Prediction Of Individual Outcome Research Articles

Improving outcome prediction in individuals with colorectal cancer and diabetes by accurate assessment of vascular complications: Implications for clinical practice

BackgroundDiabetes is considered a risk factor for mortality following a diagnosis of cancer. We hypothesised that the risk will vary due to the heterogeneous nature of the population and accurate classification of vascular complications will improve prediction of clinical outcomes. MethodsThe COloRECTal cancer data Repository (CORECT-R) was used to identify individuals with primary colorectal cancer, who underwent surgical resection in England (2005–2016). Diabetes was recorded using ICD10 codes (E10-E14) during inpatient hospital admission in the six years preceding cancer diagnosis, complication status was determined using the adapted Diabetes Complications Severity Index (aDCSI). Survival and post-operative outcomes were compared between groups. ResultsOf 232,367 individuals, 28,642 (12.3%) were recorded as having diabetes, 49.2% of whom had complications according to the aDCSI. Patients with diabetes complications had increased incidence of adverse post-operative outcomes (90-day post-operative mortality (6.6% versus 3.2%) and death during the surgical episode (7.9% versus 3.6%)), compared to those without diabetes. Those without complications had rates comparable to the population without diabetes. The odds of death within a year of diagnosis were higher for those with complicated diabetes compared to those without diabetes [OR 1.58 (95%CI 1.51–1.66) p < 0.01], but no difference was observed between those with uncomplicated diabetes and those without diabetes [OR 1.05 (95%CI 0.99–1.11) p = 0.10]. ConclusionsPrediction of outcome following surgery in colorectal cancer patients with diabetes relies on the accurate assessment of complications. This study suggests that the poor post-operative outcomes in diabetes patients may be associated with diabetes complication rather than diabetes itself.

Deep Learning Analysis of Derived Cardiac Function Metrics for the Detection of CKD and Subsequent Outcome Prediction in Community-Dwelling Individuals

Deep Learning Analysis of Derived Cardiac Function Metrics for the Detection of CKD and Subsequent Outcome Prediction in Community-Dwelling Individuals

Development and validation of a model for individualized prediction of hospitalization risk in 4,536 patients with COVID-19.

Coronavirus Disease 2019 is a pandemic that is straining healthcare resources, mainly hospital beds. Multiple risk factors of disease progression requiring hospitalization have been identified, but medical decision-making remains complex. To characterize a large cohort of patients hospitalized with COVID-19, their outcomes, develop and validate a statistical model that allows individualized prediction of future hospitalization risk for a patient newly diagnosed with COVID-19. Retrospective cohort study of patients with COVID-19 applying a least absolute shrinkage and selection operator (LASSO) logistic regression algorithm to retain the most predictive features for hospitalization risk, followed by validation in a temporally distinct patient cohort. The final model was displayed as a nomogram and programmed into an online risk calculator. One healthcare system in Ohio and Florida. All patients infected with SARS-CoV-2 between March 8, 2020 and June 5, 2020. Those tested before May 1 were included in the development cohort, while those tested May 1 and later comprised the validation cohort. Demographic, clinical, social influencers of health, exposure risk, medical co-morbidities, vaccination history, presenting symptoms, medications, and laboratory values were collected on all patients, and considered in our model development. 4,536 patients tested positive for SARS-CoV-2 during the study period. Of those, 958 (21.1%) required hospitalization. By day 3 of hospitalization, 24% of patients were transferred to the intensive care unit, and around half of the remaining patients were discharged home. Ten patients died. Hospitalization risk was increased with older age, black race, male sex, former smoking history, diabetes, hypertension, chronic lung disease, poor socioeconomic status, shortness of breath, diarrhea, and certain medications (NSAIDs, immunosuppressive treatment). Hospitalization risk was reduced with prior flu vaccination. Model discrimination was excellent with an area under the curve of 0.900 (95% confidence interval of 0.886-0.914) in the development cohort, and 0.813 (0.786, 0.839) in the validation cohort. The scaled Brier score was 42.6% (95% CI 37.8%, 47.4%) in the development cohort and 25.6% (19.9%, 31.3%) in the validation cohort. Calibration was very good. The online risk calculator is freely available and found at https://riskcalc.org/COVID19Hospitalization/. Retrospective cohort design. Our study crystallizes published risk factors of COVID-19 progression, but also provides new data on the role of social influencers of health, race, and influenza vaccination. In a context of a pandemic and limited healthcare resources, individualized outcome prediction through this nomogram or online risk calculator can facilitate complex medical decision-making.

The modified MIDA-Score predicts mid-term outcomes after interventional therapy of functional mitral regurgitation.

The preprocedural assessment of outcomes and patients' prognosis after interventional therapy of functional MR (FMR) is uncertain. Therefore, we aim to develop an easy-to-handle scoring system for adequate prediction of individual outcomes in patients with FMR after the interventional treatment. We retrospectively used medical data of patients with symptomatic FMR, who underwent transcatheter mitral valve repair (TMVR) from January 2014 to August 2016 in our heart center. All patients had the mean follow-up of 18 months. All clinical and echocardiographic data originate from the "Bonner Mitral Valve Register Database". We included 105 patients (76,7±8,8 years, 50,6% female) with symptomatic (NYHA functional class>II) moderate-to-severe or severe FMR at surgical high-risk. We modified the MIDA-Score for degenerative MR (DMR) according to the varying underlying pathomechanisms of FMR, called as "The modified MIDA Score". We found all-cause mortality of 7% within 18 months after the procedure. 34,1% of our cohort was rehospitalized; 90% of those were due to cardiovascular causes. The modified MIDA score was found to be a strong predictor for mortality and rehospitalization in patients with FMR (AUC: 0,89) and superior to the other conventional scoring systems in prediction of mortality (The modified MIDA-Score: AUC: 0,8, EuroSCORE II: AUC: 0,57, STS-Score: AUC: 0,51). The logistic regression analysis showed the modified MIDA score > 9 points to be the strongest predictor for mortality and rehospitalization after TMVR (OR: 3,35, p = 0,011). The modified MIDA score was found to be a promising, easy-to-handle, elementary scoring system for adequate prediction of individual postinterventional prognosis in patients with FMR undergoing TMVR. Further evaluation and validation of this novel scoring system in prospective multicentric studies with a large number of patients is warranted.

Exploration of the molecular characteristics of the tumor-immune interaction and the development of an individualized immune prognostic signature for neuroblastoma.

Neuroblastoma (NBL) exists in a complex tumor-immune microenvironment. Immune cell infiltration and tumor-immune molecules play a critical role in tumor development and significantly impact the prognosis of patients. However, the molecular characteristics describing the NBL-immune interaction and their prognostic potential have yet to be investigated systematically. We first employed multiple machine learning algorithms, such as Gene Sets Enrichment Analysis and cell type identification by estimating relative subsets of RNA transcripts, to identify immunophenotypes and immunological characteristics in NBL patient data from public databases and then investigated the prognostic potential and regulatory networks of identified immune-related genes involved in the NBL-immune interaction. The immunity signature combining nine immunity genes was confirmed as more effective for individual risk stratification and survival outcome prediction in NBL patients than common clinical characteristics (area under the curve [AUC] = 0.819, C-index = 0.718, p < .001). A mechanistic exploration revealed the regulatory network of molecules involved in the NBL-immune interaction. These immune molecules were also discovered to possess a significant correlation with plasma cell infiltration, MYCN status, and the level of chemokines and macrophage-related molecules (p < .001). A nomogram was constructed based on the immune signature and clinical characteristics, which showed high potential for prognosis prediction (AUC = 0.856, C-index = 0.755, p < .001). We systematically elucidated the complex regulatory mechanisms and characteristics of the molecules involved in the NBL-immune interaction and their prognostic potential, which may have important implications for further understanding the molecular mechanism of the NBL-immune interaction and identifying high-risk NBL patients to guide clinical treatment.

Interleukin 10 and Heart Fatty Acid-Binding Protein as Early Outcome Predictors in Patients With Traumatic Brain Injury.

Background: Patients with traumatic brain injury (TBI) exhibit a variable and unpredictable outcome. The proteins interleukin 10 (IL-10) and heart fatty acid-binding protein (H-FABP) have shown predictive values for the presence of intracranial lesions.Aim: To evaluate the individual and combined outcome prediction ability of IL-10 and H-FABP, and to compare them to the more studied proteins S100β, glial fibrillary acidic protein (GFAP), and neurofilament light (NF-L), both with and without clinical predictors.Methods: Blood samples from patients with acute TBI (all severities) were collected <24 h post trauma. The outcome was measured >6 months post injury using the Glasgow Outcome Scale Extended (GOSE) score, dichotomizing patients into: (i) those with favorable (GOSE≥5)/unfavorable outcome (GOSE ≤ 4) and complete (GOSE = 8)/incomplete (GOSE ≤ 7) recovery, and (ii) patients with mild TBI (mTBI) and patients with TBIs of all severities.Results: When sensitivity was set at 95–100%, the proteins' individual specificities remained low. H-FABP showed the best specificity (%) and sensitivity (100%) in predicting complete recovery in patients with mTBI. IL-10 had the best specificity (50%) and sensitivity (96%) in identifying patients with favorable outcome in patients with TBIs of all severities. When individual proteins were combined with clinical parameters, a model including H-FABP, NF-L, and ISS yielded a specificity of 56% and a sensitivity of 96% in predicting complete recovery in patients with mTBI. In predicting favorable outcome, a model consisting IL-10, age, and TBI severity reached a specificity of 80% and a sensitivity of 96% in patients with TBIs of all severities.Conclusion: Combining novel TBI biomarkers H-FABP and IL-10 with GFAP, NF-L and S100β and clinical parameters improves outcome prediction models in TBI.

Towards data-driven decision-making for breast cancer patients undergoing mastectomy and reconstruction: Prediction of individual patient-reported outcomes at two-year follow-up using machine learning.

520 Background: Post-surgical satisfaction with breasts is a key outcome for women undergoing cancer-related mastectomy and reconstruction. Current decision making relies on group-level evidence, which may not offer optimal choice of treatment for individuals. We developed and validated machine learning algorithms to predict individual post-surgical breast-satisfaction. We aim to facilitate individualized data-driven decision making in breast cancer. Methods: We collected clinical, perioperative, and patient-reported data from 3058 women who underwent breast reconstruction due to breast cancer across 11 sites in North America. We trained and evaluated four algorithms (regularized regression, Support Vector Machine, Neural Network, Regression Tree) to predict significant changes in satisfaction with breasts at 2-year follow up using the validated BREAST-Q measure. Accuracy and area under the receiver operating characteristics curve (AUC) were used to determine algorithm performance in the test sample. Results: Machine learning algorithms were able to accurately predict changes in women’s satisfaction with breasts (see table). Baseline satisfaction with breasts was the most informative predictor of outcome, followed by radiation during or after reconstruction, nipple-sparing and mixed mastectomy, implant-based reconstruction, chemotherapy, unilateral mastectomy, lower psychological well-being, and obesity. Conclusions: We reveal the crucial role of patient-reported outcomes in determining post-operative outcomes and that Machine Learning algorithms are suitable to identify individuals who might benefit from alternative treatment decisions than suggested by group-level evidence. We provide a web-based tool for individuals considering mastectomy and reconstruction. importdemo.com . Clinical trial information: NCT01723423 . [Table: see text]

S147. FUNCTIONAL BRAIN CONNECTIVITY DATA IMPROVE CLINICAL OUTCOME PREDICTION IN YOUTH AT RISK FOR PSYCHOSIS

BackgroundImproved outcome prediction in individuals at high risk for psychosis may facilitate targeted early intervention. Studies suggest that improved outcome prediction may be achieved through the use of neurocognitive or neuroimaging data, on their own or in addition to clinical data. This study examines whether adding resting-state functional connectivity data to validated clinical predictors of psychosis improve outcome prediction in the prodromal stage.MethodsThis study involves 137 adolescents and young adults at Clinical High Risk (CHR) for psychosis from the Shanghai At Risk for Psychosis (SHARP) program. Based on outcome after one-year follow-up, participants were separated into three outcome categories: good outcome (symptom remission, N = 71), intermediate outcome (ongoing CHR symptoms, N = 30), and poor outcome (conversion to psychosis or treatment-refractory, N = 36). Resting-state fMRI data were acquired for each participant and processed using the Conn toolbox, including rigorous motion correction. Multinomial logistic regression analysis and leave-one-out cross-validation were used to assess the performance of three prediction models: 1) a clinical-only model using validated clinical predictors from the NAPLS-2 psychosis-risk calculator, 2) an fMRI-only model using measures of functional connectome organization and within/between-network connectivity among established resting-state networks, and 3) a combined clinical and fMRI prediction model. Model performance was assessed using the harmonic mean of the positive predictive value and sensitivity for each outcome category. This F1 measure was compared to expected chance-levels using a permutation test with 1,000 sampled permutations in order to evaluate the statistical significance of the model’s prediction.ResultsThe clinical-only prediction model failed to achieve a significant level of outcome prediction (F1 = 0.32, F1-chance = 0.26 □ 0.06, p = .154). The fMRI-only model did predict clinical outcome to a significant degree (F1 = 0.41, F1-chance = 0.29 □ 0.06, p = .016), but the combined clinical and fMRI prediction model showed the best performance (F1 = 0.46, F1-chance = 0.29 □ 0.06, p < .001). On average, positive predictive values (reflecting the probability that an outcome label predicted by the model was correct) were 39% better than chance-level and 32% better than the clinical-only model. Analyzing the contribution of individual predictor variables showed that GAF functional decline, a family history of psychosis, and performance on the Hopkins Verbal Learning Test were the most influential clinical predictors, whereas modular connectome organization, default-mode and fronto-parietal within-network connectivity, and between-network connectivity among language, salience, dorsal attention, cerebellum, and sensorimotor networks were the leading fMRI predictors.DiscussionThis study’s findings suggest that functional brain abnormalities reflected by alterations in resting-state functional connectivity precede and may drive subsequent changes in clinical functioning. Moreover, the findings show that markers of functional brain connectivity may be useful for improving early identification and clinical decision-making in prodromal psychosis.

Radiomics Analysis of Computed Tomography helps predict poor prognostic outcome in COVID-19.

Rationale: Given the rapid spread of COVID-19, an updated risk-stratify prognostic tool could help clinicians identify the high-risk patients with worse prognoses. We aimed to develop a non-invasive and easy-to-use prognostic signature by chest CT to individually predict poor outcome (death, need for mechanical ventilation, or intensive care unit admission) in patients with COVID-19.Methods: From November 29, 2019 to February 19, 2020, a total of 492 patients with COVID-19 from four centers were retrospectively collected. Since different durations from symptom onsets to the first CT scanning might affect the prognostic model, we designated the 492 patients into two groups: 1) the early-phase group: CT scans were performed within one week after symptom onset (0-6 days, n = 317); and 2) the late-phase group: CT scans were performed one week later after symptom onset (≥7 days, n = 175). In each group, we divided patients into the primary cohort (n = 212 in the early-phase group, n = 139 in the late-phase group) and the external independent validation cohort (n = 105 in the early-phase group, n = 36 in the late-phase group) according to the centers. We built two separate radiomics models in the two patient groups. Firstly, we proposed an automatic segmentation method to extract lung volume for radiomics feature extraction. Secondly, we applied several image preprocessing procedures to increase the reproducibility of the radiomics features: 1) applied a low-pass Gaussian filter before voxel resampling to prevent aliasing; 2) conducted ComBat to harmonize radiomics features per scanner; 3) tested the stability of the features in the radiomics signature by several image transformations, such as rotating, translating, and growing/shrinking. Thirdly, we used least absolute shrinkage and selection operator (LASSO) to build the radiomics signature (RadScore). Afterward, we conducted a Fine-Gray competing risk regression to build the clinical model and the clinic-radiomics signature (CrrScore). Finally, performances of the three prognostic signatures (clinical model, RadScore, and CrrScore) were estimated from the two aspects: 1) cumulative poor outcome probability prediction; 2) 28-day poor outcome prediction. We also did stratified analyses to explore the potential association between the CrrScore and the poor outcomes regarding different age, type, and comorbidity subgroups.Results: In the early-phase group, the CrrScore showed the best performance in estimating poor outcome (C-index = 0.850), and predicting the probability of 28-day poor outcome (AUC = 0.862). In the late-phase group, the RadScore alone achieved similar performance to the CrrScore in predicting poor outcome (C-index = 0.885), and 28-day poor outcome probability (AUC = 0.976). Moreover, the RadScore in both groups successfully stratified patients with COVID-19 into low- or high-RadScore groups with significantly different survival time in the training and validation cohorts (all P < 0.05). The CrrScore in both groups can also significantly stratify patients with different prognoses regarding different age, type, and comorbidities subgroups in the combined cohorts (all P < 0.05).Conclusions: This research proposed a non-invasive and quantitative prognostic tool for predicting poor outcome in patients with COVID-19 based on CT imaging. Taking the insufficient medical recourse into account, our study might suggest that the chest CT radiomics signature of COVID-19 is more effective and ideal to predict poor outcome in the late-phase COVID-19 patients. For the early-phase patients, integrating radiomics signature with clinical risk factors can achieve a more accurate prediction of individual poor prognostic outcome, which enables appropriate management and surveillance of COVID-19.

Radiomics-Based Prediction of Long-Term Treatment Response of Vestibular Schwannomas Following Stereotactic Radiosurgery.

Stereotactic radiosurgery (SRS) is one of the treatment modalities for vestibular schwannomas (VSs). However, tumor progression can still occur after treatment. Currently, it remains unknown how to predict long-term SRS treatment outcome. This study investigates possible magnetic resonance imaging (MRI)-based predictors of long-term tumor control following SRS. Retrospective cohort study. Tertiary referral center. Analysis was performed on a database containing 735 patients with unilateral VS, treated with SRS between June 2002 and December 2014. Using strict volumetric criteria for long-term tumor control and tumor progression, a total of 85 patients were included for tumor texture analysis. All patients underwent SRS and had at least 2 years of follow-up. Quantitative tumor texture features were extracted from conventional MRI scans. These features were supplied to a machine learning stage to train prediction models. Prediction accuracy, sensitivity, specificity, and area under the receiver operating curve (AUC) are evaluated. Gray-level co-occurrence matrices, which capture statistics from specific MRI tumor texture features, obtained the best prediction scores: 0.77 accuracy, 0.71 sensitivity, 0.83 specificity, and 0.93 AUC. These prediction scores further improved to 0.83, 0.83, 0.82, and 0.99, respectively, for tumors larger than 5 cm. Results of this study show the feasibility of predicting the long-term SRS treatment response of VS tumors on an individual basis, using MRI-based tumor texture features. These results can be exploited for further research into creating a clinical decision support system, facilitating physicians, and patients to select a personalized optimal treatment strategy.

Prognostic implications of autophagy-associated gene signatures in non-small cell lung cancer.

Autophagy, a highly conserved cellular proteolysis process, has been involved in non-small cell lung cancer (NSCLC). We tried to develop a prognostic prediction model for NSCLC patients based on the expression profiles of autophagy-associated genes. Univariate Cox regression analysis was used to determine autophagy-associated genes significantly correlated with overall survival (OS) of the TCGA lung cancer cohort. LASSO regression was performed to build multiple-gene prognostic signatures. We found that the 22-gene and 11-gene signatures could dichotomize patients with significantly different OS and independently predict the OS in TCGA lung adenocarcinoma (HR=2.801, 95% CI=2.252-3.486, P<0.001) and squamous cell carcinoma (HR=1.105, 95% CI=1.067-1.145, P<0.001), respectively. The prognostic performance of the 22-gene signature was validated in four GEO lung cancer cohorts. Moreover, GO, KEGG, and GSEA analyses unveiled several fundamental signaling pathways and cellular processes associated with the 22-gene signature in lung adenocarcinoma. We also constructed a clinical nomogram with a concordance index of 0.71 to predict the survival possibility of NSCLC patients by integrating clinical characteristics and the autophagy gene signature. The calibration curves substantiated fine concordance between nomogram prediction and actual observation. Overall, we constructed and verified a novel autophagy-associated gene signature that could improve the individualized outcome prediction in NSCLC.

Interaction of arsenic trioxide and etoposide in Ewing sarcoma cell lines.

Ewing sarcomas (ES) are highly malignant mesenchymal tumors, which most often occur in children and adolescents. The current treatment of choice comprises wide resection in combination with multimodal chemotherapy including etoposide (Eto). Due to the serious side effects associated with common chemotherapeutics and prevalent multidrug resistance in recurrent and metastatic ES, there is a growing demand for alternative strategies and add‑on drugs. Previous research has demonstrated efficient cell death induction by Eto in combination with arsenic trioxide (ATO) in ES cell lines. The aim of the present study was to investigate the effect of different temporal sequences of ATO and Eto administration on apoptosis induction and to explore the effect of both drugs on inhibitory glycogen synthase kinase‑3β (GSK3‑β) phosphorylation as well as multidrug transporter gene expression. The intensity of caspase activation was mainly determined by the Eto doses in A673 and TC‑71cells, whereas in RD‑ES cells ATO application actively suppressed Eto‑induced apoptosis. This coincided with an increase in inhibitory GSK‑3β phosphorylation in ATO‑treated RD‑ES cells. Inherent mRNA expression of multidrug resistance‑associated protein1 (MRP1) was low in the ES cell lines compared to that observed in the mesenchymal stem cells (MSC), whereas multidrug resistance protein1 (MDR1) gene expression was considerably increased in the ES cell lines. ATO treatment reduced MRP1 mRNA expression in the A673 and TC‑71cells, while expression was induced in the MSC and RD‑ES cells. In contrast, MDR1 mRNA expression was specifically induced by ATO in the A673 and TC‑71cells, reinforcing the expression differences between MSC and the ES cell lines. Although a reliable cell death induction by the combination of ATO and Eto has been previously shown in ES cell lines, the present study showed marked heterogeneity of the ES cell response to ATO and Eto treatment, illustrating the difficulty of prediction of individual treatment outcome in ES.

A Distance-Based Neurorehabilitation Evaluation Method Using Linear SVM and Resting-State fMRI.

During neurorehabilitation, clinical measurements are widely adopted to evaluate behavioral improvements after treatment. However, it is not able to identify or monitor the change of central nervous system (CNS) of each individual patient. Resting-state functional magnetic resonance imaging (rs-fMRI) has been widely used to investigate brain functions in healthy controls (HCs) and patients with neurological diseases, which could find functional changes following neurorehabilitation. In this paper, a distance-based rehabilitation evaluation method based on rs-fMRI was proposed. Specifically, we posit that in the functional connectivity (FC) space, patients and HCs distribute separately. Linear support vector machines (SVM) were trained on the brain networks to firstly separate patients from HCs. Second, the FC similarity between patients and HCs was measured by the L2 distance of each subject's feature vector to the separating hyperplane. Finally, statistical analysis of the distance revealed rehabilitation program induced improvements in patients and predicted rehabilitation outcomes. An rs-fMRI dataset with 22 HCs and 18 spinal cord injury (SCI) patients was utilized to validate our method. We built whole-brain networks using five atlases to test the robustness of the method and search for features under different node resolutions. The classifier successfully separated patients and HCs. Significant improvements in FC after treatment were found for the patients for all five atlases using the proposed method, which was consistent with clinical measurements. Furthermore, distance obtained from individual patient's longitudinal data showed a similar trend with each one's clinical scores, implying the possibility of individual rehabilitation outcome tracking and prediction. Our method not only provides a novel perspective of applying rs-fMRI to neurorehabilitation monitoring but also proves the potential in individualized rehabilitation prediction.

Machine learning versus traditional risk stratification methods in acute coronary syndrome: a pooled randomized clinical trial analysis.

Traditional statistical models allow population based inferences and comparisons. Machine learning (ML) explores datasets to develop algorithms that do not assume linear relationships between variables and outcomes and that may account for higher order interactions to make individualized outcome predictions. To evaluate the performance of machine learning models compared to traditional risk stratification methods for the prediction of major adverse cardiovascular events (MACE) and bleeding in patients with acute coronary syndrome (ACS) that are treated with antithrombotic therapy. Data on 24,178 ACS patients were pooled from four randomized controlled trials. The super learner ensemble algorithm selected weights for 23 machine learning models and was compared to traditional models. The efficacy endpoint was a composite of cardiovascular death, myocardial infarction, or stroke. The safety endpoint was a composite of TIMI major and minor bleeding or bleeding requiring medical attention. For the MACE outcome, the super learner model produced a higher c-statistic (0.734) than logistic regression (0.714), the TIMI risk score (0.489), and a new cardiovascular risk score developed in the dataset (0.644). For the bleeding outcome, the super learner demonstrated a similar c-statistic as the logistic regression model (0.670 vs. 0.671). The machine learning risk estimates were highly calibrated with observed efficacy and bleeding outcomes (Hosmer-Lemeshow p value = 0.692 and 0.970, respectively). The super learner algorithm was highly calibrated on both efficacy and safety outcomes and produced the highest c-statistic for prediction of MACE compared to traditional risk stratification methods. This analysis demonstrates a contemporary application of machine learning to guide patient-level antithrombotic therapy treatment decisions.Clinical Trial Registration ATLAS ACS-2 TIMI 46: https://clinicaltrials.gov/ct2/show/NCT00402597. Unique Identifier: NCT00402597. ATLAS ACS-2 TIMI 51: https://clinicaltrials.gov/ct2/show/NCT00809965. Unique Identifier: NCT00809965. GEMINI ACS-1: https://clinicaltrials.gov/ct2/show/NCT02293395. Unique Identifier: NCT02293395. PIONEER-AF PCI: https://clinicaltrials.gov/ct2/show/NCT01830543. Unique Identifier: NCT01830543.

Pre-operative Brain Imaging Using Functional Near-Infrared Spectroscopy Helps Predict Cochlear Implant Outcome in Deaf Adults

Currently, it is not possible to accurately predict how well a deaf individual will be able to understand speech when hearing is (re)introduced via a cochlear implant. Differences in brain organisation following deafness are thought to contribute to variability in speech understanding with a cochlear implant and may offer unique insights that could help to more reliably predict outcomes. An emerging optical neuroimaging technique, functional near-infrared spectroscopy (fNIRS), was used to determine whether a pre-operative measure of brain activation could explain variability in cochlear implant (CI) outcomes and offer additional prognostic value above that provided by known clinical characteristics. Cross-modal activation to visual speech was measured in bilateral superior temporal cortex of pre- and post-lingually deaf adults before cochlear implantation. Behavioural measures of auditory speech understanding were obtained in the same individuals following 6 months of cochlear implant use. The results showed that stronger pre-operative cross-modal activation of auditory brain regions by visual speech was predictive of poorer auditory speech understanding after implantation. Further investigation suggested that this relationship may have been driven primarily by the inclusion of, and group differences between, pre- and post-lingually deaf individuals. Nonetheless, pre-operative cortical imaging provided additional prognostic value above that of influential clinical characteristics, including the age-at-onset and duration of auditory deprivation, suggesting that objectively assessing the physiological status of the brain using fNIRS imaging pre-operatively may support more accurate prediction of individual CI outcomes. Whilst activation of auditory brain regions by visual speech prior to implantation was related to the CI user’s clinical history of deafness, activation to visual speech did not relate to the future ability of these brain regions to respond to auditory speech stimulation with a CI. Greater pre-operative activation of left superior temporal cortex by visual speech was associated with enhanced speechreading abilities, suggesting that visual speech processing may help to maintain left temporal lobe specialisation for language processing during periods of profound deafness.

Confirmation of the prognostic value of pretherapeutic tumor SUR and MTV in patients with esophageal squamous cell carcinoma.

The prognosis for patients with inoperable esophageal carcinoma is still poor and the reliability of individual therapy outcome prediction based on clinical parameters is not convincing. In a recent publication, we were able to show that PET can provide independent prognostic information in such a patient group and that the tumor-to-blood standard uptake ratio (SUR) can improve the prognostic value of tracer uptake values. The present investigation addresses the question of whether the distinctly improved prognostic value of SUR can be confirmed in a similar patient group that was examined and treated at a different site. 18F-FDG PET/CT was performed in 147 consecutive patients (115 male, 32 female, mean age: 62 years) with newly diagnosed esophageal squamous cell carcinoma prior to definitive radiochemotherapy. In the PET images, the metabolic active volume (MTV) of the primary tumor was delineated with an adaptive threshold method. For the resulting ROIs, SUVmax and total lesion glycolysis (TLG = MTV × SUVmean) were computed. The blood SUV was determined by manually delineating the aorta in the low-dose CT. SUR values were computed as ratio of tumor SUV and blood SUV. Univariate Cox regression and Kaplan-Meier analysis with respect to overall survival (OS), distant-metastases-free survival (DM), and locoregional control (LRC) was performed. Additionally, a multivariate Cox regression including clinically relevant parameters was performed. Univariate Cox regression revealed MTV, TLG, and SURmax as significant prognostic factors for OS. MTV as well as TLG were significant prognostic factors for LRC while SURmax showed only a trend for significance. None of the PET parameters was prognostic for DM. In univariate analysis, SUVmax was not prognostic for any of the investigated clinical endpoints. In multivariate analysis (T-stage, N-stage, MTV, and SURmax), MTV was an independent prognostic factor for OS and showed a trend for significance for LRC. SURmax was not an independent predictor for OS or LRC. When including the PET parameters separately in multivariate analysis, MTV as well as SURmax were prognostic factors for OS indicating that SURmax is independent from the clinical parameters but not from MTV. In addition, MTV was an independent prognostic factor for LRC in this separate analysis. Our study revealed a clearly improved prognostic value of tumor SUR compared to tumor SUV and confirms our previously published findings regarding OS. Furthermore, SUR delivers prognostic information beyond that provided by the clinical parameters alone, but does not add prognostic information beyond that provided by MTV in this patient group. Therefore, our results suggest that pretherapeutic MTV is the parameter of choice for PET-based risk stratification in the considered setting but further investigations are necessary to demonstrate that this suggestion is correct.

MACHINE LEARNING VERSUS TRADITIONAL RISK STRATIFICATION METHODS IN ACUTE CORONARY SYNDROME: A POOLED RANDOMIZED CLINICAL TRIAL ANALYSIS

Traditional prediction models allow population based inferences. Machine learning explores datasets to develop algorithms that do not assume linear relationships between variables and outcomes and that account for higher order interactions to make individualized outcome predictions. Data on 24,178

Multivariate neuroimaging based prediction of individual outcome to transcranial magnetic stimulation in depression

Multivariate neuroimaging based prediction of individual outcome to transcranial magnetic stimulation in depression

Uncertainty evaluation of image-based tumour control probability models in radiotherapy of prostate cancer using a visual analytic tool.

Functional imaging techniques provide radiobiological information that can be included into tumour control probability (TCP) models to enable individualized outcome predictions in radiotherapy. However, functional imaging and the derived radiobiological information are influenced by uncertainties, translating into variations in individual TCP predictions. In this study we applied a previously developed analytical tool to quantify dose and TCP uncertainty bands when initial cell density is estimated from MRI-based apparent diffusion coefficient maps of eleven patients. TCP uncertainty bands of 16% were observed at patient level, while dose variations bands up to 8 Gy were found at voxel level for an iso-TCP approach.

Shanghai Score: A Prognostic and Adjuvant Treatment-evaluating System Constructed for Chinese Patients with Hepatocellular Carcinoma after Curative Resection.

Background:For Chinese patients with hepatocellular carcinoma (HCC), surgical resection is the most important treatment to achieve long-term survival for patients with an early-stage tumor, and yet the prognosis after surgery is diverse. We aimed to construct a scoring system (Shanghai Score) for individualized prognosis estimation and adjuvant treatment evaluation.Methods:A multivariate Cox proportional hazards model was constructed based on 4166 HCC patients undergoing resection during 2001–2008 at Zhongshan Hospital. Age, hepatitis B surface antigen, hepatitis B e antigen, partial thromboplastin time, total bilirubin, alkaline phosphatase, γ-glutamyltransferase, α-fetoprotein, tumor size, cirrhosis, vascular invasion, differentiation, encapsulation, and tumor number were finally retained by a backward step-down selection process with the Akaike information criterion. The Harrell's concordance index (C-index) was used to measure model performance. Shanghai Score is calculated by summing the products of the 14 variable values times each variable's corresponding regression coefficient. Totally 1978 patients from Zhongshan Hospital undergoing resection during 2009–2012, 808 patients from Eastern Hepatobiliary Surgery Hospital during 2008–2010, and 244 patients from Tianjin Medical University Cancer Hospital during 2010–2011 were enrolled as external validation cohorts. Shanghai Score was also implied in evaluating adjuvant treatment choices based on propensity score matching analysis.Results:Shanghai Score showed good calibration and discrimination in postsurgical HCC patients. The bootstrap-corrected C-index (confidence interval [CI]) was 0.74 for overall survival (OS) and 0.68 for recurrence-free survival (RFS) in derivation cohort (4166 patients), and in the three independent validation cohorts, the CIs for OS ranged 0.70–0.72 and that for RFS ranged 0.63–0.68. Furthermore, Shanghai Score provided evaluation for adjuvant treatment choices (transcatheter arterial chemoembolization or interferon-α). The identified subset of patients at low risk could be ideal candidates for curative surgery, and subsets of patients at moderate or high risk could be recommended with possible adjuvant therapies after surgery. Finally, a web server with individualized outcome prediction and treatment recommendation was constructed.Conclusions:Based on the largest cohort up to date, we established Shanghai Score – an individualized outcome prediction system specifically designed for Chinese HCC patients after surgery. The Shanghai Score web server provides an easily accessible tool to stratify the prognosis of patients undergoing liver resection for HCC.