Adequately Field-Validate the Efficacy (Predictive Capability) of the Simultaneously Extracted Metals/Acid Volatile Sulfides (SEM/AVS) Method

BACKGROUND The collection and examination of social media has become a useful mechanism for studying the mental activity and behavior tendencies of users. OBJECTIVE Through the analysis of a collected set of Twitter data, a model will be developed for predicting positively referenced, drug-related tweets. From this, trends and correlations can be determined. METHODS Twitter social media tweets and attribute data were collected and processed using topic pertaining keywords, such as drug slang and use-conditions (methods of drug consumption). Potential candidates were preprocessed resulting in a dataset 3,696,150 rows. The predictive classification power of multiple methods was compared including regression, decision trees, and CNN-based classifiers. For the latter, a deep learning approach was implemented to screen and analyze the semantic meaning of the tweets. RESULTS The logistic regression and decision tree models utilized 12,142 data points for training and 1041 data points for testing. The results calculated from the logistic regression models respectively displayed an accuracy of 54.56% and 57.44%, and an AUC of 0.58. While an improvement, the decision tree concluded with an accuracy of 63.40% and an AUC of 0.68. All these values implied a low predictive capability with little to no discrimination. Conversely, the CNN-based classifiers presented a heavy improvement, between the two models tested. The first was trained with 2,661 manually labeled samples, while the other included synthetically generated tweets culminating in 12,142 samples. The accuracy scores were 76.35% and 82.31%, with an AUC of 0.90 and 0.91. Using association rule mining in conjunction with the CNN-based classifier showed a high likelihood for keywords such as “smoke”, “cocaine”, and “marijuana” triggering a drug-positive classification. CONCLUSIONS Predictive analysis without a CNN is limited and possibly fruitless. Attribute-based models presented little predictive capability and were not suitable for analyzing this type of data. The semantic meaning of the tweets needed to be utilized, giving the CNN-based classifier an advantage over other solutions. Additionally, commonly mentioned drugs had a level of correspondence with frequently used illicit substances, proving the practical usefulness of this system. Lastly, the synthetically generated set provided increased scores, improving the predictive capability. CLINICALTRIAL None

Download Full-text

Comparison of pretherapeutic osseous tumor volume and standard hematology for prediction of hematotoxicity after PSMA-targeted radioligand therapy

European Journal of Nuclear Medicine and Molecular Imaging ◽

10.1007/s00259-021-05412-1 ◽

2021 ◽

Author(s):

Liam Widjaja ◽

Rudolf A. Werner ◽

Tobias L. Ross ◽

Frank M. Bengel ◽

Thorsten Derlin

Keyword(s):

Multivariate Analysis ◽

Tumor Volume ◽

Predictive Performance ◽

Castration Resistant Prostate Cancer ◽

Operating Characteristics ◽

Predictive Capability ◽

Radioligand Therapy ◽

Castration Resistant ◽

Pet Ct ◽

Late Stages

Abstract Purpose Hematotoxicity is a potentially dose-limiting adverse event in patients with metastasized castration-resistant prostate cancer (mCRPC) undergoing prostate-specific membrane antigen (PSMA)-directed radioligand therapy (RLT). We aimed to identify clinical or PSMA-targeted imaging-derived parameters to predict hematological adverse events at early and late stages in the treatment course. Methods In 67 patients with mCRPC scheduled for 177Lu-PSMA-617 RLT, pretherapeutic osseous tumor volume (TV) from 68Ga-PSMA-11 PET/CT and laboratory values were assessed. We then tested the predictive capability of these parameters for early and late hematotoxicity (according to CTCAE vers. 5.0) after one cycle of RLT and in a subgroup of 32/67 (47.8%) patients after four cycles of RLT. Results After one cycle, 10/67 (14.9%) patients developed leukocytopenia (lymphocytopenia, 39/67 [58.2%]; thrombocytopenia, 17/67 [25.4%]). A cut-off of 5.6 × 103/mm3 for baseline leukocytes was defined by receiver operating characteristics (ROC) and separated between patients with and without leukocytopenia (P < 0.001). Baseline leukocyte count emerged as a stronger predictive factor in multivariate analysis (hazard ratio [HR], 33.94, P = 0.001) relative to osseous TV (HR, 14.24, P = 0.01). After four cycles, 4/32 (12.5%) developed leukocytopenia and the pretherapeutic leukocyte cut-off (HR, 9.97, P = 0.082) tended to predict leukocytopenia better than TV (HR, 8.37, P = 0.109). In addition, a cut-off of 1.33 × 103/mm3 for baseline lymphocytes separated between patients with and without lymphocytopenia (P < 0.001), which was corroborated in multivariate analysis (HR, 21.39, P < 0.001 vs. TV, HR, 4.57, P = 0.03). After four cycles, 19/32 (59.4%) developed lymphocytopenia and the pretherapeutic cut-off for lymphocytes (HR, 46.76, P = 0.007) also demonstrated superior predictive performance for late lymphocytopenia (TV, HR, 5.15, P = 0.167). Moreover, a cut-off of 206 × 103/mm3 for baseline platelets separated between patients with and without thrombocytopenia (P < 0.001) and also demonstrated superior predictive capability in multivariate analysis (HR, 115.02, P < 0.001 vs.TV, HR, 12.75, P = 0.025). After four cycles, 9/32 (28.1%) developed thrombocytopenia and the pretherapeutic cut-off for platelets (HR, 5.44, P = 0.048) was also superior for the occurrence of late thrombocytopenia (TV, HR, 1.44, P = 0.7). Conclusions Pretherapeutic leukocyte, lymphocyte, and platelet levels themselves are strong predictors for early and late hematotoxicity under PSMA-directed RLT, and are better suited than PET-based osseous TV for this purpose.

Download Full-text

Free Energy of Metals from Quasi-Harmonic Models of Thermal Disorder

Metals ◽

10.3390/met11020195 ◽

2021 ◽

Vol 11 (2) ◽

pp. 195

Author(s):

Pavel A. Korzhavyi ◽

Jing Zhang

Keyword(s):

Free Energy ◽

First Principles ◽

Density Functional ◽

Helmholtz Free Energy ◽

Complex Materials ◽

Predictive Capability ◽

Temperature Dependent ◽

Disordered Alloys ◽

Thermal Disorder ◽

Structure Calculations

A simple modeling method to extend first-principles electronic structure calculations to finite temperatures is presented. The method is applicable to crystalline solids exhibiting complex thermal disorder and employs quasi-harmonic models to represent the vibrational and magnetic free energy contributions. The main outcome is the Helmholtz free energy, calculated as a function of volume and temperature, from which the other related thermophysical properties (such as temperature-dependent lattice and elastic constants) can be derived. Our test calculations for Fe, Ni, Ti, and W metals in the paramagnetic state at temperatures of up to 1600 K show that the predictive capability of the quasi-harmonic modeling approach is mainly limited by the electron density functional approximation used and, in the second place, by the neglect of higher-order anharmonic effects. The developed methodology is equally applicable to disordered alloys and ordered compounds and can therefore be useful in modeling realistically complex materials.

Download Full-text