disrupt the players' immersion. This research uses an optimized Random Forest to construct a non-invasive prediction of a game education player's Memorization via in-game data. Firstly, we obtained the dataset from a 3-month survey to record in-game data of 50 players who play 4-15 game stages of the Chem Fight (a test case game). Next, we generated three variants of datasets via the preprocessing stages: resampling method (SMOTE), normalization (min-max), and a combination of resampling and normalization. Then, we trained and optimized three Random Forest (RF) classifiers to predict the player's Memorization. We chose RF because it can generalize well given the high-dimensional dataset. We used RF as the classifier, subject to optimization using its hyperparameter: n_estimators. We implemented a Grid Search Cross Validation (GSCV) method to identify the best value of n_estimators. We utilized the statistics of GSCV results to reduce the weight of n_estimators by observing the region of interest shown by the graphs of performances of the classifiers. Overall, the classifiers fitted using the BEST n_estimators (i.e., 89, 31, 89, and 196 trees) from GSCV performed well with around 80% accuracy. Moreover, we successfully identified the smaller number of n_estimators (OPTIMAL), at least halved the BEST n_estimators. All classifiers were retrained using the OPTIMAL n_estimators (37, 12, 37, and 41 trees). We found out that the performances of the classifiers were relatively steady at ~80%. This means that we successfully optimized the Random Forest in predicting a player's Memorization when playing the Chem Fight game. An automated technique presented in this paper can monitor student interactions and evaluate their abilities based on in-game data. As such, it can offer objective data about the skills used.

Imaging prognostication and tumor biology in hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) is the most common primary liver malignancy, and represents a significant global health burden with rising incidence rates, despite a more thorough understanding of the etiology and biology of HCC, as well as advancements in diagnosis and treatment modalities. According to emerging evidence, imaging features related to tumor aggressiveness can offer relevant prognostic information, hence validation of imaging prognostic features may allow for better noninvasive outcomes prediction and inform the selection of tailored therapies, ultimately improving survival outcomes for patients with HCC.

Molecular-subtype guided automatic invasive breast cancer grading using dynamic contrast-enhanced MRI

Background and objectivesHistological grade and molecular subtype have presented valuable references in assigning personalized or precision medicine as the significant prognostic indicators representing biological behaviors of invasive breast cancer (IBC). To evaluate a two-stage deep learning framework for IBC grading that incorporates with molecular-subtype (MS) information using DCE-MRI. MethodsIn Stage I, an innovative neural network called IOS2-DA is developed, which includes a dense atrous-spatial pyramid pooling block with a pooling layer (DA) and inception-octconved blocks with double kernel squeeze-and-excitations (IOS2). This method focuses on the imaging manifestation of IBC grades and performs preliminary prediction using a novel class F1-score loss function. In Stage II, a MS attention branch is introduced to fine-tune the integrated deep vectors from IOS2-DA via Kullback-Leibler divergence. The MS-guided information is weighted with preliminary results to obtain classification values, which are analyzed by ensemble learning for tumor grade prediction on three MRI post-contrast series. Objective assessment is quantitatively evaluated by receiver operating characteristic curve analysis. DeLong test is applied to measure statistical significance (P < 0.05). ResultsThe molecular-subtype guided IOS2-DA performs significantly better than the single IOS2-DA in terms of accuracy (0.927), precision (0.942), AUC (0.927, 95% CI: [0.908, 0.946]), and F1-score (0.930). The gradient-weighted class activation maps show that the feature representations extracted from IOS2-DA are consistent with tumor areas. ConclusionsIOS2-DA elucidates its potential in non-invasive tumor grade prediction. With respect to the correlation between MS and histological grade, it exhibits remarkable clinical prospects in the application of relevant clinical biomarkers to enhance the diagnostic effectiveness of IBC grading. Therefore, DCE-MRI tends to be a feasible imaging modality for the thorough preoperative assessment of breast biological behavior and carcinoma prognosis.

Impact of signal intensity normalization of MRI on the generalizability of radiomic-based prediction of molecular glioma subtypes

ObjectivesRadiomic features have demonstrated encouraging results for non-invasive detection of molecular biomarkers, but the lack of guidelines for pre-processing MRI-data has led to poor generalizability. Here, we assessed the influence of different MRI-intensity normalization techniques on the performance of radiomics-based models for predicting molecular glioma subtypes.MethodsPreoperative MRI-data from n = 615 patients with newly diagnosed glioma and known isocitrate dehydrogenase (IDH) and 1p/19q status were pre-processed using four different methods: no normalization (naive), N4 bias field correction (N4), N4 followed by either WhiteStripe (N4/WS), or z-score normalization (N4/z-score). A total of 377 Image-Biomarker-Standardisation-Initiative-compliant radiomic features were extracted from each normalized data, and 9 different machine-learning algorithms were trained for multiclass prediction of molecular glioma subtypes (IDH-mutant 1p/19q codeleted vs. IDH-mutant 1p/19q non-codeleted vs. IDH wild type). External testing was performed in public glioma datasets from UCSF (n = 410) and TCGA (n = 160).ResultsSupport vector machine yielded the best performance with macro-average AUCs of 0.84 (naive), 0.84 (N4), 0.87 (N4/WS), and 0.87 (N4/z-score) in the internal test set. Both N4/WS and z-score outperformed the other approaches in the external UCSF and TCGA test sets with macro-average AUCs ranging from 0.85 to 0.87, replicating the performance of the internal test set, in contrast to macro-average AUCs ranging from 0.19 to 0.45 for naive and 0.26 to 0.52 for N4 alone.ConclusionIntensity normalization of MRI data is essential for the generalizability of radiomic-based machine-learning models. Specifically, both N4/WS and N4/z-score approaches allow to preserve the high model performance, yielding generalizable performance when applying the developed radiomic-based machine-learning model in an external heterogeneous, multi-institutional setting.Clinical relevance statementIntensity normalization such as N4/WS or N4/z-score can be used to develop reliable radiomics-based machine learning models from heterogeneous multicentre MRI datasets and provide non-invasive prediction of glioma subtypes.Key Points• MRI-intensity normalization increases the stability of radiomics-based models and leads to better generalizability.• Intensity normalization did not appear relevant when the developed model was applied to homogeneous data from the same institution.• Radiomic-based machine learning algorithms are a promising approach for simultaneous classification of IDH and 1p/19q status of glioma.

68Ga-PSMA PET/CT-based multivariate model for highly accurate and noninvasive diagnosis of clinically significant prostate cancer in the PSA gray zone

BackgroundThe prostate-specific antigen (PSA) has been widely used in screening and early diagnosis of prostate cancer (PCa). However, in the PSA grey zone of 4–10 ng/ml, the sensitivity and specificity for diagnosing PCa are limited, resulting in considerable number of unnecessary and invasive prostate biopsies, which may lead to potential overdiagnosis and overtreatment. We aimed to predict clinically significant PCa (CSPCa) by combining the maximal standardized uptake value (SUVmax) based on 68Ga‑PSMA PET/CT and clinical indicators in men with gray zone PSA levels.Methods81 patients with suspected PCa based on increased serum total PSA (TPSA) levels of 4 − 10 ng/mL who underwent transrectal ultrasound/magnetic resonance imaging (MRI)/PET fusion-guided biopsy were enrolled. Among them, patients confirmed by histopathology were divided into the CSPCa group and the non-CSPCa group, and data on PSA concentration, prostate volume (PV), PSA density (PSAD), free PSA (FPSA)/TPSA, Prostate Imaging-Reporting and Data System version 2.1 (PI-RADS v2.1) score, 68Ga-PSMA PET/CT imaging evaluation results and SUVmax were compared. Multivariate logistic regression analysis was performed to identify the independent predictors for CSPCa, thereby establishing a predictive model based on SUVmax that was evaluated by analyzing the receiver operating characteristic (ROC) curve and decision curve analysis.ResultsCompared to non-CSPCa, CSPCa patients had smaller PVs (median, 31.40 mL), lower FPSA/TPSA (median, 0.12), larger PSADs (median, 0.21 ng/mL2) and higher PI-RADS scores (P < 0.05). The prediction model comprising 68Ga-PSMA PET/CT maximal standardized uptake value, PV and FPSA/TPSA had the highest AUC of 0.927 compared with that of other predictors alone (AUCs of 0.585 for PSA, 0.652 for mpMRI and 0.850 for 68Ga-PSMA PET/CT). The diagnostic sensitivity and specificity of the prediction model were 86.21% and 86.54%, respectively.ConclusionGiven the low diagnostic accuracy of regular PSA tests, a new prediction model based on the 68Ga-PSMA PET/CT SUVmax, PV and FPSA/TPSA was developed and validated, and this model could provide a more satisfactory predictive accuracy for CSPCa. This study provides a noninvasive prediction model with high accuracy for the diagnosis of CSPCa in the PSA gray zone, thus may be better avoiding unnecessary biopsy procedures.